2012 Asco

AMERICAN SOCIETY OF CLINICAL ONCOLOGY
2012 EDUCATIONAL BOOK

48th Annual Meeting | June 1-5, 2012 | Chicago, Illinois
American Society of Clinical Oncology

Educational Book
The American Society of Clinical Oncology Educational Book (ISSN 1548-8748) is published by the American Society of
Clinical Oncology.
Requests for permission to reprint all or part of any article published in this title should be directed to Permissions,
American Society of Clinical Oncology, 2318 Mill Road, Suite 800, Alexandria, VA 22314. Tel: 571-483-1300; fax:
571-366-9550; or email: permissions@asco.org.
All other questions should be addressed to Managing Editor, American Society of Clinical Oncology, 2318 Mill Road, Suite
800, Alexandria, VA 22314. Tel: 703-299-1183; email: EdBook@asco.org.
Single issues, both current and back, exist in limited quantities and are offered for sale subject to availability. For further
information, email EdBook@asco.org or call 888-273-3508.
Copyright 2012 American Society of Clinical Oncology. All rights reserved. No part of this publication may be reproduced
or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information
storage and retrieval system, without permission in writing from ASCO or the author.
Copies of articles in this publication may be made for personal use. This consent is given on the condition, however, that
the copier pay the stated per-copy fee through the Copyright Clearance Center, Inc. (222 Rosewood Drive, Danvers, MA
01923) for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to
other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new
collective works, or for resale.
The American Society of Clinical Oncology assumes no responsibility for errors or omissions in this publication. The reader
is advised to check the appropriate medical literature and the product information currently provided by the manufacturer
of each drug to be administered to verify the dosage, the method and duration of administration, or contraindications. It is
the responsibility of the treating physician or other health care professional, relying on the independent experience and
knowledge of the patient, to determine drug dosages and the best treatment for the patient.
American Society of Clinical

Oncology
Educational Book
Editor: Ramaswamy Govindan, MD

Associate Editor: Natasha Leighl, MD
Managing Editor: Christine Smith
Editorial Coordinator: Lauren Burke
Production Manager: Donna Dottellis
Executive Editor: Lisa Greaves
2012 American Society of Clinical Oncology

Alexandria, VA
ASCO gratefully acknowledges Amgen

for their support of the 48th Annual Meeting Educational Book.

Oncology
Educational Book
TABLE OF CONTENTS
ASCO Annual Meeting Disclosure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
20112012 Cancer Education Committee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv
2012 Annual Meeting Faculty List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv
2012 Educational Book Expert Panel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xx
2012 Annual Meeting Supporters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxii
Letter from the Editor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
The 2012 ASCO Educational Book is published online at asco.org/edbook. Articles can also be accessed
from the Attendee Resource Center at chicago2012.asco.org/attendee. Articles that are included in this
print edition include a page number reference in the table of contents below.
Breast Cancer
Controversies in the Adjuvant Treatment of Breast Cancer
Adjuvant Therapy for Older Women with Early-Stage Breast Cancer: Treatment Selection in a Complex Population
Cynthia Owusu, Arti Hurria, and Hyman Muss. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Management of Small T1a/b N0 Breast Cancers
Anthony D. Elias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Controversies in Adjuvant Endocrine Therapy for Breast Cancer
Stephen R. Johnston. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
A Dickens Tale of the Treatment of Advanced Breast Cancer: The Past, the Present, and the Future
A Dickens Tale of the Treatment of Advanced Breast Cancer: The Past, the Present, and the Future
George W. Sledge Jr., Fatima Cardoso, Eric P. Winer, and Martine J. Piccart. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
A Fresh Look at Ductal Carcinoma In Situ: Basic Science to Clinical Management
Ductal Carcinoma In Situ: Challenges, Opportunities, and Uncharted Water
Abigail W. Hoffman, Catherine Ibarra-Drendall, Virginia Espina, Lance Liotta, and Victoria Seewaldt . . . . . . . . . . . . . . . . 40
Ductal Carcinoma In Situ, and the Influence of the Mode of Detection, Population Characteristics, and Other Risk
Factors
Beth A. Virnig, Shi-Yi Wang, and Todd M. Tuttle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Key Questions in the Loco-regional Treatment of Breast Cancer
Postmastectomy Radiation and Partial Breast Irradiation
Richard C. Zellars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
The Appropriate Extent of Surgery for Early-Stage Breast Cancer
Monica Morrow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
How Does Biology Affect Local Therapy Decisions?

Thomas A. Buchholz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
The Many Options for Breast Imaging: What to Use When
Clinical and Imaging Surveillance Following Breast Cancer Diagnosis
Chris I. Flowers, Blaise P. Mooney, and Jennifer S. Drukteinis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Advanced Imaging Techniques for the Detection of Breast Cancer
Maxine Jochelson. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Update of the Oxford Overview: New Insights and Perspectives in the Era of Personalized Medicine
Update of the Oxford Overview: New Insight and Perspectives in the Era of Personalized Medicine
Kathleen I. Pritchard, Jonas Bergh, and Harold J. Burstein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Cancer Genetics
Gene Patenting: Effects on Biotechnology and Oncology
Gene Patents and Personalized Medicine
Roger D. Klein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Cancer Prevention/Epidemiology
Breast Cancer Chemoprevention: If Not Now, When?
Chemoprevention for Breast Cancer: Overcoming Barriers to Treatment
Abenaa M. Brewster, Nancy E. Davidson, and Worta McCaskill-Stevens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Controversies in Prostate Cancer: PSA Screening, Chemoprevention, and Treatment of Early Disease
The Case for Prostate Cancer Risk Reduction by 5-Alpha Reductase Inhibitors
Youssef S. Tanagho and Gerald L. Andriole . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Screening for Prostate Cancer with Prostate-Specific Antigen: Whats the Evidence?
Pamela M. Marcus and Barnett S. Kramer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Central Nervous System Tumors
Glioblastoma: Taking the Standard of Care to the Cutting Edge
Glioblastoma: Biology, Genetics, and Behavior
Daniel J. Brat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Future Directions in Glioblastoma Therapy
Howard Colman. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Establishing the Standard of Care for Patients with Newly Diagnosed and Recurrent Glioblastoma
Mark R. Gilbert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Imaging in Neuro-oncology: Practical Primer for the Practicing Physician
Current Concepts in Brain Tumor Imaging
Andrew D. Norden, Whitney B. Pope, and Susan M. Chang . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Perspectives on Headline-Making News in Neuro-oncology
Noninvasive Application of Alternating Electric Fields in Glioblastoma: A Fourth Cancer
Treatment Modality
Philip H. Gutin and Eric T. Wong . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Clinical Trials
Adaptive Clinical Trial Designs: What Are They and Should They Be Used?
Limitations of Adaptive Clinical Trials
Marc Buyse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Endpoints for Cancer Trials in 2012: Statistical, Regulatory, and Clinical Perspectives (eQ&A)
Capturing the Patient Perspective: Patient-Reported Outcomes as Clinical Trial Endpoints
Deborah Watkins Bruner, Benjamin Movsas, and Ethan Basch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
New Looks and Challenges for Cooperative Groups

The New National Cancer Institute National Clinical Trials Network
Stephen S. Grubbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
Successful Integration of Cooperative Groups: The Origin of the Childrens Oncology Group
Gregory H. Reaman. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Overcoming Disparities in Clinical Trial Accrual among African Americans
A Critical Review of the Enrollment of Black Patients in Cancer Clinical Trials
Deliya R. Banda, Diane St. Germain, Worta McCaskill-Stevens, Jean G. Ford, and Sandra M. Swain . . . . . . . . . . . . . . . . 153
Developmental Therapeutics
Antibody Conjugates: Targeted Therapy Meets Chemotherapy in One Drug
Trastuzumab Emtansine (T-DM1): Hitching a Ride on a Therapeutic Antibody
Howard A. Burris III . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Targeting CD30 in Hodgkin Lymphoma: Antibody-Drug Conjugates Make a Difference
Catherine S. M. Diefenbach and John P. Leonard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Early Drug Development: Casting a Wide Net versus Preselecting a Narrow Audience
Approval of New Agents after Phase II Trials
Bruce Chabner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Drug Development in the Era of Personalized Oncology: From Population-Based Trials to Enrichment and
Prescreening Strategies
Rodrigo Dienstmann, Jordi Rodon, and Josep Tabernero. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
Immunotherapy with the T-cell Checkpoint Antibodies: Implications for the Practitioner
Practical Management of Immune-Related Adverse Events from Immune Checkpoint Protein Antibodies for
the Oncologist
Jeffrey S. Weber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Targeting Critical Molecular Aberrations Early in the Course of Solid Tumors: Is It about Time?
Treatment of Chronic Myelogenous Leukemia as a Paradigm for Solid Tumors: How Targeted Agents in Newly Diagnosed
Disease Transformed Outcomes
Jason R. Westin, Hagop Kantarjian, and Razelle Kurzrock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
Targeting Molecular Aberrations in Breast Cancer: Is It about Time?
Laura Esserman, Christopher Benz, and Angela DeMichele . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
Ethics
Ethical Challenges of Health Care
Core Elements of the Patient Protection and Affordable Care Act and Their Relevance to the Delivery of
High-Quality Cancer Care
Beverly Moy, Amy P. Abernethy, and Jeffrey M. Peppercorn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
International Variation in Understanding Oncologists Professional Duties toward Patients, Families, and Themselves
A Balanced Approach to Physician Responsibilities: Oncologists Duties toward Themselves
Amy P. Abernethy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Are Oncologists Accountable Only to Patients or Also to Their Families? An International Perspective
Antonella Surbone and Lea Baider . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
The Oncologists Duty to Provide Hope: Fact or Fiction?
Simon Wein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Medical Errors in Cancer Care: Prevention, Disclosure, and Patient and Family Member Responses
Oncologists Difficulties in Facing and Disclosing Medical Errors: Suggestions for the Clinic
Antonella Surbone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Preventing Errors in Oncology: Perspective of a Physician Who Is Also a Cancer Patient
Itzhak Brook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Gastrointestinal (Colorectal) Cancer

Personalized Oncology for Colorectal Cancer: Ready for Prime Time or Stop the Train?
Is There Currently an Established Role for the Use of Predictive or Prognostic Molecular Markers in the Management
of Colorectal Cancer? A Point/Counterpoint
Alan P. Venook, Johanna C. Bendell, and Robert S. Warren . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Curative-Intent Treatment of Colorectal Cancer Metastases
The Interventional Radiologist Role in Treating Liver Metastases for Colorectal Cancer
Stephen B. Solomon and Constantinos T. Sofocleous . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
Curative-Intent Treatment for Colorectal Liver Metastases: A Medical Oncologists Perspective
Leonard B. Saltz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
Surgical Treatment of Colorectal Cancer Liver Metastases
Steven A. Curley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
Rectal Cancer: New Paradigms beyond Standard Chemotherapy and Radiation
Minimally Invasive Surgery of Rectal Cancer: Current Evidence and Options
Atthaphorn Trakarnsanga and Martin R. Weiser. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
Radiation in Rectal Cancer: What Are the Options and If/When Can It Be Avoided?
Karyn A. Goodman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Stage III Colon Cancer: What Works, What Doesnt and Why, and Whats Next (eQ&A)
Stage III Colon Cancer: What Works, What Doesnt and Why, and Whats Next
Thierry Andre, Bert H. ONeil, and Jeffrey A. Meyerhardt. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Gastrointestinal (Noncolorectal) Cancer
Evolving Paradigms in the Management of Unresectable Pancreatic Cancer
A New Direction for Pancreatic Cancer Treatment: FOLFIRINOX in Context
Hedy Lee Kindler. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
A Matter of Timing: Is There a Role for Radiation in Locally Advanced Pancreatic Cancer, and If So, When?
Theodore S. Hong, Jennifer Y. Wo, and Eunice L. Kwak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
A Myriad of Symptoms: New Approaches to Optimizing Palliative Care of Patients with Advanced Pancreatic Cancer
Lauren A. Wiebe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Global Perspective of Locally Advanced Gastric Cancer: Different Treatment Paradigms and Their Rationale
Varying Lymphadenectomies for Gastric Adenocarcinoma in the East Compared with the West: Effect on Outcomes
Benjamin Schmidt and Sam S. Yoon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
Will Disease Heterogeneity Help Define Treatment Paradigms for Gastroesophageal Adenocarcinoma? A Global
Perspective
Manish A. Shah . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
Adjuvant Treatments for Localized Advanced Gastric Cancer: Differences among Geographic Regions
Yoon-Koo Kang and Changhoon Yoo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Liver-Directed Therapeutic Options for Hepatocellular Carcinoma: Patient Selection and Clinical Outcomes
Stereotactic Body Radiation Therapy for Hepatocellular Carcinoma
Laura A. Dawson, Sameh Hashem, and Alexis Bujold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Patient Selection, Resection, and Outcomes for Hepatocellular Carcinoma
Claudius Conrad and Kenneth K. Tanabe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
The Management of Less Common but Complex Upper Gastrointestinal Malignancies: Hepatocellular Carcinoma,
Pancreatic Neuroendocrine Tumors, and Biliary Tract Tumors
A Renaissance in Therapeutic Options for Pancreatic Neuroendocrine Tumors
Pamela L. Kunz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
A Review of the Management of Hepatocellular Carcinoma: Standard Therapy and a Look to New Targets
Andrew X. Zhu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Advanced Biliary Tract Cancers
Laura Williams Goff and Jordan D. Berlin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
Genitourinary Cancer
Best Use of Imaging Techniques, Old and New, for Genitourinary Cancers in Clinical Practice
Optimal Use of Imaging to Guide Treatment Decisions for Kidney Cancer
Walter M. Stadler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
Castration-Resistant Prostate Cancer: New Insights into Biology and Treatment (eQ&A)
State-of-the-Art Management for the Patient with Castration-Resistant Prostate Cancer in 2012
Oliver Sartor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Prognostic, Predictive, and Surrogate Factors for Individualizing Treatment for Men with Castration-Resistant Prostate
Cancer
Rhonda L. Bitting and Andrew J. Armstrong . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
Kidney Cancer Biology and Therapeutics: VEGFR, mTOR, and Beyond
The Evolving Landscape of Metastatic Renal Cell Carcinoma
Daniel Y. C. Heng and Toni K. Choueiri . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
Recent Innovations in the Management of Genitourinary Cancers (eQ&A)
New Developments in Urothelial Cancer
Matthew D. Galsky . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
New Developments in Prostate Cancer Therapy
Madhuri Bajaj and Elisabeth I. Heath . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
Geriatric Oncology
Adjuvant Therapy for Older Patients
Adjuvant Treatment of Older Patients with Lung Cancer
Jeffrey Crawford . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Treatment of Older Patients with Advanced Cancer: Balancing Efficacy with Toxicity (eQ&A)
Considerations and Controversies in the Management of Older Patients with Advanced Cancer
Supriya Gupta Mohile, Heidi D. Klepin, and Arati V. Rao. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
Gynecologic Cancer
Recent Clinical Highlights in Gynecologic Oncology
International Perspective on the Global Advances in Gynecologic Oncology
Lynette Denny . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330
The European Society of Gynaecological Oncology: Update on Objectives and Educational and Research Activities
Renata Brantnerova, Ranjit Manchanda, and Nicoletta Colombo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
Upfront Treatment of Ovarian Cancer: International Consensus and Variation
Biologicals in the Upfront Treatment of Ovarian Cancer: Focus on Bevacizumab and Poly (ADP-Ribose) Polymerase
Inhibitors
Rene Roux, Martin Zweifel, and Gordon J. S. Rustin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
Intraperitoneal Treatment in Ovarian Cancer: The Gynecologic Oncology Group Perspective in 2012
Deborah K. Armstrong, Keiichi Fujiwara, and Danijela Jelovac . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
Dose-Dense Chemotherapy and Neoadjuvant Chemotherapy for Ovarian Cancer
Keiichi Fujiwara, Noriyuki Katsumata, and Takashi Onda. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
Uterine Sarcoma: Challenging Cases for the Interdisciplinary Team
Uterine Sarcomas: Histology and Its Implications on Therapy
Martee L. Hensley . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
Surgical Options for Recurrent Uterine Sarcomas
Sharmilee B. Korets and John P. Curtin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
Head and Neck Cancer

Patients with HPV-Positive Oropharynx Tumors: Communicating Diagnosis to Family and Treatment Options
Management of Human Papillomavirus-Induced Oropharynx Cancer
David Brizel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
Translating Biologic Discoveries into Clinical Trials
Important Early Advances in Squamous Cell Carcinoma of the Head and Neck
Eric Bissada, Irene Brana, and Lillian L. Siu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
Application of Genomic and Proteomic Technologies in Biomarker Discovery
Elana J. Fertig, Robbert Slebos, and Christine H. Chung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
Treatment of Thyroid Cancers: New Information for Medical Oncologists
Evaluation of Patients with Disseminated or Locoregionally Advanced Thyroid Cancer: A Primer for
Medical Oncologists
A. Dimitrios Colevas and Manisha H. Shah . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Systemic Therapeutic Approaches to Advanced Thyroid Cancers
Michael E. Menefee, Robert C. Smallridge, and Keith C. Bible, and the Mayo Clinic Endocrine Malignancies
Disease-Oriented Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
Health Services Research
Avoiding Overdiagnosis and Overtreatment of Common Cancers
Overdiagnosis and Overtreatment of Prostate Cancer
Ian M. Thompson Jr. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Overdiagnosis and Overtreatment of Breast Cancer
Michael Alvarado, Elissa Ozanne, and Laura Esserman . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Costs of Cancer Care: Affordability, Access, and Policy
Reducing the Cost of Cancer Care: How to Bend the Curve Downward
Thomas J. Smith, Bruce E. Hillner, and Ronan J. Kelly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
New Diagnostics and Devices in the Era of Comparative Effectiveness
Why Hasnt Genomic Testing Changed the Landscape in Clinical Oncology?
Daniel F. Hayes, Muin J. Khoury, and David Ransohoff. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Leukemia, Myelodysplasia, and Transplantation
Management of Chronic Lymphocytic Leukemia
Predicting Clinical Outcome in B-Chronic Lymphocytic Leukemia
Neil E. Kay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
Transplant in Chronic Lymphocytic Leukemia: To Do It or Not and If So, When and How?
John G. Gribben . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
New Developments in Myeloproliferative Neoplasms
Therapeutic Advances in Myeloproliferative Neoplasms: The Role of New-Small Molecule Inhibitors
Srdan Verstovsek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
Insights into the Molecular Genetics of Myeloproliferative Neoplasms
Huong (Marie) Nguyen and Jason Gotlib . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
Classification of Myeloproliferative Neoplasms and Prognostic Factors
Francesco Passamonti . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
Lung Cancer
Around the World in Almost 80 Minutes: Lung Cancer Care and Research
Lung Cancer in Brazil
Gilberto Schwartsmann . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
Research and Standard Care: Lung Cancer in China
Tony S. Mok, Qing Zhou, and Yi-Long Wu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
Research and Standard of Care: Lung Cancer in Romania

Tudor E. Ciuleanu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
Leveraging Virtual Patient Communities for Optimal Clinical Care and Research
A New Model: Physician-Patient Collaboration in Online Communities and the Clinical Practice of Oncology
Howard J. West, Dave deBronkart, and George D. Demetri . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443
Lung Cancer Screening 101
Lung Cancer Screening: Promise and Pitfalls
Christine D. Berg, Denise R. Aberle, and Douglas E. Wood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
Personalized Medicine in Lung Cancer in 2012
Molecular Testing of NonSmall Cell Lung Carcinoma Biopsy and Cytology Specimens
Ignacio I. Wistuba . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
Thymoma and Thymic Carcinoma: Update on Management
Multidisciplinary Management of Thymic Carcinoma
Gregory J. Riely, James Huang, and Andreas Rimner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
The Creation of the International Thymic Malignancies Interest Group as a Model for Rare Diseases
Frank C. Detterbeck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
Thymoma: From Chemotherapy to Targeted Therapy
Nicolas Girard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
Lymphoma and Plasma Cell Disorders
Controversies in Indolent Lymphoma
What Is the Best Strategy for Incorporating New Agents into the Current Treatment of Follicular Lymphoma?
Sonali M. Smith . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 481
What Is the Best First-Line Treatment Strategy for Patients with Indolent Lymphomas?
Gilles Salles, Herve Ghesquie`res, and Emmanuel Bachy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
What Is the Role of Transplantation for Indolent Lymphoma?
Ryan D. Cassaday and Ajay K. Gopal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
Controversies in Myeloma: Induction, Transplant, and Maintenance
Stem Cell Transplantation for Multiple Myeloma: Who, When, and What Type?
Amrita Krishnan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502
Upfront Therapy for Myeloma: Tailoring Therapy across the Disease Spectrum
S. Vincent Rajkumar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508
Maintenance Therapy for Myeloma: How Much, How Long, and at What Cost?
Michel Attal and Murielle Roussel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515
Melanoma/Skin Cancers
New Options, New Questions: How to Select and Sequence Therapies for Metastatic Melanoma
New Options and New Questions: How to Select and Sequence Therapies for Patients with Metastatic Melanoma
Keith T. Flaherty, Jeff A. Sosman, and Michael B. Atkins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524
Patient and Survivor Care
Antiemetics: Current Standards, Emerging Approaches, and Persistent Gaps
Antiemetic Use in Oncology: Updated Guideline Recommendations from ASCO
Ethan Basch, Ann Alexis Prestrud, Paul J. Hesketh, Mark G. Kris, Mark R. Somerfield, and Gary H. Lyman . . . . . . . . . . 532
Chemotherapy-Induced Nausea and Vomiting Incidence and Prevalence
Steven M. Grunberg. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 541
Mucosal Injury in Patients with Cancer: Targeting the Biology
New Frontiers in Mucositis
Douglas E. Peterson, Dorothy M. Keefe, and Stephen T. Sonis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
Short- and Long-term Cardiovascular Complications of Cancer Treatment: Overview for the Practicing Oncologist
Short- and Long-term Cardiovascular Complications of Cancer Treatment: Overview for the Practicing Oncologist
Chetan Shenoy and Gretchen Kimmick . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553
Recent Advances in Cardiotoxicity of Anticancer Therapies
Marianne Ryberg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 555
When Cancer Becomes Personal: The Physicians View of Patient Care
The Oncologist as the Patient with Cancer or Relative
Teresa Gilewski, Martin Raber, and George W. Sledge Jr. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
Pediatric Oncology
Advances in Infection Management in Pediatric Oncology
Prolonged Febrile Neutropenia in the Pediatric Patient with Cancer
Andrew Y. Koh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565
Initial Management of Low-Risk Pediatric Fever and Neutropenia: Efficacy and Safety, Costs, Quality-of-Life
Considerations, and Preferences
Lillian Sung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
Genetic Counseling of the Patient with Pediatric Cancer
Identification, Management, and Evaluation of Children with Cancer-Predisposition Syndromes
Sara Knapke, Kristin Zelley, Kim E. Nichols, Wendy Kohlmann, and Joshua D. Schiffman . . . . . . . . . . . . . . . . . . . . . . . . . 576
How to Manage Very Rare Pediatric Cancers
Treating Rare Cancer in Children: The Importance of Evidence
Alberto S. Pappo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 586
Genetic Alterations in Childhood Melanoma
Fariba Navid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589
Desmoid-Type Fibromatosis in Children: A Step Forward in the Cooperative Group Setting
Natalie Pounds and Stephen X. Skapek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
Molecular Pathways for the Practicing Pediatric Oncologist
Targeting the Insulin-Like Growth Factor (IGF) System Is Not as Simple as Just Targeting the Type 1 IGF Receptor
Katia Scotlandi and Antonino Belfiore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
Hedgehog Pathway in Pediatric Cancers: Theyre Not Just for Brain Tumors Anymore
Tobey J. MacDonald . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
Pediatric Oncology Educational Session in Honor of Dr. James Nachman
Development and Refinement of Augmented Treatment Regimens for Pediatric High-Risk Acute Lymphoblastic Leukemia
Stephen P. Hunger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611
Refining the Role of Radiation Therapy in Pediatric Hodgkin Lymphoma
Melissa M. Hudson and Louis S. Constine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 616
Role of Doxorubicin in Rhabdomyosarcoma: Is the Answer Knowable?
Carola A. S. Arndt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
Pontine Gliomas in Children: To Biopsy or Not to Biopsy
Identification of Novel Biologic Targets in the Treatment of Newly Diagnosed Diffuse Intrinsic Pontine Glioma
Nathan J. Robison and Mark W. Kieran . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625
Is Biopsy Safe in Children with Newly Diagnosed Diffuse Intrinsic Pontine Glioma?
Stephanie Puget, Thomas Blauwblomme, and Jacques Grill. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 629
Ethics of Biopsy in Children with Newly Diagnosed Diffuse Intrinsic Pontine Glioma
Nicholas K. Foreman. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 634
Transition and Decision Making for Palliative Care of Patients with Pediatric Cancer
Communication and Decision Support for Children with Advanced Cancer and Their Families
Jennifer W. Mack, Chris Feudtner, and Pamela S. Hinds. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 637
Practice Management and Information Technology

Addressing the Imbalance of Supply and Demand: Integrating Advanced Practice Providers into Survivorship Care
Planning for the Future: The Role of Nurse Practitioners and Physician Assistants in Survivorship Care
Mary S. McCabe and Todd Alan Pickard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Doing It Right, and for Less: Implementing Practice Changes to Manage the Growing Complexities, Inefficiencies,
and Costs of Cancer Care
Driving Evidence-Based Standardization of Care within a Framework of Personalized Medicine
Adam Brufsky, Kathleen Lokay, and Melissa McDonald . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Patient Portals in Oncology
The Future of Oncology Care with Personal Health Records
Henry Feldman and Elizabeth S. Rodriguez . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
The ASCO Quality Oncology Practice Initiative (QOPI) and Beyond
Measuring and Improving Value of Care in Oncology Practices: ASCO Programs from Quality Oncology Practice
Initiative to the Rapid Learning System
Joseph O. Jacobson, Michael N. Neuss, and Robert Hauser . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Professional Development
Physician Wellness: Coping with Repetitive Loss and Work-Related Stress
Buoyancy: A Model for Self-Reflection about Stress and Burnout in Oncology Providers
Michael J. Fisch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Encountering Grief in Patient Care
Teresa Gilewski. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Talking in Sync with Patients across Cultural Barriers
New Insights in Cross-Cultural Communication
Lidia Schapira. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
The Oncologist, the Patient, and the Media
Patients with Cancer, Internet Information, and the Clinical Encounter: A Taxonomy of Patient Users
Paul R. Helft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
Sarcoma
Global Variations in Access to New Therapies for Sarcomas and Gastrointestinal Stromal Tumor
How to Decide Whether to Offer and Use Nonstandard Therapies in Patients with Advanced Sarcomas and
Gastrointestinal Stromal Tumors: Global Variations in Clinical Practice, Assessment, and Access to Therapies in
Diseases with Limited Incidence and Data
Stefan Sleijfer, Ian Judson, and George D. Demetri . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645
Targeted Therapies in Targeted or Untargeted Sarcomas
Targeted Therapy in Sarcoma: Should We Be Lumpers or Splitters?
Richard F. Riedel, Robert G. Maki, and Andrew J. Wagner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652
What the Busy Oncologist Needs to Know about Gastrointestinal Stromal Tumor
Adjuvant Treatment of Gastrointestinal Stromal Tumor: The Proof, the Pro, and the Practice
Jaap Verweij . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
Management of Tyrosine Kinase InhibitorResistant Gastrointestinal Stromal Tumors
Christine M. Barnett and Michael C. Heinrich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
Tumor Biology
Biologic Principles of Targeted Combination Therapy
Opportunities and Pitfalls of Targeted Therapeutic Combinations in Solid Tumors
Joaquin Mateo, Michael Ong, Timothy A. Yap, and Johann S. de Bono . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
Combination Therapies Building on the Efficacy of CTLA4 and BRAF Inhibitors for Metastatic Melanoma
Antoni Ribas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 675
Mechanisms of Resistance to Targeted Anticancer Agents
Mechanisms of Resistance to Mitogen-Activated Protein Kinase Pathway Inhibition in BRAF-Mutant Melanoma
Eva M. Goetz and Levi A. Garraway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680
Mechanisms of Resistance to Targeted Therapies in Acute Myeloid Leukemia and Chronic Myeloid Leukemia
Catherine C. Smith and Neil P. Shah . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685
Toward Successful Targeting of the PI3 Kinase Pathway in Cancer
Translating PI3K-Delta Inhibitors to the Clinic in Chronic Lymphocytic Leukemia: The Story of CAL-101 (GS1101)
John C. Byrd, Jennifer A. Woyach, and Amy J. Johnson . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691
PI3 Kinase in Cancer: From Biology to Clinic
Paul Workman and Paul Clarke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . eArticle
2012 ASCO Annual Meeting Disclosure

In compliance with the ASCO Conflict of Interest Policy and the Accreditation Council for Continuing Medical
Education (ACCME) Standards for Commercial Support, authors and faculty who participate in any ASCOsponsored meeting must disclose financial interests and other relationships with entities that have an
investment, licensing, or other commercial interest in the subject matter under consideration in their paper
or presentation. Author and faculty disclosures may be published in ASCO publications where appropriate.
Definition of notations: L Leadership Position; I Immediate Family Member; B Myself and immediate
Family Member; U Uncompensated.
Authors and faculty will be asked to disclose financial interests and other relationships in the following
categories:
Employment or Leadership Positions

Any full- or part-time employment or service as an officer or board member for an entity having an investment,
licensing, or other commercial interest in the subject matter under consideration must be disclosed.
Advisory Role
Consultant or advisory arrangements with an entity having an investment, licensing, or other commercial
interest in the subject matter under consideration must be disclosed if consultation was performed or payments
made for such consultation within 2 years of the activity or subject matter in question.
Stock Ownership
Any ownership interest (except when invested in a diversified fund not controlled by the covered individual) in
a start-up company, the stock of which is not publicly traded, or in any publicly traded company must be disclosed
if the company is an entity having an investment, licensing, or other commercial interest in the subject matter
under consideration.
Honoraria
Honoraria are reasonable payments for specific speeches, seminar presentations, or appearances. Disclosure of
honoraria is required when paid directly to the covered individual by an entity having an investment, licensing,
or other commercial interest in the subject matter under consideration and when provided within 2 years of the
activity or subject matter in question.
Research Funding
All payments associated with the conduct of the clinical research project in question must be disclosed if provided
by the trial sponsor or agents employed by the sponsor.
Expert Testimony
Provision of expert testimony must be disclosed when the testimony relates to the subject matter under
consideration.
Other Remuneration
The value of trips, travel, gifts, or other in-kind payments not directly related to research activities must be
disclosed if received from an entity having an investment, licensing, or other commercial interest in the subject
matter under consideration and when received within 2 years of the activity or subject matter in question. These
payments exclude research-related costs and travel.
Continuing Medical Education Information

This activity has been approved for AMA PRA Category 1 Credit. For information detailing the
Continuing Medical Education components of the 48th ASCO Annual Meeting, please visit
chicago2012.asco.org.
20112012 Cancer Education Committee

Harold J. Burstein MD, PhD, Chair
Antoinette R. Tan, MD, Chair-Elect
Charles Blanke, MD, Immediate Past
Chair
Gary I. Cohen, MD, Board Liaison
Breast Cancer
Anna Maria Storniolo, MD (Track
Leader)
Angelo Di Leo, MD, PhD
Stephen R. Grobmyer, MD
Erica L. Mayer, MD, MPH
Karen A. Gelmon, MD
Jonas Bergh, MD, PhD
Barbara A. Parker, MD
Richard C. Zellars, MD
Cancer Genetics
Sancy Ann Leachman, MD (Track
Leader)
James M. Ford, MD
Kenneth Offit, MD, MPH
Michael J. Hall, MD
Zsofia Kinga Stadler, MD
Cancer Prevention/Epidemiology
Pamela Jean Goodwin, MD (Track
Leader)
Christopher S. Lathan, MD, MS, MPH
Abenaa M. Brewster, MD, MHS
James Roger Marshall, PhD
Melanie R. Palomares, MD
Central Nervous System Tumors
Joon H. Uhm, MD (Track Leader)
Jay Steven Loeffler, MD
Timothy Francis Cloughesy, MD
Clinical Trials
David M. Dilts, PhD, MBA (Track
Leader)
Donna S. Neuberg, ScD
Tatiana Michelle Prowell, MD
Philip Jordan Gold, MD
Mary W. Redman, PhD
Tim Eisen, PhD
Developmental Therapeutics
Keith T. Flaherty, MD (Track Leader)
Howard A. Burris III, MD
Razelle Kurzrock, MD
Timothy W. Synold, PharmD
Ethics
Antonella Surbone, MD, PhD (Track
Leader)
Amy Pickar Abernethy, MD
Eileen P. Smith, MD
Rudolph M. Navari, MD, PhD
Gastrointestinal (Colorectal) Cancer

Jeffrey A. Meyerhardt, MD, MPH
(Track Leader)
Johanna C. Bendell, MD
Alan Paul Venook, MD
David P. Ryan, MD
Gastrointestinal (Noncolorectal)
Cancer
Jordan Berlin, MD (Track Leader)
Hedy Lee Kindler, MD
Matthew Kulke, MD
Pamela L. Kunz, MD
Manish A. Shah, MD
Cliff P. Connery, MD
Genitourinary Cancer
Kim N. Chi, MD (Track Leader)
Toni K. Choueiri, MD
Johann Sebastian De Bono, MD, PhD,
MSc
Matt D. Galsky, MD
Primo Lara Jr., MD
Andrew J. Stephenson, MD
Geriatric Oncology
Supriya Gupta Mohile, MD, MS (Track
Leader)
Arti Hurria, MD
Arati Rao, MD
Homayoon Sanati, MD
Gynecologic Cancer
Don S. Dizon, MD (Track Leader)
Gini F. Fleming, MD
David E. Cohn, MD
Head and Neck Cancer
David Sidransky, MD (Track Leader)
Ezra E.W. Cohen, MD
Vassiliki Papadimitrakopoulou, MD
Health Services Research
Jennifer J. Griggs, MD, MPH (Track
Leader)
Dawn L. Hershman, MD
Jennifer Malin, MD
Rinaa S. Punglia, MD
Steven J. Katz, MD
Leukemia, Myelodysplasia, and
Transplantation
Martin S. Tallman, MD (Track Leader)
Daniel J. DeAngelo, MD, PhD
Koen Van Besien, MD
Lung Cancer
Renato Martins, MD, MPH (Track
Leader)
David E. Gerber, MD
Pasi A. Janne, MD, PhD
Jyoti D. Patel, MD
Charles Andrew Butts, MD
Natasha B. Leighl, MD
Lymphoma and Plasma Cell

Disorders
Amrita Y. Krishnan, MD (Track Leader)
Ranjana Advani, MD
Nancy Bartlett, MD
Steven M. Horwitz, MD
Jeremy S. Abramson, MD, MSc
Melanoma/Skin Cancers
Leslie Anne Fecher, MD (Track Leader)
F. Stephen Hodi, MD
Vernon K. Sondak, MD
Denise L. Johnson Miller, MD
Hensin Tsao, MD
Patient and Survivor Care
Julia Howe Rowland, PhD (Track
Leader)
Sydney Morss Dy, MD
Teresa Gilewski, MD
Gretchen Genevieve Kimmick, MD
Pediatric Oncology
Stephen L. Lessnick, MD, PhD (Track
Leader)
Mark W. Kieran MD, PhD
Stephen Skapek, MD
Judith K. Sato, MD
Practice Management and
Information Technology
Craig A. Bunnell, MD (Track Leader)
Robert Stephen Miller, MD
William Charles Penley, MD
Alan M. Keller, MD
David R. Artz, MD
Professional Development
Michael Anthony Carducci, MD (Track
Leader)
Emily K. Bergsland, MD
Christine Marie Lovly, MD, PhD
Vikki A. Canfield, MD
Katherine Elizabeth Reeder-Hayes, MD
Robert A. Wolff, MD
Sarcoma
George D. Demetri, MD (Track Leader)
Richard F. Riedel, MD
Warren Allen Chow, MD
Jayesh Desai, MD
R. Lor Randall, MD
Tumor Biology
Levi Garraway, MD, PhD (Track
Leader)
Ross L. Levine, MD
Lee M. Ellis, MD
Josep Tabernero, MD
2012 Annual Meeting Faculty List

Matti S. Aapro, MD
Clinique De Genolier
Denise R. Aberle, MD
University of California, Los
Angeles
Amy Pickar Abernethy, MD
Duke University Medical Center
Thierry Andre,
Pitie-Salpetriere Hospital
Gerald L. Andriole, MD
Washington University School of
Medicine
Andrew J. Armstrong, MD, ScM
Duke Cancer Institute
Deborah Kay Armstrong, MD
Sidney Kimmel Comprehensive
Cancer Center at Johns Hopkins
University
Carola A. S. Arndt, MD
Mayo Clinic
Carlos L. Arteaga, MD
Vanderbilt University
Michael B. Atkins, MD
Beth Israel Deaconess Medical
Center
Michel Attal, MD
CHU Purpan
Peter Bach, MD
Memorial Sloan-Kettering Cancer
Center
Deliya Banda, PhD, MPH
Washington Hospital Center
Alberto Bardelli, PhD
Istituto Ricerca Cura Cancro and
FIRC
Ethan M. Basch, MD, MSc
Center
The Sarah Cannon Research
Institute
Christine D. Berg, MD
National Cancer Institute
Carl Bergh, MD
Karolinska Institutet
Jordan Berlin, MD
Vanderbilt-Ingram Cancer Center
Donald A. Berry, PhD
University of Texas M. D.
Anderson Cancer Center
Keith C. Bible, MD, PhD
Mayo Clinic
Douglas W. Blayney, MD
Stanford University
Diane Blum, MSW
Lymphoma Research Foundation
Daniel Brat, MD, PhD
Emory University

David M. Brizel, MD
Itzhak Brook, MD, MSc
Georgetown University
Adam Brufsky, MD, PhD
University of Pittsburgh School
of Medicine
Deborah Watkins Bruner, PhD, RN
Abramson Cancer Center,
University of Pennsylvania
Thomas A. Buchholz, MD
Howard A. Burris, MD
The Sarah Cannon Research
Institute
Harold J. Burstein, MD, PhD
Dana-Farber Cancer Institute
Marc E. Buyse, ScD
International Drug Development
Institute
John C. Byrd, MD
The Ohio State University
Comprehensive Cancer Center
D. Ross Camidge, MD, PhD
University of Colorado Cancer
Center
Fatima Cardoso, MD
Breast Unit, Champalimaud
Cancer Center
Bruce Allan Chabner, MD
Massachusetts General Hospital
Susan Marina Chang, MD
University of California, San
Francisco
E. Antonio Chiocca, MD, PhD
Medical Center
Toni K. Choueiri, MD
Dana-Farber Cancer Institute/
Harvard Medical School
Christine H. Chung, MD
Johns Hopkins University
Tudor-Eliade Ciuleanu, MD
Institute Ion Chiricuta
Rebecca Anne Clark-Snow, RN
University of Kansas Cancer
Center
Timothy Francis Cloughesy, MD
Angeles
Kenneth J. Cohen, MD, MBA
Johns Hopkins School of
Medicine
A. Dimitrios Colevas, MD
Stanford University School of
Medicine
Howard Colman, MD, PhD

University of Utah
Nicoletta Colombo, MD
University of Milan-Bicocca
Robert Leo Comis, MD
Coalition of Cancer Cooperative
Groups
Carolyn C. Compton, MD, PhD
National Institutes of Health
Fergus J. Couch, PhD
Mayo Clinic
Jeffrey Crawford, MD
Steven A. Curley, MD
John Patrick Curtin, MD
New York University School of
Medicine
Nancy E. Davidson, MD
University of Pittsburgh Cancer
Institute
Laura A. Dawson, MD
Princess Margaret Hospital
Johann Sebastian De Bono, MB,
ChB, MSc, PhD
Royal Marsden Hospital
Dave Debronkart
Boston, Massachusetts
John F. Deeken, MD
Lombardi Comprehensive
Cancer Center, Georgetown
University
George D. Demetri, MD
Dana-Farber Cancer Institute/
Harvard Medical School
Lynette Denny, MD, PhD
University of Cape Town
Frank C. Detterbeck, MD
Yale School of Medicine
Craig Earle, MD
Institute for Clinical Evaluative
Sciences
Anthony D. Elias, MD
University of Colorado Cancer
Center
Lee M. Ellis, MD
William F. Elmquist, PhD,
PharmD
University of Minnesota
Laura Esserman, MD, MBA
Francisco Helen Diller Family
Ruth D. Etzioni, PhD
University of Washington
Ann T. Farrell, MD
U. S. Food and Drug
Administration
Henry Feldman, MD
Division of Clinical Informatics
Chris Feudtner, MD, PhD, MPH
Childrens Hospital of
Philadelphia
Michael Fisch, MD, MPH
Keith T. Flaherty, MD
Chris Flowers, MD
H. Lee Moffitt Cancer Center &
Research Institute
Jean G. Ford, MD
Medicine
Nicholas K. Foreman, MD
The Childrens Hospital
Josef J. Fox, MD
Center
Keiichi Fujiwara, MD, PhD
Saitama Medical University
International Medical Center
Gwendolyn A. Fyfe, MD
San Francisco, California
Matt D. Galsky, MD
The Tisch Cancer Institute,
Mount Sinai School of Medicine
Patricia A. Ganz, MD
Angeles Schools of Medicine and
Public Health
Levi A. Garraway, MD, PhD
Karen A. Gelmon, MD
British Columbia Cancer
Agency
Mark R. Gilbert, MD
Teresa Gilewski, MD
Center
Maura L. Gillison, MD, PhD
Nicolas Girard, MD, MSc
Hopital Louis Pradel
Michael Goldstein, MD
Center and Harvard Medical
School
Paul J. Goodfellow, PhD
Medicine
Karyn A. Goodman, MD
Center
Ajay K. Gopal, MD
Jason R. Gotlib, MD
Stanford University School of
Medicine
Bernardo Haddock Lobo Goulart,
MD
Fred Hutchinson Cancer
Research Center
John G. Gribben, DSc, MD
St. Bartholomews Hospital
Stephen S. Grubbs, MD
Helen F. Graham Cancer Center
at Christiana Care
Stephen B. Gruber, MD, PhD,
MPH
University of Michigan
Steven M. Grunberg, MD
Fletcher Allen Health Care
Eva Grunfeld, MD, DPhil
University of Toronto
Philip H. Gutin, MD
Center
Paul Han, MD, MA, MPH
Maine Medical Center Reseach
Institute
Robert Hauser, PhD, PharmD
Oncology
Joshua Hauser, MD
Northwestern University
Daniel F. Hayes, MD
University of Michigan Medical
Center
Elisabeth I. Heath, MD
Karmanos Cancer Institute
Michael C. Heinrich, MD
Oregon Health & Science
University Knight Cancer
Institute
Paul R. Helft, MD
Indiana University Simon Cancer
Center
Daniel Yick Chin Heng, MD,
MPH
Tom Baker Cancer Centre,
University of Calgary
Martee Leigh Hensley, MD, MSc
Center
Dawn L. Hershman, MD, MS
Columbia University Medical
Center
William J. Hicks, MD
Peter Hillmen, PhD
Leeds General Infirmary
Bruce E. Hillner, MD
Virginia Commonwealth
University
Pamela S. Hinds, PhD, RN

Childrens National Medical
Center
F. Stephen Hodi, MD
Theodore S. Hong, MD
Melissa M. Hudson, MD
St. Jude Childrens Research
Hospital
Stephen Hunger, MD
University of Colorado Denver
School of Medicine
Arti Hurria, MD
City of Hope
Eun-Sil Shelley Hwang, MD,
MPH
Duke Unversity Medical Center
Nadine Jackson McCleary, MD
Joseph O. Jacobson, MD
North Shore Medical Center
Anuja Jhingran, MD
Maxine S. Jochelson, MD
Center
Stephen R. D. Johnston, MA, MD,
PhD
Ian Robert Judson, MD
Royal Marsden Hospital and
NHS Foundation Trust
William G. Kaelin, MD
Yoon-Koo Kang, MD, PhD
Asan Medical Center
Hagop Kantarjian, MD
Neil E. Kay, MD
Mayo Clinic
Dorothy Mary Kate Keefe, MD,
MBBS
Royal Adelaide Hospital
Karla Kerlikowske, MD
Francisco
Muin Khoury, MD, PhD
Office of Public Health
Genomics, Centers for Disease
Control and Prevention
Mark W. Kieran, MD, PhD
Gretchen Genevieve Kimmick,
MD, MS

The University of Chicago
Medical Center
Roger David Klein, MD, JD
Bloodcenter of Wisconsin
Heidi D. Klepin, MD, MS
Wake Forest University School of
Medicine
Sara Knapke, MS
Cincinnati Childrens Hospital
Medical Center
Andrew Y. Koh, MD
University of Texas Southwestern
Medical Center
Barnett S. Kramer, MD, MPH
Amrita Y. Krishnan, MD
City of Hope National Medical
Center
Pamela L. Kunz, MD
Stanford University Medical
Center
Sancy Ann Leachman, MD
University of Utah Hospitals and
Clinics
J. Jack Lee, PhD, DDS
Fred T. Lee, MD
University of Wisconsin
Natasha B. Leighl, MD
John Leonard, MD
Weill Cornell Medical College
Stephen L. Lessnick, MD, PhD
Huntsman Cancer Institute at the
University of Utah
Douglas A. Levine, MD
Center
Lisa F. Licitra, MD
Fondazione IRCCS Istituto
Nazionale dei Tumori
Patricia LoRusso, DO
Tobey MacDonald, MD
Emory University
Jennifer W. Mack, MD
Michael L. Maitland, MD, PhD
Robert G. Maki, MD, PhD
Jeremy Manier
Elizabeth Mansfield, PhD
U. S. Food and Drug
Administration
John M. Maris, MD
Philadelphia
James D. Marks, MD, PhD
Francisco
Mary S. McCabe, RN, MA
Center
Worta J. McCaskill-Stevens, MD
Division of Cancer Prevention,
D. Scott McMeekin, MD
University of Oklahoma Health
Sciences Center
Jerry Menikoff, MD, JD
U.S. Department of Health and
Human Services
Robert Stephen Miller, MD
Tetsuya Mitsudomi, MD, PhD
Aichi Cancer Center Hospital
Supriya Gupta Mohile, MD, MS
University of Rochester Medical
Center
Tony Mok, MD
Prince of Wales Hospital
Monica Morrow, MD
Center
Robert John Motzer, MD
Center
Beverly Moy, MD
Therese M. Mulvey, MD
Southcoast Hospitals Group
Hyman Bernard Muss, MD
University of North Carolina
Lineberger Comprehensive
Cancer Center
Fariba Navid, MD
Hospital
Michael N. Neuss, MD
Vanderbilt-Ingram Cancer
Center
Kim Nichols, MD
Philadelphia
Andrew David Norden, MD, MPH
Kenneth Offit, MD, MPH
Center
Bert H. ONeil, MD
University of North Carolina at
Chapel Hill
Henry Orlando Otero, MD

Virginia Mason Medical Center
Cynthia Owusu, MD, MS
Case Western Reserve
University
William Pao, MD, PhD
Vanderbilt-Ingram Cancer
Center
Alberto S. Pappo, MD
Hospital
Donald W. Parsons, MD, PhD
Texas Childrens Hospital
Francesco Passamonti, MD
Fondazione IRCCS Policlinico
San Matteo
Jeffrey M. Peppercorn, MD, MPH
Duke University Medical
Center
Douglas E. Peterson, DMD, PhD
University of Connecticut Health
Center
Martine J. Piccart-Gebhart, MD,
PhD
Jules Bordet Institute
Todd Alan Pickard, PA-C
Whitney Pope, MD, PhD
Angeles
Kathleen I. Pritchard, MD
Sunnybrook Health Sciences
Centre
Stephanie Puget, MD, PhD
Necker Hospital, Universite
Paris-Descartes
Martin N. Raber, MD
Derek Raghavan, MD, PhD
Carolinas Medical Center
S. Vincent Rajkumar, MD
Mayo Clinic
David Ransohoff, MD
University of North Carolina at
Chapel Hill
Arati Rao, MD
Center
Gregory H. Reaman, MD
Childrens National Medical
Center
Mary Weber Redman, PhD
Fred Hutchinson Cancer
Research Center
Robert Evan Reiter, MD
Angeles
Antoni Ribas, MD
Angeles
Duke Cancer Institute, Duke
University Medical Center
Gregory J. Riely, MD, PhD
Center
Elizabeth Schmidt Rodriguez, RN
Center
Lee S. Rosen, MD
Angeles Division of HematologyOncology
Michael Rowe, PhD
Gordon J. S. Rustin, MD, MBBS
Mount Vernon Cancer Centre
David P. Ryan, MD
Cancer Center
Charles J. Ryan, MD
Francisco
Marianne Ryberg, MD
Herlev University Hospital
Gilles A. Salles, MD, PhD
Hospices Civils de Lyon and
Universite de Lyon
Leonard Saltz, MD
Center
Daniel J. Sargent, PhD
Mayo Clinic
Oliver Sartor, MD
Tulane Cancer Center
Hans Sauer, PhD, JD
Biotechnology Industry
Organization
Doug Sawyer, MD, PhD
Lidia Schapira, MD
Joshua David Schiffman, MD
University of Utah
Lowell E. Schnipper, MD
Center
Gilberto Schwartsmann, MD, PhD
Federal University of Rio Grande
do Sul
Katia Scotlandi, PhD
Istituto Ortopedico Rizzoli
Victoria Louise Seewaldt, MD
Center
Lecia V. Sequist, MD, MPH
Manish A. Shah, MD
Manisha H. Shah, MD
Neil P. Shah, MD, PhD

Francisco
Geoffrey Shapiro, MD, PhD
Alice Tsang Shaw, MD, PhD
Cancer Center
Lillian L. Siu, MD
Stephen Skapek, MD
Martha L. Slattery, PhD
University of Utah
George W. Sledge, MD
Indiana University Simon Cancer
Center
Stefan Sleijfer, MD, PhD
Erasmus University Medical
Center, Daniel den Hoed Cancer
Center
Sonali M. Smith, MD
Thomas J. Smith, MD
Stephen Barnett Solomon, MD
Center
Stephen T. Sonis, DMSc, DDS
Biomodels
Jeffrey Alan Sosman, MD
Vanderbilt Medical Center
Paul Spellman, PhD
University
Margaret R. Spitz, MD
Zsofia Kinga Stadler, MD
Center
Walter Michael Stadler, MD
Richard Sullivan, PhD, MBBS
Kings Health Partners Integrated
Cancer Centre
Lillian Sung, MD, PhD
The Hospital for Sick Children
Antonella Surbone, MD, PhD
New York University
Sandra M. Swain, MD
Mario Sznol, MD
Yale Cancer Center
Josep Tabernero, MD
Vall dHebron University
Hospital
Kenneth Tanabe, MD
Cancer Center
William D. Tap, MD
Angeles
Ian Murchie Thompson, MD
University of Texas Health
Science Center at San Antonio
Michael A. Thompson, MD, PhD
ProHealth Regional Cancer
Center
Anthony W. Tolcher, MD
South Texas Accelerated
Research Therapeutics (START)
Sean R. Tunis, MD, MSc
Center for Medical Technology
Policy
Alan Paul Venook, MD
Francisco
Srdan Verstovsek, MD, PhD
Jaap Verweij, MD, PhD
Erasmus University Medical
Center
Louise Villejo, MPH
Beth A. Virnig, PhD
University of Minnesota, School
of Public Health
Andrew J. Wagner, MD, PhD
James Ross Waisman, MD
Breastlink Medical Group Inc
Robert S. Warren, MD
Francisco
Jeffrey S. Weber, MD, PhD
H. Lee Moffitt Cancer Center &
Research Institute
Simon Wein, MD
Davidoff Cancer Center, Rabin
Medical Center
Martin R. Weiser, MD
Center
Howard Jack West, MD
Swedish Cancer Institute
William H. Westra, MD
Medicine
Lauren Allison Wiebe, MD
Medical College of Wisconsin
Timothy J. Wilt, MD, MPH
United States Department of
Veterans Affairs
Eric P. Winer, MD
Ignacio I. Wistuba, MD
Douglas E. Wood, MD
Paul Workman, BSc, PhD, DSc
The Institute of Cancer Research
Michael Xing, MD, PhD
Sam S. Yoon, MD
Anas Younes, MD
Theoklis Zaoutis, MD
Philadelphia
Andrew X. Zhu, MD, PhD
2012 Educational Book Expert Panel

The Expert Panel is a group of well-recognized physicians and researchers in oncology
and related fields who have served as peer reviewers on the Educational Book articles.
David Adelstein, MD
Cleveland Clinic
Alex Adjei, MD
Doug Adkins, MD
Medicine
Stefan Anker, MD
Charite, Universitatsmedizin
Berlin
Robert Arceci, MD, PhD
Sidney Kimmel Cancer Center
Voichita Bar-Ad, MD
Thomas Jefferson University
Hospital
Helen Barraclough, MS
Eli Lilly
Leif Bergsagel, MD
Mayo Clinic
Jan Beumer, PharmD, PhD
University of Pittsburgh
George Blumenschein, MD
Louise Bordeleau, MD, MSc
McMaster University
Alan Bryce, MD
Mayo Clinic
Kevin Camphausen, MD
James Cassidy, MD
Roche
Nelson Chao, MD
Duke University
Carien Creutzberg, MD
Leiden University Medical
Center
Brian Czito, MD
Duke University
Adam Dicker, MD
Brenda Diergaarde, PhD
Benjamin Djulbegovic, MD
University of South Florida
Laura Dominici, MD
Brigham and Womens Hospital
Fritz Eilber, MD
University of California,
Los Angeles
Rebecca Elstrom, MD
Edward Garon, MD
University of California,
Los Angeles
Julia Glade, MD
Columbia University
Stewart Goldman, MD
Childrens Memorial Hospital
Mary Gospodarowicz, MD
Victor Grann, MD
Columbia University
Parameswaran Hari, MD
Michael Hassett, MD
Michael Isakoff, MD
Conneticut Childrens Medical
Center
Michael Kelly, MD, PhD
Scott Kopetz, MD
Mario Lacouture, MD
Edward Levine, MD
Wake Forest University
Gerry Linette, MD, PhD
Medicine
Steven Lipshultz, MD
University of Miami
Christophe Louvet, MD
lInstitut Mutualiste Montsouris
Bo Lu, MD, PhD
Mario Malogolowkin, MD
Guido Marcucci, MD
Medical Center
Erica Mayer, MD
Minesh Mehta, MD
Vicki Morrison, MD
University of Minnesota
Joanne Mortimer, MD
City of Hope
Craig Moskowitz, MD
Center
Timothy Moynihan, MD
Mayo Clinic
Barbara Murphy, MD
P.B. Ottevanger, MD
Radboud University Nijmegen
Medical Centre
Rebecca Pentz, PhD
Emory University
James Pingpank, MD
Katherine Pisters, MD
Dennis Priebat, MD
Rinaa Punglia, MD
David Quinn, MD
University of Southern California
Nithya Ramnath, MD
University of Michigan
Paul Gerard Guy Richardson, MD
Jonathan Rosenberg, MD
David Rosenthal, MD
Harvard University
Anna Roshal, MD
Medicine
John Ruckdeschel, MD
Hope Rugo, MD
Francisco
Rachel Sanborn, MD
Providence Portland Medical
Center
Alan Sandler, MD
University
Rafael Santana-Davilla, MD
Takami Sato, MD, PhD
Jefferson Medical College
Lee Schwartzberg, MD
The West Clinic
Sunil Sharma, MD
University of Utah
Wenyin Shi, MD
Hospital
Timothy Showalter, MD
Hospital
Mary Lou Smith, JD, MBA

Research Advocacy Network,
Chicago
Andrew Stephenson, MD
Cleveland Clinic
Timothy Synold, PharmD
City of Hope
Ben Tan, MD
Medicine
Meaghan Tenney, MD
University of Chicago
Brian Van Tine, MD, PhD
Medicine
Gauri Varadhachary, MD
Ravi Vij, MD
Medicine
Nina Wagner-Johnston, MD
Medicine
Heather Wakelee, MD
Stanford University
Ellen Warner, MD
Odette Cancer Center
Theresa Werner, MD
University of Utah
Mier Wetzler, MD
Roswell Park Cancer Institute
William Wierda, MD
Dennis Wigle, MD, PhD
Mayo Clinic
Tanya Wildes, MD
Medicine
Els Witteveen, MD
University Medical Center
Utrecht
Sandra Wong, MD
University of Michigan Health
Systems
Seiko Yamada, MD
University of Chicago
Darrell Yamashiro, MD, PhD
Columbia University
Jonathan Zager, MD
H. Lee Moffitt Cancer Center
Robin Zon, MD
Michiana Hematology-Oncology
2012 Annual Meeting Supporters*

CONQUER CANCER FOUNDATION MISSION ENDOWMENT FOUNDING DONORS
Genentech BioOncologyTM
GlaxoSmithKline Oncology
Novartis Oncology
Sanofi Oncology
CONQUER CANCER FOUNDATION MISSION ENDOWMENT SUSTAINING DONORS

Abbott Oncology
Amgen
Celgene Corporation
Lilly USA, LLC

Millennium: The Takeda Oncology Company
Sanofi Oncology
EDUCATIONAL SUPPORT
Merck Oncology
Melanoma/Skin Cancer Track
Amgen
Educational Book
Astellas
AVEO Pharmaceuticals, Inc.
Medivation
Genitourinary Cancer Track
Bristol-Myers Squibb
Gastrointestinal (Colorectal) Cancer Track
Gastrointestinal (Noncolorectal) Cancer Track
Head and Neck Cancer Track
Celgene Corporation
Best of ASCO Meetings
Breast Cancer Track
Leukemia, Myelodysplasia, and Transplantation Track
Lung Cancer Track
Lymphoma and Plasma Cell Disorders Track
Daiichi Sankyo, Inc.
Lung Cancer Track
Eisai Inc.
Breast Cancer Track
Patient and Survivor Care Track
Genentech BioOncologyTM
Clinical Trials Track
Gynecologic Cancer Track
Melanoma/Skin Cancer Track
Professional Development Track
Tumor Biology Track
Kidney Cancer Association
Lilly USA, LLC
Head and Neck Cancer Track
Lung Cancer Track
* This list reflects commitments as of April 16, 2012

Professional Development Track
Myriad Genetics Laboratories, Inc.
Breast Cancer Track
Cancer Genetics Track
Novartis Oncology
Breast Cancer Track
Onyx Pharmaceuticals
Pfizer Oncology
Breast Cancer Track
Lung Cancer Track
Patient and Survivor Care Track
Sanofi Oncology
Sanofi Oncology and Regeneron Pharmaceuticals
Seattle Genetics
GENERAL SUPPORT
Abbott Oncology
Young Investigator Award
American Cancer Society
ASCO-American Cancer Society Award and Lecture
Amgen
Career Development Award
International Development and Education Awards
Long-term International Fellowship (LIFe)
Merit Awards
ASCO and Conquer Cancer Foundation Boards of

Directors
Jane C. Wright, MD, Young Investigator Award
Doris Duke Charitable Foundation

Medical Student Rotation for Underrepresented
Populations
ASCO Cancer Research Committee

Eisai Inc.
Charging Stations
Interactive Locator Maps
Internet Stations
Patient Advocate Scholarship Program
Laptop Stations (Wi-Fi Zones)
Young Investigator Award (2)
ASCO Clinical Practice Committee

Astellas
Avon Foundation for Women
International Development and Education Awards in
Breast Cancer
Bayer Healthcare Pharmaceuticals Inc.
Boehringer Ingelheim Pharmaceuticals, Inc.
iMeetings Mobile Application
Virtual Meetings
Bradley Stuart Beller Endowed Fund
Bradley Stuart Beller Special Merit Award
The Breast Cancer Research Foundation
Advanced Clinical Research Award in Breast Cancer
Career Development Award in Breast Cancer
Comparative Effectiveness Research Professorship in
Breast Cancer
Long-term International Award (LIFe) in Breast or
Related Cancers
Young Investigator Award in Breast Cancer
Young Investigator Award in Breast Cancer in honor of
Susan Hirschhorn and in memory of her mother
Bristol-Myers Squibb
Annual Meeting Program
Merit Awards
Celgene Corporation
Merit Awards
Celltrion Healthcare Co., Ltd.
Annual Meeting Program Mobile Application (iPlanner)
Coalition of Cancer Cooperative Groups
Clinical Trials Participation Awards
Daiichi Sankyo, Inc.
Dendreon Corporation
Annual Meeting ASCO TV
Internet Stations
Don Shula Foundation, Inc.
Young Investigator Award in Breast Cancer
Estate of Sandra Syms

Young Investigator Award in Kidney Cancer Research
Genentech BioOncology
Advanced Clinical Research Award in Colorectal Cancer
Advanced Clinical Research Award in Glioma
Advanced Clinical Research Award in Hematologic
Malignancies
Career Development Award (5)
Program Announcement
Registration and Housing Announcement
Translational Research Professorship
GlaxoSmithKline Oncology
Gianni Bonadonna Breast Cancer Award and Fellowship
Translational Research Professorship
Incyte Corporation
Annual Meeting ASCO TV
Charging Stations
Interactive Locator Maps
Internet Stations
Jackson Simpson
Merit Awards in Lung Cancer and Lymphoma
James B. Nachman Pediatric Oncology Fund
James B. Nachman ASCO Junior Faculty Award in
Pediatric Oncology
Journal of Clinical Oncology
Kidney Cancer Association
Young Investigator Award in Renal Cell Carcinoma
Lilly USA, LLC
Merit Awards
Merck Oncology
Merit Awards
Oncology Trainee Travel Awards

Palliative Care
Novartis Oncology
Open Society Foundations
Palliative Care
Palo Alto Medical Foundation
Leaders in PracticeSilver Supporter
Pepper Hamilton LLC
Pro Bono Legal Services: Diversity in Oncology Initiative
Pfizer Oncology
Virtual Meetings
Raj Mantena, RPh
Leaders in PracticePrimary Supporter
Roche
Career Development Award (2)
Sanofi Oncology
Drug Development Research Professorship
Sarcoma Foundation of America
Career Development Award in Sarcoma Research
Young Investigator Award in Sarcoma Research
Seattle Genetics
iMeetings Mobile Application
Internet Stations
Strike 3 Foundation
Young Investigator Award in Pediatric Oncology
Susan G. Komen for the Cure
The Susan G. Komen for the Cure/Conquer Cancer
Foundation Research Initiative
Conquer Cancer Foundation of ASCO Improving
Cancer Care Grant, funded by Susan G. Komen for the
Cure
The Susan G. Komen for the Cure/ASCO Quality of Care
Initiative
The ASCO Breast Cancer Registry Pilot Program,
funded by Susan G. Komen for the Cure
The ASCO Diversity in Oncology Initiative, funded by
Susan G. Komen for the Cure
The ASCO Study of Collaborative Practice
Arrangements, funded by Susan G. Komen for the
Cure
The ASCO Study of Geographic Access to Oncology
Care, funded by Susan G. Komen for the Cure
Assessing and Monitoring Demand for Cancer Care
Services, with funding from Susan G. Komen for the
Cure
William D. Piety Living Trust
Young Investigator Award in Cholangiocarcinoma
Research
WWWW Foundation, Inc. (QuadW) and The Sarcoma
Fund of the QuadW Foundation
Young Investigator Award in Sarcoma in memory of Willie
Tichenor
Letter from the Editor
he theme of the 2012 American Society of Clinical

Oncology Annual Meeting is Collaborating to Conquer
Cancer, and the sessions being presented at this years
meeting certainly look toward this theme with the goal of
advancing the education of oncologists so that they can
improve the care provided to their patients.
The ASCO Educational Book is meant to serve as an
enduring record of the information presented at these sessions. This year, 92 educational sessions are represented by
the articles published in this book, written by some of the
most recognized clinicians and investigators in their field. In
order to ensure the utmost accuracy and clarity, each article
is reviewed by our editorial team, as well as by peer
reviewers from our extensive expert panel. In addition,
Natasha Leighl, MD, an expert in lung cancer and cost-ofcare issues and the Associate Editor of the Educational
Book, has provided valuable insights to the review process.
We are grateful to the authors who, despite their demanding schedules, took the time to write and, in some cases,
revise their articles according to the peer reviewers comments. We appreciate the efforts of our peer reviewers for
their careful, thorough reviews that undoubtedly contributed to the high quality of this book.
Please visit asco.org/edbook for access to all of the 2012
Educational Book articles. The following are examples for
citing both print and electronic articles:
Girard N. Thymoma: From Chemotherapy to Targeted

Therapy. In: Govindan R, ed. 2012 ASCO Educational Book.
Alexandria, VA: American Society of Clinical Oncology;
2012;475-479.
Surbone A, Baider L. Are Oncologists Accountable Only to
Patients or Also to Their Families? An International Perspective. In: Govindan R, ed. 2012 ASCO Educational Book.
Alexandria, VA: American Society of Clinical Oncology;
2012;e15-e19.
The ASCO Educational Book would not exist today but for
the remarkable staff at ASCO. I want to specifically thank
Christine Smith and Lauren Burke, who have worked very
hard year-round to make this publication possible. Dr.
Leighl and I welcome your feedback and suggestions on how
we can improve the content, so please feel free to email me
with your comments at EdBook@asco.org.
Ramaswamy Govindan, MD
Editor
CONTROVERSIES IN THE ADJUVANT TREATMENT

OF BREAST CANCER
CHAIR
Stephen R. Johnston, MA, PhD
London, United Kingdom
SPEAKERS
Anthony D. Elias, MD
University of Colorado Cancer Center
Aurora, CO
Hyman Muss, MD
University of North Carolina Lineberger
Chapel Hill, NC
Adjuvant Therapy for Older Women with

Early-Stage Breast Cancer: Treatment
Selection in a Complex Population
By Cynthia Owusu, MD, MS, Arti Hurria, MD, and Hyman Muss, MD
Overview: Breast cancer is a disease of aging. However,

older women with breast cancer are less likely to participate in
clinical trials or to receive recommended treatment. This
undertreatment has contributed to a lag in breast cancer
survival outcomes for older women compared with that for
their younger counterparts. The principles that govern recommendations for adjuvant treatment of breast cancer are the
same for younger and older women. Systemic adjuvant treatment recommendations should be offered on the basis of
tumor characteristics that divide patients into three distinct
subgroups. These include (1) older women with hormone
receptor (HR)-positive and human epidermal growth factor 2
(HER2)-negative breast cancer who should be offered endocrine therapy; (2) older women with HR-negative and HER2negative breast cancer who should be offered adjuvant
REAST CANCER is the most common cancer in American women and the second leading cause of cancerrelated deaths among women. In 2011, approximately
230,480 new cases were diagnosed in the United States,
with an expected 39,520 deaths.1,2 The most important risk
factor for breast cancer is age. The estimated lifetime risk of
a new breast cancer is 1 in 15, 1 in 29, 1 in 27 and 1 in 207
for women 70 years or older, 60 to 69, 40 to 59, and 39 or
younger, respectively.1 The median age at the time of breast
cancer diagnosis is currently 61 years and an estimated 45%
of women are 65 or older at the time of initial diagnosis.2,3
Recent gains in life expectancy, coupled with aging as a risk
factor for breast cancer, makes breast cancer primarily a
disease of older women, with increasing public health importance. In 1980, persons 65 and older represented 11.3%
of the total population, but by 2030 this proportion is
expected to increase to 20%.3 In addition, by 2030, persons
older than 75 will be expected to account for just under 50%
of the total cohort older than 65.4 Given the nonlinear
age-risk relationship and increasing life expectancy, a substantial proportion of older women are expected to be affected by breast cancer.
Age-Related Cancer Health Disparities
Evaluation of the biology of breast cancer by patient age

has shown that hormone receptor-positive, low S-phase, low
tumor grade and HER2-negative tumors are more common
among older than younger women,5 although these differences are relatively modest. Despite the favorable tumor
profile of breast cancer in older women, this has not translated into any major survival advantage for older women
with breast cancer in comparison with their younger counterparts. In a study that drew data from National Vital
Statistics Reports and the Surveillance Epidemiology and
End Results database of the National Cancer Institute,
Smith and colleagues6 found that although the rate of breast
cancer death in the general population and the adjusted risk
of death among women with newly diagnosed disease are
declining among all age groups, the least decline has been
among older women. Relative to 1990, the rate of breast
chemotherapy; and (3) older women with HER2-positive disease who should be offered chemotherapy with trastuzumab.
Exceptions to these guidelines may be made for older women
with small node-negative tumors or frail older women with
limited life expectancy, where close surveillance may be a
reasonable alternative. Addressing the current age-related
disparities in breast cancer survival will require that older
women are offered the same state-of-the-art-treatment as
their younger counterparts, with a careful weighing of the
risks and benefits of each treatment in the context of the
individuals preferences. In addition, older women should be
encouraged to participate in breast cancer clinical trials to
generate additional chemotherapy efficacy, toxicity, and quality of life data.
cancer death in the general population decreased 2.5% per

year for women age 20 to 49, 2.1% per year for those age 50
to 64, 2% per year for those age 65 to 74, but 1.1% per year
for those age 75 years and older. Moreover, among women
with newly diagnosed breast cancer between 1980 and 1997,
the adjusted risk of death decreased by 3.6% per year among
women younger than age 75 compared with 1.3% per year
among those age 75 and older (p 0.01). These differences
were even greater for older black women. The age-related
disparity in survival outcomes was hypothesized to be related to the undertreatment of older women with breast
cancer. In an analysis of 9,766 patients enrolled in the
TEAM (Tamoxifen Exemestane Adjuvant Multinational)
randomized controlled trial conducted with postmenopausal
women with hormone receptor-positive breast cancer, increasing age was associated with a higher disease-specific
mortality.7 Treatments received and tumor characteristics
did not completely explain the age-related differences in
survival outcomes but older patients in this trial were much
less likely to receive chemotherapy. Together, these data
clearly underscore the fact that breast cancer is an important disease of older women who bear a disproportionate
burden of the morbidity and mortality associated with the
disease and who should be offered proven treatments that
improve survival outcomes. Given that treatment differences do not completely explain the age-related disparities
in survival outcomes, additional population and translational studies are needed to provide further insight into the
reasons for these disparities.
From the Case Western Reserve University School of Medicine, Cleveland, OH; City of
Hope Medical Center and Beckman Research Institute, Duarte, CA; and, Lineberger
Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC.
Authors disclosures of potential conflicts of interest are found at the end of this article.
Address reprint requests to Cynthia Owusu, MD, MS, Case Western Reserve University
School of Medicine, Division of Hematology/Oncology and Case Comprehensive Cancer
Center-BHC 5055, 11100 Euclid Avenue, Cleveland, OH 44106; e-mail: cynthia.owusu@
case.edu.
2012 by American Society of Clinical Oncology.
1092-9118/10/1-10
OWUSU, HURRIA, AND MUSS

Adjuvant Systemic Therapy
Adjuvant systemic therapy refers to the administration of

anticancer therapy after primary breast surgery for early
stage breast cancer with the goal of eradicating occult
micrometastatic disease thought to be responsible for distant recurrence. Systemic treatment modalities include
endocrine therapy and chemotherapy with or without trastuzumab. Treatment decision making regarding adjuvant
systemic therapy for older women should involve consideration of such factors as the risk of morbidity and mortality
from breast cancer, life expectancy and treatment tolerance,
and patient preference. A geriatric assessment that includes
evaluation of functional, cognitive, nutritional, and psychosocial status, and review of comorbidities and concomitant
medications is particularly helpful in estimating the health
status and life expectancy of older adults. Each of the
components of a geriatric assessment can identify older
adults at increased risk for morbidity and mortality. This
assessment can also identify areas of vulnerability that may
affect the patients ability to participate in a treatment plan
(for example, ability to take medications on ones own).
Items in a geriatric assessment are able to identify older
adults at risk of chemotherapy toxicity.8,9 In a multisite
study of 500 older adults with cancer, conducted by the
Cancer and Aging Research Group, five geriatric assessment
questions were independent predictors of the risk of chemotherapy toxicity, in addition to age, tumor, treatment, and
laboratory variables. The geriatric assessment questions
that were predictive of chemotherapy toxicity included hearing impairment (rated at fair or worse), difficulty in walking
one block, need for assistance with taking medications, one
or more falls in the past 6 months, and decrease in social
activities because of either physical or emotional health.8
This predictive model for chemotherapy toxicity is being
validated specifically in older adults receiving adjuvant
chemotherapy for breast cancer (clinicaltrials.gov
NCT01472094). In another study of 518 evaluable older
adults with cancer (Chemotherapy Risk Assessment Scale
for High Age Patients), predictors of chemotherapy toxicity
included measures of functional status (ability to complete
KEY POINTS
The undertreatment of older women with breast

cancer has contributed to poorer survival outcomes
for older women than for younger women.
The principles that guide breast cancer treatment
recommendations for younger and older women are
fundamentally the same.
Breast cancer treatment decision making should consider the risk of breast cancer relapse, patient health
status, life expectancy, and patient preference.
The use of systemic therapy in the adjuvant setting
should be based on tumor characteristics and include
endocrine therapy, and/or chemotherapy with or
without trastuzumab.
Omission of postoperative radiation therapy is a
reasonable strategy for older women with small
hormone-receptor positive tumors.
instrumental activities of daily living), cognition (MiniMental Status score), and nutrition (Mini-Nutritional Assessment score).9 This information can be used in the
treatment decision-making process in order to estimate life
expectancy and risk of chemotherapy side effects and to
identify areas of intervention to assist the patient during her
treatment course.
Endocrine Therapy
The majority of older patients present with hormone

receptorpositive breast cancer. Endocrine therapy is the
mainstay of adjuvant therapy for these patients and results
in proportional reductions in the risk of relapse and dying
of breast cancer that exceed any available chemotherapy
regimen. Current consensus guidelines recommend adjuvant systemic endocrine therapy for hormone receptorpositive breast cancer. The National Comprehensive Center
Network (NCCN) guidelines10 recommend the use of adjuvant endocrine therapy for women with hormone receptorpositive breast cancer regardless of age, menopausal status,
or HER2 status, with the possible exception of women
with lymph node-negative cancers 0.5 cm or less, or 0.6 to
1.0 cm in diameter with favorable prognostic features;
benefit from endocrine therapy is likely to be small for
tumors of this size. In contrast, the St. Gallen International
Consensus Panel recommends adjuvant systemic endocrine
therapy for all women with endocrine-responsive disease
with no exceptions and has defined endocrine-responsive
tumors as those with as few as 1% of cells staining positive
for hormone receptor proteins.11 The Society of International Geriatric-Oncology (SIOG) breast cancer task force
recommended that older women with endocrine-responsive
breast cancer be offered systemic endocrine therapy. However, for women with minimal risk disease, the decision to
offer endocrine therapy should be based on a risk-benefit
analysis.12
Tamoxifen is the most firmly established adjuvant endocrine therapy for both premenopausal and postmenopausal
women. Supporting these recommendations are results
from the Early Breast Cancer Trialists Collaborative
Group (EBCTCG) overview analysis, which demonstrated
that over a 15-year period, use of adjuvant tamoxifen
therapy for women with known estrogen receptor-positive
disease, compared with no tamoxifen, decreased the 15-year
risk of recurrence and death by 39% and 31%, respectively,
regardless of age.13 It is clear from these data that
tamoxifen is of benefit for older women. In addition to
decreasing the risk of disease relapse and death, there are
also potential nonbreast cancer benefits of tamoxifen therapy in postmenopausal women. Tamoxifen may prevent
osteoporosis14 and reduce the risk of cardiovascular disease.15 Adjuvant tamoxifen therapy is, however, underutilized in older women. Women 80 years or older are half as
likely to report having had a discussion about tamoxifen
with their doctor compared with women 65 to 79 years, and
women age 85 to 92 years are 25% less likely to receive a
tamoxifen prescription than those 80 to 84 years.16 Additionally, older women are more likely to self-discontinue and
to be nonadherent to tamoxifen before the recommended
treatment period of 5 years,17,18 undercutting the treatment
benefit from tamoxifen. Oncologists should always ask their
patients if they are taking their prescribed medications and
TREATMENT FOR OLDER WOMEN WITH BREAST CANCER
should reinforce the importance of adherence to maximize

treatment benefit.
The adjuvant use of aromatase inhibitors (anastrozole,
letrozole, exemestane) for postmenopausal women with
early breast cancer has been evaluated in several studies.
These studies have involved the use of aromatase inhibitors
either as initial adjuvant therapy,19 as sequential therapy
after 2 to 3 years of tamoxifen,20 or as extended therapy
after 4.5 to 6 years of tamoxifen.21 The findings of these
studies are consistent in demonstrating that the use of a
third-generation aromatase inhibitor for postmenopausal
women with hormone receptorpositive breast cancer, regardless of patient age, is superior in decreasing the risk of
disease recurrence, including ipsilateral and contralateral
breast cancer, and distant metastatic disease compared with
tamoxifen. Additionally, sequential use of aromatase inhibitors20 or extended therapy21 has been shown to provide an
overall survival advantage compared with tamoxifen use for
5 years. No survival advantage has been demonstrated with
the upfront use of aromatase inhibitors for 5 years compared
with tamoxifen.
There are differences in the toxicity profiles of aromatase
inhibitors and tamoxifen. The incidence of venous thromboembolic disease, cerebrovascular events, endometrial cancer, vaginal bleeding, and hot flashes are less likely to be
associated with aromatase inhibitors than tamoxifen,
whereas the incidence of musculoskeletal pain, osteoporosis,
and bone fractures have been found to be higher with
aromatase inhibitors.19 Emerging data also suggest that
aromatase inhibitors may be associated with a small but
higher risk of cardiovascular events compared with tamoxifen,22,23 but not compared with placebo.24 In a metaanalysis of seven trials in which aromatase inhibitors were
compared with tamoxifen, longer duration of aromatase
inhibitor use or aromatase inhibitor use alone for 5 years
was associated with a higher likelihood of cardiovascular
events compared with tamoxifen alone (odds ratio [OR]
1.26, 95% CI 1.10 1.43).25 Because studies of aromatase
inhibitors have not included extensive follow-up, the full
effect of aromatase inhibitors on cardiovascular disease and
coronary heart risk remains to be determined.
Although there is little evidence of age-related differences
in the benefits of aromatase inhibitors for postmenopausal
women, results of studies designed to examine age-related
differences in the pattern of toxicity have been mixed. In
general, the incidence of grade 35 nonfracture-related adverse events is higher among women 75 and older than
among women less than 75 years.26 However, a comparison
of the quality of life and the side effect profile for women who
participated in MA-17, a study in which 5 years of letrozole
was compared with placebo, showed no age-related differences in side effects.24 The long-term consequences and
implications of these side effects and any-age-related differences remain to be well characterized.
Based on the results from recent studies that favor aromatase inhibitors over tamoxifen, current guidelines recommend that aromatase inhibitors should be offered to all
postmenopausal women with hormone receptor-positive
early stage breast cancer, either alone, as sequential therapy after 23 years of tamoxifen. Given the lack of overall
survival advantage associated with aromatase inhibitor use
for 5 years, for women with pre-existing heart disease or
bone loss, use of tamoxifen for 5 years or a switching
strategy is a reasonable approach. For women with low

grade, node-negative tumors 1 cm or smaller, endocrine
therapy may be optional and observation acceptable, although the risks and benefits should be discussed with the
patient.
Adjuvant Chemotherapy
Cytotoxic chemotherapy can be considered for older

women with either node-positive or high-risk node-negative
disease, particularly, triple-negative breast cancer. An
abundance of literature has demonstrated the benefit of
adjuvant chemotherapy for younger women, with benefit
decreasing as age increases. The EBCTCG overview analysis,13 which has 15 years of follow-up data, demonstrated
that adjuvant chemotherapy reduced the annual risk of
recurrence by 37% and 19%, for women younger than 50 and
50 to 69, respectively. The annual risk of death was reduced
by 30% and 12%, for women younger than 50 and 50 to 69,
respectively. The benefit of adjuvant chemotherapy for
women with early stage breast cancer over age 70 was
difficult to assess in the EBCTCG overview analysis
because of the paucity of randomized trials that incorporated this age group. Of 29,000 women included in 60
adjuvant polychemotherapy trials, 4% were 70 and older.
This paucity of data has prevented definitive estimates
regarding the magnitude of benefit of chemotherapy for
women age 70 and older. In addition, with advancing age,
organ function and performance status decline, and comorbidities increase, making the risks associated with chemotherapy even greater. Moreover, the risk reductions for
chemotherapy are lower for postmenopausal women than for
premenopausal women, although the reasons are unclear.
Coupled with the apparent decline in the efficacy of chemotherapy with age is the increased risk of death from competing illnesses (comorbidity), leading to additional decline in
the benefit from chemotherapy. As a result of all these
factors, older women with early stage breast cancer receive
adjuvant chemotherapy considerably less frequently than do
younger women.
However, a growing body of evidence suggests that adjuvant chemotherapy leads to improved survival outcomes
for older women with breast cancer, particularly older
women with hormone receptor-negative and node-positive
breast cancer. A retrospective analysis of four Cancer and
Leukemia Group B (CALGB) randomized clinical trials
showed superior disease-free and overall survival benefits
with the use of more aggressive chemotherapy (compared
with less aggressive chemotherapy), among 6,487 women
with node-positive breast cancer in all age groups, including
70 and older.27 This benefit, however, came at the cost of
increased risk of toxicity in older women, with older women
more likely to discontinue treatment and to have an increased risk of treatment-related mortality. Additionally,
data from large population studies have demonstrated a
survival benefit from adjuvant chemotherapy for older
women with hormone-negative or node-positive early stage
breast cancer.28,29
Results from randomized controlled clinical trials that
have specifically focused on older women with breast cancer,
though scant, have helped to fill the gap on chemotherapy
benefit for older women with early stage breast cancer.
In the largest study in this population to date, a CALGB and
Breast Cancer Intergroup study, 33 patients 65 and older
with early-stage breast cancer were randomly assigned to

either standard treatment (doxorubicin and cyclophosphamide [AC] for four cycles or cyclophosphamide, methotrexate, and 5-fluorouracil [CMF] for six cycles) or to an
experimental arm of single-agent capecitabine for six cycles.30 At a median follow-up of 2.4 years, the relapse-free
survival for patients receiving single-agent capecitabine was
inferior to that for women receiving standard therapy (hazard ratio [HR] 2.09 (1.38 3.17) p 0 0.0001), as was
overall survival (HR 1.85 (1.113.08) p 0.02). An
unplanned subset analysis demonstrated that standard
combination chemotherapy was particularly effective in
patients with hormone receptor-negative disease. Overall,
these results demonstrate that standard combination chemotherapy in the adjuvant setting, provides an overall
survival benefit for older women with breast cancer, particularly those with hormone-receptor negative disease.
The optimum chemotherapy regimen for treating breast
cancer in the adjuvant setting remains debatable. Regardless, anthracycline-based regimens have become the norm,
particularly for high-risk disease. However, the long-term
complications associated with anthracycline use include,
but are not limited to, dose-dependent cardiomyopathy. Age
is a risk factor for cardiac disease, including anthracyclinerelated cardiomyopathy. Often, this precludes the use of
anthracycline-based regimens in older women with breast
cancer. Jones and colleagues,31 in a US Oncology trial,
compared an anthracycline-based regimen (AC for four cycles) with a nonanthracycline taxane-based regimen docetaxel and cyclophosphamide (TC) for four cycles in 1,016
women with node-negative and node-positive disease. At a
median follow-up of 7 years, TC use was associated with
superior disease-free survival (HR 0.74, 95% CI (0.56
0.98) p 0 0.03) and overall survival (HR 0.69, 95%
(0.50 0.97) p 0 0.03) compared with AC. Sixteen percent
of the study population were 65 and older. Unplanned
subgroup analysis showed that TC was associated with
superior disease-free and overall survival in all age groups,
including older ages. TC therefore is a reasonable treatment
regimen in the adjuvant setting for older women, particularly those with pre-existing heart disease and other contraindications to anthracycline-based regimens. A recent study
has shown that this regimen is well tolerated in older
women.32 In another EBCTCG overview analysis in which
different polychemotherapy regimens for early stage breast
cancer were compared, adding a taxane to an anthracyclinebased chemotherapy regimen or a higher cumulative-dosage
anthracycline-based regimen reduced breast cancer mortality, on average, by one-third. This benefit was irrespective
of node status, tumor size, tumor grade, or age (70).33
No definitive conclusion could be drawn regarding women
70 and older because few women in that age group were
included in the meta-analyses.
For older women with node-negative, hormone-positive
breast cancer, gene-expression profiling analysis can be used
to identify women with high-risk disease who are likely to
benefit from chemotherapy. The most widely used assay for
this purpose is the 21-gene assay, which quantifies the
likelihood of breast cancer recurrence in women with nodenegative, estrogen receptor-positive breast cancer and predicts the magnitude of endocrine therapy and chemotherapy
benefit.34 To the extent that the 21-gene assay allows for
individualization of cancer treatment, it is indeed a useful

tool for the management of older patients with breast
cancer, who were well represented in the validation cohorts
(NSABP-14 and NSABP-20) for the assay. Moreover, the
predictive ability of the 21-gene assay has been found to be
independent of age.34 Older patients with low scores are not
likely to derive substantial benefit from chemotherapy,
whereas those with high scores may derive great benefit.
The benefit of adjuvant chemotherapy among patients with
intermediate scores on the assay is being evaluated in the
Tailor Rx study.
Adjuvant! Online is another tool that can assist in decision
making regarding the benefits of adjuvant endocrine therapy and chemotherapy for an individual patient. This online
tool (available at www.adjuvantonline.com) summarizes the
absolute benefit of chemotherapy and endocrine therapy,
taking into account the patients age, brief assessment of
comorbid medical illnesses, and tumor characteristics.35 To
further inform this discussion, the risks associated with
chemotherapy can be calculated with the predictive models
for chemotherapy toxicity8,9 as described earlier (See Adjuvant Systemic Therapy).
Adjuvant Trastuzumab for HER2-Positive
Breast Cancer
Amplification or overexpression of HER2 is seen in approximately 10% to 15% of invasive breast cancers in older
women,5 and it is associated with an unfavorable prognosis.
A substantial body of literature from phase III trials in
the adjuvant setting has demonstrated considerable benefit
in disease-free and overall survival when trastuzumab is
used either concurrently or sequential to chemotherapy
compared with chemotherapy alone.36-38 The main adverse
effect associated with trastuzumab use is cardiotoxicity. In
five phase III trials of adjuvant trastuzumab, the incidence
of severe heart failure (New York Heart Association class III
or IV), ranged from 0 to 3.9% among patients receiving
trastuzumab, compared with 0 to 1.3% among patients who
did not receive trastuzumab.39 In the Breast Cancer International Research Group (BCIRG) 006 study,38 two
trastuzumab-containing regimens (AC plus docetaxel and
trastuzumab and a nonanthracycline regimen of docetaxel,
carboplatin, and trastuzumab [TCH]) were compared with
standard chemotherapy alone. This study demonstrated
disease-free and overall survival benefits with the use of
trastuzumab plus chemotherapy compared with chemotherapy alone. There was no substantial difference between the
two trastuzumab-containing arms. Moreover, the incidence
of cardiotoxicity associated with the nonanthracycline-based
trastuzumab regimen was lower than that associated with
the anthracycline-based trastuzumab regimen.
Consistent with the under-representation of older women
in breast cancer clinical trials of chemotherapy, older women
have also been under-represented in clinical trials of trastuzumab therapy. With the notable exception of cardiac
dysfunction, which was found to be associated with increasing age (older than 50), limitations in data collection precluded a determination of whether the toxicity profile of
trastuzumab in older patients was different from that in
younger patients. The reported clinical experience was also
not adequate to determine whether the efficacy improvements (overall and disease-free survival) associated with

Table 1. Summary of Recommendations for Adjuvant Systemic Therapy in Early Stage Breast Cancer in Older Women
Node-negative,
Tumor size 1 cm
Node-negative,
Tumor size 1 cm
Node-positive
Endocrine-positive,
HER2-negative
No adjuvant therapy or
Consider hormonal therapy if tumor size 0.6 cm,
grade 2, or other high risk features
Hormonal therapy
Consider chemotherapy based on geneexpression profiling results
Hormonal therapy
Chemotherapy
Endocrine-negative,
HER2-negative
Consider chemotherapy if tumor size 0.6 cm plus
other high risk features
Chemotherapy alone
Chemotherapy alone
HER2-positive
Consider chemotherapy with trastuzumab if tumor
size 0.6 cm plus high risk features
Chemotherapy with trastuzumab

Add hormonal therapy if hormone
positive

Add hormonal therapy
if hormone positive
patients younger than 65. Regardless, in the absence of

contraindications, trastuzumab is currently recommended
for the adjuvant treatment of HER2-positive breast cancer,
even for older women. In older women, a nonanthracycline,
trastuzumab-containing regimen of TCH is often used because of the increased risk of cardiotoxicity associated with
the anthracycline and trastuzumab regimen.
In summary, the principles that govern recommendations
for systemic adjuvant treatment of breast cancer are the
same for younger and older women. Older women should
be offered guideline-recommended therapies (Table 1).
Broadly, these recommendations should be offered along
three clinically distinct subgroups based on tumor characteristics. (1) Older women with hormone receptor-positive
and HER2-negative breast cancer who should be offered
endocrine therapy regardless of node status. Gene expression profiling assay may be used to determine the added
benefit of chemotherapy for those with node-negative
disease; (2) older women with hormone receptornegative
and HER2-negative breast cancer (triple-negative breast
cancer) who should be offered adjuvant chemotherapy; and
(3) older women with HER2-positive disease who should
be offered chemotherapy with trastuzumab. In the last
group, women with hormone receptor-positive tumors
should also be offered endocrine therapy. Exceptions to
these guidelines may be made for older women in any of the
three subgroups who have node-negative disease and a
tumor less than 1 cm or for frail older women with limited
life expectancy, where close surveillance may be a reasonable alternative.
Adjuvant Radiation Therapy
Until recently, the guideline recommendation, regardless

of age, was for all women to receive radiation therapy
after breast-conserving surgery and for postmastectomy
radiation to be offered to women with a high probability of
local recurrence. A recent meta-analysis of the EBCTCG
supports these recommendations, showing that radiation
therapy decreased the 10-year risk of any first recurrence
from 35% to 19% and the 15-year risk of breast cancer death
from 25% to 21% among women treated with breastconserving surgery.40 Although the proportional reductions
in relapse were similar among all women, the absolute
benefits varied substantially by age, grade, estrogen receptor status, tamoxifen use, and extent of surgery. The authors
concluded that radiation therapy after breast-conserving
surgery halves the rate at which the disease recurs and
reduces the breast cancer death rate by about a sixth.
However, older women with small tumors can be spared
radiation therapy. In a landmark randomized controlled

study by Hughes and colleagues41,42 636 women 70 or older
who had undergone lumpectomy for stage I hormone
receptor-positive breast cancer were randomly assigned to
receive either radiation therapy and adjuvant tamoxifen
for 5 years or to adjuvant tamoxifen for 5 years alone. The
results demonstrated no substantial differences between
the two groups with regard to mastectomy rates for local
recurrence, distant metastases, or overall survival (median
follow-up of 12 years). Of the 49% of patients who died
during follow-up, 3% died of breast cancer. The only statistically significant difference was found in the rate of local
or regional recurrence at 5 years, (2% among women who
had radiation therapy compared with 9% who did not).
Based on results from this study, one may reasonably
consider lumpectomy (with surgically clear margins)
without radiation therapy for women 70 and older with
clinically negative lymph nodes, a tumor 2 cm or smaller,
and hormone receptor-positive breast cancer who agree
to take endocrine therapy. This strategy is limited by the
high rate of nonadherence and early discontinuation of
adjuvant systemic endocrine therapy among older
women.43,44 Omission of postoperative radiation therapy,
coupled with nonadherence to adjuvant systemic endocrine
therapy, may result in earlier recurrences and, ultimately,
poorer survival outcomes for older women. A favorable
outcome from this approach can be achieved only when
older women are adherent to prescribed oral endocrine
therapies. Adherence to prescribed endocrine therapy
should therefore be discussed and encouraged at follow-up
visits.
Conclusion
Breast cancer is a disease of aging. With minor differences, existing data support similar recommendations for
both younger and older women. However, there are agerelated differences in treatment patterns, with older women
less likely than younger women to receive standard therapies. Furthermore, survival outcomes lag behind that of
younger women. Closing the current gap in age-related
disparities in breast cancer survival will require that
older women are offered the same state-of-the-art treatment
as younger women, with a careful weighing of the risks
and benefits of each treatment in the context of the individuals preferences. Newer tools that estimate life
expectancy and toxicity as well as the potential benefits of
therapy should make it easier for oncologists to make better
treatment decisions with older patients. In addition, older
women should be encouraged to participate in breast cancer
clinical trials to generate additional efficacy and toxicity

data. Such information will provide further knowledge so
that oncologists can offer older women the best treatment

options.
Authors Disclosures of Potential Conflicts of Interest
Author
Cynthia Owusu*
Arti Hurria
Hyman Muss
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Amgen;
Genentech; GTx
Research
Funding
Expert
Testimony
Other
Remuneration
Abraxis
BioScience;
Celgene;
GlaxoSmithKline
Boehringer
Ingelheim; Eisai;
Pfizer; Sandoz
*No relevant relationships to disclose.
REFERENCES
1. Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011: the impact of
eliminating socioeconomic and racial disparities on premature cancer deaths.
CA Cancer J Clin. 2011;61:212-236.
2. DeSantis C, Siegel R, Bandi P, et al. Breast cancer statistics, 2011. CA
Cancer J Clin. 2011;61:409-418.
3. SEER stat fact sheets. http://seer.cancer.gov/csr/1975_2008/results_
single/sect_01_table.11_2pgs.pdf. Accessed March 8, 2012.
4. Chu KC, Tarone RE, Kessler LG, et al. Recent trends in U.S. breast
cancer incidence, survival, and mortality rates. J Natl Cancer Inst. 1996;88:
1571-1579.
5. Diab SG, Elledge RM, Clark GM. Tumor characteristics and clinical
outcome of elderly women with breast cancer. J Natl Cancer Inst. 2000;92:
550-556.
6. Smith BD, Jiang J, McLaughlin SS, et al. Improvement in breast cancer
outcomes over time: are older women missing out? J Clin Oncol. 2011;29:
4647-4653.
7. van de Water W, Markopoulos C, van de Velde CJ, et al. Association
between age at diagnosis and disease-specific mortality among postmenopausal women with hormone receptor-positive breast cancer. JAMA. 2012;
307:590-597.
8. Hurria A, Togawa K, Mohile SG, et al. Predicting chemotherapy toxicity
in older adults with cancer: a prospective multicenter study. J Clin Oncol.
2011;29:3457-3465.
9. Extermann M, Boler I, Reich RR, et al. Predicting the risk of chemotherapy toxicity in older patients: The Chemotherapy Risk Assessment Scale for
High-Age Patients (CRASH) score. Cancer. Epub 2011 Nov 9. doi: 10.1002/
cncr.26646
10. National Comprehensive Cancer Network. Clinical Practice Guidelines
in Breast Cancer. Version 1. 2012. www.nccn.org. Accessed February 28,
2012.
11. Goldhirsch A, Wood WC, Gelber RD, et al. Meeting highlights: updated
international expert consensus on the primary therapy of early breast cancer.
J Clin Oncol. 2003;21:3357-3365.
12. Wildiers H, Kunkler I, Biganzoli L, et al. Management of breast cancer
in elderly individuals: recommendations of the International Society of
Geriatric Oncology. Lancet Oncol. 2007;8:1101-1115.
13. Effects of chemotherapy and hormonal therapy for early breast cancer
on recurrence and 15-year survival: an overview of the randomised trials.
Lancet. 2005;365:1687-1717.
14. Love RR, Mazess RB, Barden HS, et al. Effects of tamoxifen on bone
mineral density in postmenopausal women with breast cancer. N Engl J Med.
1992;326:852-856.
15. Love RR, Wiebe DA, Feyzi JM, et al. Effects of tamoxifen on cardiovascular risk factors in postmenopausal women after 5 years of treatment. J Natl
Cancer Inst. 1994;86:1534-1539.
16. Blackman SB, Lash TL, Fink AK, et al. Advanced age and adjuvant
tamoxifen prescription in early-stage breast carcinoma patients. Cancer.
2002;95:2465-2472.
17. Partridge AH, Wang PS, Winer EP, et al. Nonadherence to adjuvant
tamoxifen therapy in women with primary breast cancer. J Clin Oncol.
2003;21:602 606.
18. Owusu C, Buist DS, Field T, et al. Tamoxifen discontinuance among
older women with estrogen-receptor-positive breast cancer. J Clin Oncol.
2006;24: (suppl; abstr 648).
19. Howell A, Cuzick J, Baum M, et al. Results of the ATAC (Arimidex,
Tamoxifen, Alone or in Combination) trial after completion of 5 years

adjuvant treatment for breast cancer. Lancet. 2005;365:60-62.
20. Coombes RC, Hall E, Gibson LJ, et al. A randomized trial of exemestane after two to three years of tamoxifen therapy in postmenopausal women
with primary breast cancer. N Engl J Med. 2004;350:1081-1092.
21. Goss PE, Ingle JN, Martino S, et al. A randomized trial of letrozole in
postmenopausal women after five years of tamoxifen therapy for early-stage
breast cancer. N Engl J Med. 2003;349:1793-1802.
22. Mouridsen H, Keshaviah A, Coates AS, et al. Cardiovascular adverse
events during adjuvant endocrine therapy for early breast cancer using
letrozole or tamoxifen: safety analysis of BIG 1-98 trial. J Clin Oncol.
2007;25:5715-5722.
23. Buzdar A, Howell A, et al. Comprehensive side-effect profile of anastrozole and tamoxifen as adjuvant treatment for early-stage breast cancer:
long-term safety analysis of the ATAC trial. Lancet Oncol. 2006;7:633-643.
24. Muss HB, Tu D, Ingle JN, et al. Efficacy, toxicity, and quality of life in
older women with early-stage breast cancer treated with letrozole or placebo
after 5 years of tamoxifen: NCIC CTG intergroup trial MA.17. J Clin Oncol.
2008;26:1956-1964.
25. Amir E, Seruga B, Niraula S, et al. Toxicity of adjuvant endocrine
therapy in postmenopausal breast cancer patients: a systematic review and
meta-analysis. J Natl Cancer Inst. 2011;103:1299-1309.
26. Crivellari D, Sun Z, Coates AS, et al. Letrozole compared with tamoxifen for elderly patients with endocrine-responsive early breast cancer: the
BIG 1-98 trial. J Clin Oncol. 2008;26:1972-1979.
27. Muss HB, Woolf S, Berry D, et al. Adjuvant chemotherapy in older and
younger women with lymph node-positive breast cancer. JAMA. 2005;293:
1073-1081.
28. Du XL, Jones DV, Zhang D. Effectiveness of adjuvant chemotherapy for
node-positive operable breast cancer in older women. J Gerontol A Biol Sci
Med Sci. 2005;60:1137-1144.
29. Elkin EB, Hurria A, Mitra N, et al. Adjuvant chemotherapy and
survival in older women with hormone receptor-negative breast cancer:
assessing outcome in a population-based, observational cohort. J Clin Oncol.
2006;24:2757-2764.
30. Muss HB, Berry DA, Cirrincione CT, et al. Adjuvant chemotherapy in
older women with early-stage breast cancer. N Engl J Med. 2009;360:20552065.
31. Jones S, Holmes FA, OShaughnessy J, et al. Docetaxel with cyclophosphamide is associated with an overall survival benefit compared with doxorubicin and cyclophosphamide: 7-year follow-up of US Oncology Research
Trial 9735. J Clin Oncol. 2009;27:1177-1183.
32. Freyer G, Campone M, Peron J, et al. Adjuvant docetaxel/cyclophosphamide in breast cancer patients over the age of 70: results of an observational
study. Crit Rev Oncol Hematol. 2011;80:466-473. Epub 2011 May 11.
33. Peto R, Davies C, et al. Comparisons between different polychemotherapy regimens for early breast cancer: meta-analyses of long-term
outcome among 100,000 women in 123 randomised trials. Lancet. 2012;379:
432-444.
34. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence
of tamoxifen-treated, node-negative breast cancer. N Engl J Med. 2004;351:
2817-2826.
35. Ravdin PM, Siminoff LA, Davis GJ, et al. Computer program to assist
in making decisions about adjuvant therapy for women with early breast
cancer. J Clin Oncol. 2001;19:980-991.
36. Piccart-Gebhart MJ, Procter M, Leyland-Jones B, et al. Trastuzumab

after adjuvant chemotherapy in HER2-positive breast cancer. N Engl J Med.
2005;353:1659-1672.
37. Romond EH, Perez EA, Bryant J, et al. Trastuzumab plus adjuvant
chemotherapy for operable HER2-positive breast cancer. N Engl J Med.
2005;353:1673-1684.
38. Slamon D, Eiermann W, Robert N, et al. Adjuvant trastuzumab in
HER2-positive breast cancer. N Engl J Med. 2011;365:1273-1283.
39. Ewer SM, Ewer MS. Cardiotoxicity profile of trastuzumab. Drug Saf.
2008;31:459-467.
40. Darby S, McGale P, et al. Effect of radiotherapy after breast-conserving
surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis
of individual patient data for 10,801 women in 17 randomised trials. Lancet.
2011;378:1707-1716.
41. Hughes KS, Schnaper LA, Berry D, et al. Lumpectomy plus tamoxifen
with or without irradiation in women 70 years of age or older with early
42. Hughes KS, Schnaper LA, Cirrincione C, et al. Lumpectomy plus
tamoxifen with or without irradiation in women age 70 or older with early
breast cancer. J Clin Oncol. 2010;28:15s (suppl; abstr 507).
43. Owusu C, Buist DS, Field TS, et al. Predictors of tamoxifen discontinuation among older women with estrogen receptor-positive breast cancer.
J Clin Oncol. 2008;26:549-555.
44. Partridge AH, LaFountain A, Mayer E, et al. Adherence to initial
adjuvant anastrozole therapy among women with early-stage breast cancer.
J Clin Oncol. 2008;26:556-562.
Management of Small T1a/b N0

Breast Cancers
By Anthony D. Elias, MD
Overview: T1ab N0 breast cancer generally has excellent

prognosis. Adverse prognostic factors include HER2 disease, ER-negative disease, high-grade histology, T1b, and
young age of patient. These patients are largely excluded from
most trials, and to date, no prospective studies for this group
HE INCIDENCE of small ( 1 cm) node-negative breast

cancers (BC) is increasing in mammography-screened
populations.1 About two-thirds of all breast cancers (BCs) in
the Surveillance, Epidemiology, and End Results (SEER)
registry are T1; approximately 72% and 28% are T1b and
T1a, respectively.1
Approximately 50% to 60% of T1abN0 human epidermal
growth factor receptor 2 (HER2)-positive tumors are estrogen receptor (ER) positive.2-4 Approximately 10% to 15% of
patients with T1abN0 have HER2-positive disease.2-9 Several studies have reported that T1mic and T1a are more
likely to be ER negative and HER2 positive compared with
T1b, although an early study of OncotypeDx suggested that
among the ER-positive tumors, a greater percentage of the
T1b had low recurrence score (RS).4,10,11 In the Kwon report,
HER2 was positive in 46%, 23%, and 13%, and ER positive
in 42%, 66%, and 78% of T1mic, T1a, and T1b, respectively.4
Small node-negative BC are generally associated with
excellent outcomes, with disease-free (DFS) and relapse-free
survivals (RFS) in excess of 90%, and with survivals exceeding 95% at 5 to 10 years. Older studies have shown 10-year
RFS between 83% and 91%.12-14 However, these tumors are
quite heterogeneous, and some subsets of tumors have a far
higher risk of relapse. Because these patients are generally
not included in clinical trials, outcomes are poorly characterized, management is uncertain, and treatment varies
considerably from one oncologist to another.
This article will review the outcomes literature for T1ab
N0M0 BCs, focusing on articles presented in the last decade,
at a time where molecular testing of ER, progesterone
receptor (PR), and HER2 have become routine. For more
details of earlier literature, refer to the excellent review by
Hanrahan et al.15 Because of a recent spike in interest in the
oncology community, the majority of articles on this topic
have been published in manuscript or abstract form in 2010
and 2011.
T1a tumors are 0.5 cm or smaller, while T1b are larger
than 0.5 to 1 cm or smaller. Typical definitions include RFS:
date of diagnosis or initial surgery to date of recurrence
(distant or local), death from any cause, or last follow-up.
DFS includes contralateral new BCs as events. Distant DFS
only includes metastatic events.
yet reported. Treatment guidelines are vague and treatment

inconsistent. As yet, in the HER2 population, little experience with targeted therapy has been reported. Prospective
trials are needed.
were chemotherapy naive with ER-negative BC, the 8-year

RFS was only 79% with an overall survival (OS) of 93%. A
90% 8-year RFS was achieved when chemotherapy was
given. For the 264 patients with ER-positive BC, an 8-year
RFS of 86% and an OS of 90% were achieved with surgery
alone. For those treated with tamoxifen, a 93% 8-year RFS
was noted.
Wood et al studied the outcomes of 282 consecutive women
with T1abN0 BCs at Emory.17 The mean follow-up was 7
years. Tamoxifen and chemotherapy was administered to 93
and 20 patients, respectively. One patient had local recurrence and two suffered metastatic spread despite tamoxifen
for an estimated 10-year distant DFS (DDFS) of 98.7%.
Joensuu described a cohort of 852 Finnish patients with
T1N0 BC, of whom 12% had HER2-positive disease and 75%
had ER-positive disease.18 Median follow-up was 114
months. Three hundred and thirteen patients had T1ab
disease with a 9-year DFS of 93%; for T1a 100%. Nine-year
DDFS was much better for HER2-negative compared with
HER2-positive BCs (89% vs. 73%, p 0.0003). In the T1b
group, this difference was greater (95% vs. 67%). Thirty
patients had TN BC, of whom nine received chemotherapy.
No relapses were observed. Adverse prognostic factors included T size (in mm), HER2 positivity, Ki67 greater than
20%, and possibly grade 2 to 3 disease.
Fisher et al reported on the outcomes of the patients with
T1abN0 disease from the NSABP B21 study who were
randomly selected to receive lumpectomy plus tamoxifen,
lumpectomy plus radiation therapy, or all three.19 Median
follow-up was 87 months. HER2 status was unknown. Twothirds received endocrine therapy, and the remainder received observation. Radiation therapy reduced ipsilateral
breast tumor recurrence (IBTR; hazard ratio [HR] 0.19) as
did tamoxifen (HR 0.37). Eight-year DFS for the best arm
(preoperative cetuximab and radiation [XRT] plus tamoxifen) was 84.4%. An update was provided in 2007 with
14-year follow-up.20 Adverse prognostic factors were grade
3, nontubular histology, and, for survival, lymphovascular
invasion (LVI).
Chia et al reported the outcomes of 2,026 patients in
British Columbia with N0 BC treated in 1986 to 1992.21
Seventy percent had no adjuvant therapy. Ten-year RFS
was 76% (and only 66% in the HER2 population [10% of
Literature Review
Studies in Which 10% of Patients Received Adjuvant Chemotherapy
and No Trastuzumab
Fisher et al presented the National Surgical Adjuvant

Breast and Bowel Project (NSABP) experience with T1abN0
BC in the B-06, B-13, B-14, B-19, and B-20 studies (Table
1).16 Median follow-up was 8 years. For the 61 patients who
10
From the University of Colorado Cancer Center, Anschutz Medical Campus, Aurora, CO.
Address reprint requests to Anthony D. Elias, MD, University of Colorado Cancer Center,
Anschutz Medical Campus, MS 8117, Research Tower South, 12801 East 17th Avenue,
Room 8111, Aurora, CO 80045; email: anthony.elias@ucdenver.edu.
1092-9118/10/1-10
MANAGEMENT OF T1 BREAST CANCERS
the group]). Of the 307 T1abN0 HER2-negative BCs, 10-year

RFS was approximately 94%. Of the 21 T1abN0 HER2positive BCs, the 10-year RFS was approximately 85%. T1b
fared a bit worse than T1a.
Gonzalez-Angulo et al analyzed a University of Texas
M. D. Anderson Cancer Center (MDACC) dataset (later
reported by Theriault et al9) plus a confirmatory dataset
from the Institutes Bordet and Leoben.22 Nine hundred and
sixty five patients from MDACC and 350 patients from the
other two institutions with T1abN0 were included. Endocrine therapy was given to 55% of the MDACC patients;
treatment details were sparse from the other group. Overall
5-year RFS was 92% and 96% for the two groups but only
77% and 87% for the HER2-positive subset. Adverse prognostic factors included HER2-positivity, ER-negativity, age
younger than 50, and grade 3 tumor. Tumor size was not a
factor. The TN group had a 5-year RFS of 83% at MDACC,
but it was substantially better (95%) at the other institutions. It is uncertain whether this was a result of differences
in treatment.
Theriault et al summarized the MDACC experience for
1,012 patients with T1abN0 BC in 1990 to 2002.9 Median
follow-up was roughly 60 months. Patients who had received
chemotherapy or trastuzumab were excluded from this article. Of the 771 patients with ER-positive BC, 61% received
endocrine therapy. Of the 98 and 143 patients with HER2positive and TN BC, endocrine therapy was given to 46%
and 21%, respectively. Five-year DDFS was 97.5%, 86.9%,
and 96.3% for the patients with HR-positive, HER2-positive,
and TN disease, respectively. Five-year RFS was 94.5%,
77.2%, and 84.9%, respectively. Adverse prognostic factors
included age younger than 35, HER2 positivity, and less so
TN BC.
Cancello et al reported on the European Institute of
Oncology experience for patients with T1mic/a/bN0 BC
treated in 1997 to 2005.8 It is not clear how these patients
overlap with a prior report from Curigliano et al.23 Median
follow-up was 76 months for the 1,691 patients. Overall
treatment consisted of endocrine therapy in 35% (93% of all
ER) and chemotherapy in 10% (54% and 71% of HER2
and TN BC). No trastuzumab was administered. Using
cutoffs of 1% for ER/PR and 14% for Ki67, patients were
categorized by IHC to have BCs that were luminal A (881,
52%), luminal B (532, 31%), luminal B HER2 positive (101,
6%), HER2 positive/ER negative (82, 5%), and TN (95, 6%).
Five-year BC-specific survival was approximately 98%, 97%,
95%, 91%, and 91%, respectively. Adverse prognostic factors
included age younger than 35, HER2 positivity, TN, and
LVI.
KEY POINTS
T1ab N0 breast cancer generally has excellent prognosis.

Adverse prognostic factors include HER2 disease,
ER disease, high-grade histology, T1b, and younger
age.
Vague treatment guidelines exist.
No prospective trials have yet been reported.
For the HER2 population, there is almost no trastuzumab experience in the absence of chemotherapy.
Rouanet et al reported the outcomes of 703 French patients with T1abN0 BC treated in 1999 to 2004.10 Only 6%
had HER2-positive disease. Of these, 50% had dual positive
disease. Treatment consisted of endocrine therapy in 41%,
chemotherapy in 10%, and observation in 55%, which
achieved a 5-year DFS of only 74%. For the 661 patients
with HER2-negative disease, 10% had TN disease and 90%
ER-positive disease. Treatment for this group was endocrine
for 80%, chemotherapy for 5%, and observation for 17%,
which achieved a 5-year DFS of 95%. Adverse prognostic
factors included HER2-positivity, ER-negativity, age
younger than 50, no adjuvant therapy, and grade 3 disease.
Five-year DFS for the TN cohort was 91%.
Livi et al presented 704 Florentines with T1abN0 BC
treated in 2002 to 2008.2,24 Median follow-up was 59
months. No chemotherapy or trastuzumab was given; endocrine therapy was given in 64%. Five hundred and fifty-nine
had ER-positive/HER2-negative disease, 75 had HER2positive disease of whom 55% had ER-positive disease
and 70 had TN BC.24 Five-year DDFS was 97.8%, 92%, and
91.8%, respectively. Overall 5-year DFS was 96.5%. Adverse
prognostic factors included age 50 or younger and ERnegative and HER2-positive disease.
Studies in Which >10% Patients Received Adjuvant Chemotherapy,
but No Trastuzumab
Colleoni et al reported on a cohort of patients with T1mic/

a/b N0 disease from the European Institute of Oncology
treated in 1997 to 2001.14 Median follow-up was only 43
months. Of the 425 patients, 358 had ER-positive BC (11%
HER2), and 60 had ER-negative BC (40% HER2, 60%
TN). Most received adjuvant therapy. Of the patients who
were ER positive, only six relapsed following treatment
(endocrine 88%, chemotherapy 12%, observation 11%). Of
the patients who were ER negative, only three relapsed
following treatment (chemotherapy 60%, observation 40%).
Adverse prognostic factors included Ki67 greater than 20%,
age younger than 35, and PR-negative disease.
Hanrahan et al analyzed the outcomes of patients with
T1abN0 disease in the SEER registry from 1988 to 2001.26
Treatment details are not available. By 10 years, only
approximately 4% died of BC, whereas an additional 20%
died of other causes, mostly cardiovascular. For OS, prognostic factors included black race, T1b, and PR-negative
disease. For each mm tumor size, the HR for OS was 1.02.
Prognostic factors for BC-specific mortality were age
younger than 50, grade 3, ER negativity, PR negativity, and
fewer than six lymph nodes in the axillary lymph node
dissection (ALND).
Curigliano et al reviewed 2,130 patients with T1abN0 BC
treated at the European Institute of Oncology in 1999 to
2006.23 The extent of overlap with Cancellos presentation is
not clear. Median follow-up was 55 months. A subset had
immunohistochemical (IHC) testing: 79 with HER2-positive/
ER-positive disease and 71 with HER2-positive/ER-negative
disease; 158 patients with ER-positive/HER2-negative BC
were matched by HR status, age, and year of surgery.
Treatment for the ER-positive/HER2-negative group was
endocrine in 93%, chemotherapy in 1% and observation in
6% resulting in a 5-year DFS of 99%. Treatment for the dual
positive group was endocrine in 66% and chemotherapy in
25%, resulting in a 5-year DFS of 92%. Treatment for the
HER2-positive/ER-negative group was endocrine in 3%, che-
11
ANTHONY D. ELIAS
Table 1. T1ab N0 Breast Cancer Literature Review (Papers Published 20022011)
Author
F/U
(mos)
Types of Treatment
Institution
Fisher
16
NSABP B21
2002 19891998
Wood
Joensuu
17
18
Emory
Finland
2002
?
84 mean T1abN0
2003 19911992
114
T1N0
T1ab
TN
HRHER2-
Fisher
Chia
GonzalezAngulo
19,20 NSABP
Subset of
1009
in B21
21
9,22
Year
Pt Accrual
Yrs
Ref(s)
2007 19891998
British
Columbia
2008 19861992
MDACC
2009 19902002
87
134
Stage
T1abN0
T1abN0
N0
T1ab HER2T1ab HER2
74
T1abN0
Bordet
Leoben
Theriault
Cancello
Rouanet
Livi
10
2,24
MDACC
2011 19902002
Florence
2011 19992004
60
76
77
2011 20022008
59
Colleoni
16
Eur Inst
2004 19972001
Oncology
43
Hanrahan
26
SEER
2007 19882001
64
Curigliano
23
Eur Inst Onc 2009 19992006
55
12
T1abN0
ER
HER2
TN
T1m/a/bN0
Luminal A
Luminal B
Luminal B HER2
HER2
TN
T1abN0
T1a
T1b
T1a/bN0
T1a/b H
T1a/b TN
T1mic/a/b N0
ER
ERT1abN0
T1abN0
ER/HER2
matched ER/HER-/HER2
Matched ER-/H-
%ER
1009
57% (30% unk)
332
57
334
54
334
59
282
NR
852
75
313
30
396*
*#s dont add up
638
n/a
204
228
206
2026
307
21
965
350
Eur Inst Onc 2011 19972005
ONCO
# Pts
1012
771
98
143
1691
881
532
101
82
95
703*
131
572
424 others*
661
42
704
75
70
425
358
60
51,246
2130
79
158
71
71
%HER2
nd
nd
12
%OBS
70
93
nd
%ET %Chemo %anti HER2
33
67
XRT only
0
XRT
100
no XRT
100
NR
33
95
32
68
XRT only
0
XRT
100
no XRT
100
10
70
82
76
77
86
61
48
10
45
55
76
10
39
54
79
NR
61
46
21
35
73% any T
39% any T
11
100
100
100
0
0
93
77
97
0
0
100
100
0
6
14
4
unk
0
100
58% (30% unk)
30
0
10
20
92
94
94
1
3
83
1
7
7
54
71
6
17
55
36
80
36
64
5
5
0
0
0
0
16
11
40
NR
66
88
0
NR
16
12
60
NR
0
0
0
0
10
6
54
31
89
94
4
3
25
1
43
66
0
0
0
0
0
90
90
50
84
55
0
84
13
nd

Table 1. T1ab N0 Breast Cancer Literature Review (Papers Published 20022011) (Contd)
Measure of Efficacy
Time
%DFS
%RFS
%DDFS
%OS
%IBTR
Adverse Factors
81.6
84.4
77.8
96.7
98.5
96.8
98.7
7y
4.2 y
94
93.4
94
90.7
97.8
83.5
size, HER2, Ki67
20%, G 1
93
100
5y
5 y BCS
66 versus
76 HER2
94
85
92
71 versus
82 HER2
94
77 (HR 2.7)
HR 3.9
96
97
86 (HR 5.3)
HR 2.8
T1b a bit worse

96
97
87
91.5
96.3
94.5
97.5
77.2 (HR 4.98) 86.9 (HR 4.7)
84.9 (HR 2.71) 96.3 (HR 2.1)
5y
49 T1a 100% DFS
HER2 Test
HER2 unk
HER2 unk
IHC 3 10%
or CISH
DFS: high nuclear

IBTR: nuclear grade, HER2 unk
grade, non-tubular
DCIS present
82.1
89.2
OS: LVI
77.8
89.8
82.2
80.5
60.6
56
61.5
10 y
Comments
71% T1b
IBTR: HR 0.19 for XRT;
HR 0.37 for Tam
8y
IHC 3 10%
or FISH
HER2HER2
TN
5y RFS TN 95% (n 125);
IHC 3 10%
no diff T1a versus T1b
or FISH
HER2HER2
Age 35, HER2 worse Fig 1
IHC 3 10%
than TN
excluded CT and trastuzumab
or FISH
more T1a
94.6
HER2, ER-, age

50, high grade
ER-5y RFS 83.3%
Age 35, HER2 then

TN, LVI
ER/PR at 1% threshold;
Ki67 14%
Fig 1
IHC 3 10%
or FISH
SP3 Ab
98
97
95
91
91
IHC 3 10%
Tmic overrepresented (27%)
5y
97
HER2, ER-, 50 y,
no adj therapy, G3
T1a more HER2, ER50% HER also ER

TN 5 y DFS 91%
IHC 3 or DISH
age 50; ER-, HER2
HR 3.66 for HER2 DFS

63% T1b
IHC 10% 3
or FISH
Ki67 20% HR 12.9;

35, PR-
76% T1b; more G3 in

T1mic, no events in T1mic
IHC 3 10%
BCSM: 50, G3,

ER-, PR-, 6LN in
ALND; OS: 50,
AA, T1b, PR-
4% died of BC;
24% all-cause
HER2 unk
95
74
5y
96.5
92
91.8
93.7
90.3
1
1
1
4y
6 events
3 events
10 y
5y
HR 1.02 for
1 mm T size
HR 2.4
92 (HR 5.2)
99
91
92
HR 5.1 for HER2

T1b, multifocal
T1a
T1a
T1a
T1a
DFS 88%; T1b 95%

97%; T1b 99%
93%; T1b 85%
87%; T1b 94%
IHC 3 10%
or FISH
13
ANTHONY D. ELIAS
Author
Kwon
Park
Lai
Tanaka
Ref(s)
4
27
28
Institution
Seoul
Seoul
Taiwan
Japan
Year
Pt Accrual
Yrs
2010 20002006
F/U
(mos)
61
2010 19942004
61
2011 19952006
289
54
31
33
48
2011 20012007
46
Fehrenbacher
29
KPNC
2010 20002006
70
Amar
30
Mayo
2010 20012005
34
Horio
Wong
McArthur
Nagoya
3
Singapore
subset
subset
subset
6
MSKCC
2011 20032007
2011 19892009
2011 20022008
20022004
52
57
Shao
14
T1ab
N0 or Nx
TN
non-TN
T1N0
T1N0 HT1N0 H
T1a/b H
T1abN0 HER2
T1a
T1b
T1abN0
ER
HER2
TN
T1a/b
HT1a
T1b
H
T1a
T1b
TN
T1N0
T1N0 HER2
T1c
T1a/b HT1a/b H
T1N0 HER2
Beth Israel NYC
2011 19952008
42
%ER
%HER2
375
120
93
162
31
65
56
370
57(13%)
67
42
66
78
26
46
23
13
HR/HER2 any
ER/HER2
HER2/ERTN
377
323
47 of 311
454
376
78
28
237
116
121
421
364
28
29
%OBS %ET
26
62
%Chemo
%anti HER2
16
34% of ER82
9 (not dual )
NR
74
36
94
25
65
64
65
29
36
18
17
0
0
0
NR
NR
77
44
14
45
100
NR
NR
NR
44
48
36
4
25
12.9
36.3
6
3
39
28
37
36
55
62
9
4
41
48
32
24
17
32
43
35
53
12
9
16
NR
78
47
8
24
0
9
0
60
59
63
61
66
42
100
100
49
59
52
65
87
50
T1a/b H
T1abN0 HER2
HR
HR-
59
100
0
100
100
100
T1abN0
HR
HER2
TN
658
494
109
55
75
100
27
0
17
0
100
0
36
41
# Pts
267
225
42
183
42
23
19
10
519
89
315
170
34
261
106
45
155
54
205
121
84
T1a/b H
20052008
subset
31
AERIO/Unicancer 2011 20002010
32
T1mic/a/bN0
T1mic
T1a
T1b
HER2HR
HER2HRTN
T1abN0
T1abN1
78
subset
Peron
Types of Treatment
Stage
86
16
36
8.4
6.8
9.9
1
18
17
18
NR
54
90
NR
NR
NR
55
NR
NR
NR
NR
10
31
42
100
100
45
0
0
17

Measure of Efficacy
Time
%DFS
5y
%RFS
%DDFS
%OS
%IBTR
Adverse Factors
age 35, TN
97.2
Comments
48% MRM
HER2 Test
IHC 3 (10%?)
96.8
98.5
92.6
5y
HER2, TN
for N1, multi,
HER2, TN
HR 7.2
HR 5.5
5y
98
92
90 (HR 5.7)
90 (HR 6)
T1b TN DDFS 84%,

HER2 87%
N1, HER2, TN
85.6
95.5 (HR 2.01)
97.4
3 M1, 54 N1 (15%)
IHC 3 30%
or FISH
HR 4.16 HER2, HR 3.37 TN
5y
HER2, ER-, grade 3,

premenopausal, LVI
97.2
88
5y
IHC 3 10%
or FISH
94.2
97.4
91.1
96.5
99.1
94
2.9
ER- 15%
ER- 51%
1 cm tumors have 12.5% 5 y
distant recurrence risk
3y
DAKO 3
or FISH
IHC 3 10%
or FISH
IHC 3 10%
or FISH
99
89
83
5y
IHC 3 or FISH
97.7
90.5
5 y,
10 y OS
T1b 79%
92.3
95.7
98.6
83.7
95.7
85.1
89.9
83.3
100
99.2
HER2 (HR 4.1 for DFS);

size for DDFS
5 y DFS HER2 83.7%

ER- 39%
3.5
14.9
HER2 (but not OS)
3y
More LVI, T1c in T-treated group

82
78
97
95
95
97
100
100
5y
IHC 3 or FISH
97
98
99
98
96 if treated w/chemo/
trastuzumab
86 if not treated
5y
DAKO 3
or FISH
25% T1a; 3 pts trastuzumab alone
TN
Chemo given for ER-, G2-3,

high mitoses
35% T1a
97.9
95.6 (HR 1.64)
93.5 (HR 3.13)
Abbreviations: Pt, patient; yrs, years; F/U, follow-up; mos, months; ER, estrogen receptor; HER, human epidermal growth factor receptor; OBS, observation; ET,
endocrine therapy; DFS, disease-free survival; RFS, recurrence-free survival; DDFS, distant disease-free survival; OS, overall survival; IBTR, ipsilateral breast tumor
recurrence; NSABP, National Surgical Breast and Bowel Project; unk, unknown; nd, not done; XRT, radiation therapy; HR, hazard ratio; Tam, tamoxifen; NR, not
reported; IHC, immunohistochemical; CISH, chromagen in situ hybridization; PR, progesterone receptor; LVI, lymphovascular invasion; SEER, Surveillance,
Epidemiology, and End Results; BCSM, breast cancer-specific mortality; LN, lymph node; ALND, axillary lymph node dissection; AA, African American; BC, breast
cancer; FISH, fluorescence in situ hybridization; T, tumor; SP3 Ab, SP3 antibody; MDACC, M. D. Anderson Cancer Center; TN, triple-negative; MRM, modified radical
mastectomy; BCS, breast cancer survival; CT, computed tomography; DAKO, DAKO, Inc.; MSKCC, Memorial Sloan-Kettering Cancer Center; KPNC, Kaiser Permanente
Northern California; ONCO, ONCO Languedoc-Roussillon Network.
15
ANTHONY D. ELIAS
motherapy in 43%, and observation in 54%, resulting in a

5-year DFS of 91%. Adverse prognostic factors included T1b
compared with T1a and multifocal BC in the HER2-positive
group.
Kwon et al treated 375 Koreans with T1mic/a/bN0 BC in
2000 to 2006.4 Median follow-up was 61 months. Treatment
consisted of observation in 26%, chemotherapy in 16%, and
endocrine therapy in 62%. No trastuzumab was given.
Five-year RFS was 97.2%. Adverse prognostic factors included age younger than 35, and TN disease. Five-year RFS
was 96.8%, 98.5%, and 92.5% for the HER2-positive/ERpositive, HER2-positive/ER-negative, and TN groups.
Park et al reported on 370 Korean patients with T1abN0
BC treated in 1994 to 2004.27 Median follow-up was 61
months. Treatment consisted of endocrine therapy in 74%
and chemotherapy in 36%. No trastuzumab was given.
Five-year DDFS was approximately 98%, 92%, 90%, and
90% for HR-positive, HER2-positive/ER-positive, HER2positive/ER-negative, and TN disease. Adverse prognostic
factors included T1b compared with T1a, TN, and HER2positive disease. The 5-year DDFS for the T1b TN and
HER2-positive groups was 84% and 87%, respectively.
Lai et al reported on 377 Taiwanese patients with T1ab
BC treated in 1995 to 2006.28 Fifty-four patients had node
involvement and three had metastatic disease. The overall
five-year DFS was 85.6% and the prognostic factors were N1,
HER2-positive, and TN disease.
Tanaka et al reported on 454 Japanese patients with
T1N0 BC treated from 2001 to 2007.5 Median follow-up was
46 months. Treatment followed St. Gallen guidelines, thus
only 14% of the T1N0 HER2-negative group and 45% of the
HER2-positive group received chemotherapy. No trastuzumab was given. Five-year RFS was 97.2% and 88% for the
HER2-negative and HER2-positive groups. By univariate
analysis, HER2-positivity, ER-negativity, grade 3 histology,
premenopausal status, and LVI were adverse prognostic
factors. By multivariate analysis, HER2-positivity and LVI
retained significance. Only 28 patients had T1abN0 HER2positive disease.
Fehrenbach et al reported the outcomes of 237 patients
with T1abN0 HER2-positive disease treated at Kaiser Permanente Northern California (KPNC) in 2000 to 2006.29
Median follow-up was 70 months. T1a disease was present
in 49% of patients and ER-positive disease in 59%. Chemotherapy was given to 17% (29% in T1b, 4% in T1a) and
trastuzumab to 8% (generally with chemotherapy). Fiveyear DDFS was excellent (99.1 in T1a and 94% in T1b).
Studies Including the Use of Trastuzumab
Amar et al reviewed the experience of the Mayo Clinic

with patients with T1abN0 in 2001 to 2005.30 Median
follow-up was 34 months. Of 421 patients, 87% had ERpositive, 7% HER2-positive, and 7% TN disease. Treatment
consisted of hormone therapy in 44%, chemotherapy in 7%,
and trastuzumab in 1%. Of the 28 patients with HER2positive disease, four had trastuzumab plus chemotherapy
and seven had chemotherapy alone. Endocrine therapy was
given to 48% of ER-positive BC and 36% of HER2-positive
BC. Chemotherapy was given to 28% of patients with TN
disease. Three-year RFS was 99% in ER-positive BC but
only 89% and 83% in patients with HER2-positive and TN
disease.
Horio et al reported on 267 patients from Nagoya with
16
T1abN0 BC treated in 2003 to 2007. Median follow-up was

52 months. Overall treatment was observation in 27%,
chemotherapy in 9%, and endocrine therapy in 55%. Of the
42 patients with HER2-positive BC, 12% received trastuzumab together with chemotherapy. Five-year RFS was
97.7% and 90.5% in the ER-positive/HER2-negative and
HER2-positive groups, respectively. There were no obvious
differences in relapse in the T1a group compared with the
T1b group. Only 10 patients had TN BC.
Wong et al published their institutional outcomes for 519
patients with T1N0 BC treated in Singapore from 1989 to
2009.3 Treatment details were not provided, although endocrine therapy was given to most patients with ER-positive
disease, chemotherapy to approximately 10%, and chemotherapy plus trastuzumab to three patients. Median
follow-up was 57 months. Five-year DFS, DDFS, and OS
were 92.3%, 95.7%, and 98.6%, respectively. Major prognostic factors were the presence of HER2 (HR 4.1 for DFS)
and tumor size (for DDFS). The 17% who had HER2-positive
disease had a 5-year DFS of only 83.7% and an OS of 83.3%.
Two hundred and four patients had T1abN0 disease, of
whom 34 had tumors positive for HER2. The patients with
T1ab HER2-negative disease did well with a 5-year DFS of
95.7%, OS of 100%, and an IBTR of only 3.5%. However, the
patients with T1ab HER2-positive disease fared less well
with a 5-year DFS of 85.1%, OS of 99.2%, and an IBTR of
14.9%.
McArthur et al summarized the MSKCC experience with
the T1N0 HER2-positive BCs in the pre-trastuzumab era of
2002 to 2004 and the post-trastuzumab era of 2005 to 2008.6
Median follow-up was 78 and 36 months for the two groups.
In the pre-trastuzumab era, 66% of the cohort received
chemotherapy (42% in the T1ab subset), and 59% (essentially all of those ER) received endocrine therapy. Threeyear DFS was only 82% (78% for the T1ab group), although
3-year DDFS was 95% to 97%. A high rate of local recurrence
was observed. In the post-trastuzumab era, all patients
received chemotherapy plus trastuzumab and 61% (essentially all of those ER) received endocrine therapy. The
3-year DFS was 97% (95% in the T1ab subset) and the DDFS
was 100%. This trial certainly suggests that aggressive
treatment, including anti-HER2 treatment, can reduce relapses but at the risk of overtreatment in a large proportion.
Peron et al updated the outcomes of 205 patients with
T1abN0 HER2-positive BC treated in 2000 to 2010.31 Median follow-up was 41 months. Adjuvant therapy consisted
of endocrine therapy in 54% (90% of the HER2/ER group)
and chemotherapy with trastuzumab in 44%, chemotherapy
in 5%, and trastuzumab alone in three patients. Five-year
RFS was 96% if treated with chemotherapy plus trastuzumab and only 86% if not treated. Chemotherapy was
selectively given for ER-negative disease, grade 2 to 3
disease, and high mitotic count.
Shao et al presented a cohort of 658 patients with T1abN0
BC from Beth Israel Deaconess Medical Center treated in
1995 to 2008.32 Median follow-up was 42 months. Four
hundred and ninety-four patients had ER-positive BC, of
whom 10% received chemotherapy, and endocrine therapy
was not reported. One hundred and nine patients had
HER2-positive BC (27% ER as well), of whom 14% received
chemotherapy, and 17% received chemotherapy plus trastuzumab. Fifty-five patients had TN BC, of whom 42% received chemotherapy. Five-year DFS was 97.9%, 95.6%, and
93.5%, respectively. Adverse prognostic factors included TN

disease. Hazard function did not reach statistical significance for HER2-positive disease.
Local-Regional Management
A major trend in BC therapy is the lessening extent of

axillary surgery. ALN dissection is completed for patients
with clinically node-positive disease but not for patients
with sentinel node (SN)-negative disease and is controversial even for the patients with clinically negative, SNpositive disease. SN sampling is occasionally omitted for the
elderly. Since ALN involvement is less common for small
primary tumors, there is a great temptation to avoid this
surgery. In various series of small primary tumors, the
likelihood of ALN involvement is between 3% and 37%.33 In
this series of 888 T1ab tumors with a 12% ALN involvement
rate, factors associated with ALN involvement in these
small tumors included LVI (HR 12.63), perineural invasion (HR 5.47), grade 3 tumor (HR 3.07), and ERnegative disease (HR 1.84).33
Rivadeneira et al found that 18% of 919 patients (199 T1a)
had ALN metastases following a standard ALND.34 Although tumor size was predictive, even the T1a group had a
16% rate of ALN involvement. Other factors included grade
3, LVI, and age younger than 50. These risk factors are
consistently observed in other series.35 HER2 status was not
tested. In the very best group of older patients with welldifferentiated T1a tumors without LVI, 13% still had metastases.34 The authors recommend routine use of SLN biopsy.
Literature Conclusions
T1abN0 BC generally has an excellent prognosis. Consistent adverse prognostic factors include HER2-positive disease, ER-negative disease, high grade, T1b, and age younger
than 50 years. Because most of the articles treated these
patients variably, these prognostic factors actually are
mixed prognostic/predictive factors demonstrating benefit of
specific therapies. Grade is problematic because of poor
concordance between expert pathologists in its determination. Ki67prognostic in many articleslacks standardization between pathology laboratories. Age is a complex
variable: low comorbidity rates and high life expectancy
allow extended time for BC events; premenopausal status
with different hormonal milieu; different genetic drivers to
develop BC in the young; a different mix of molecular
subtypes; and independent of subtypes, a more aggressive
metastasis pattern in the young, perhaps related to the
immunologic and inflammatory signals generated by pregnancy and weaning.36
In addition to these factors, which ultimately reflect
molecular subtype, size, and premenopausal status, the fact
remains that most events in these patients with good prognosis are unrelated to BC. Comorbiditiesparticularly cardiovascular healthas competing risks may represent up to
80% of all events, particularly in the elderly.22 Thus, the
best endpoints to help us make decisions may not be OS,
DFS, or RFS but the balancing of DDFS (or BCSS) and the
risks of treatment.
Most adjuvant trials find a strong correlation between an
intervention and the relative risk reduction independent of
stage. Thus, we expect approximately 25% risk reduction
with older polychemotherapy, approximately 40% risk re-
duction with docetaxel plus cyclophosphamide (TC), a reduction of approximately 41% with tamoxifen, approximately
50% with aromatase inhibitors, and approximately 50%
reduction with the addition of trastuzumab to chemotherapy. Thus, one can estimate the magnitude of absolute risk
reduction with these interventions by estimating the absolute risk if untreated.
Guidelines
Both major guidelines (National Comprehensive Cancer

Network [NCCN] and St. Gallens Expert Panel) require the
oncologist to consider much, but detailed instructions are
nebulous, as paraphrased below.
NCCN 2009. If T1a/T1mic N0, or if T1b and G1, no
adjuvant therapy is recommended, although endocrine therapy can be considered. If T1b N0 and grade 2 to 3, then
adjuvant endocrine therapy is recommended if ER positive.
If younger than age 60, adjuvant chemotherapy could be
considered. If T1b HER2 positive, endocrine therapy with or
without chemotherapy with or without trastuzumab could
be considered, but no further guidance is provided.
St. Gallen 2011. For luminal T1N0 ER-positive BC, antiestrogen therapies are recommended but not chemotherapy.
For the HER2-positive subtype, the panel of experts was
willing to extrapolate the chemotherapy/trastuzumab studies to T1bN0 BCs but would not recommend any adjuvant
therapy for T1a tumors. For the TN subtype, chemotherapy
is considered; no specific recommendation was made based
on tumor size. Trastuzumab with or without endocrine
therapy alone without chemotherapy was not considered to
be an acceptable adjuvant treatment for small HER2positive BC by 78% of the experts unless a contraindication
to chemotherapy existed.37 Chemotherapy would be more
strongly considered for large tumor size, involved nodes, or
bad biology (HER2 or TN, grade 3).
In My Opinion
ER-positive/HER2-negative BC. Adjuvant endocrine therapy should be considered for all, particularly for the T1b
group, subject to the patient tolerance of the endocrine
therapy. DFS/RFS is uniformly greater than 90% and typically approaches 97%. OncotypeRx is considered valid for
the T1bN0 ER-positive subset but has not been tested in the
T1mic or T1a groups. This test would be potentially useful in
T1b tumors that have adverse features such as higher grade,
low ER positivity, PR negativity, and possibly high Ki67
scores. Size and grade remain weak prognostic factors despite RS, and a new integrated RS is anticipated to account
for size.11 Chemotherapy could be considered for T1b disease
with high RS.
HER-positive/ER-positive BC. Treatment for this group is
highly controversial.38 HER2 amplification increases recurrence risk. Observation with or without endocrine therapy
alone has resulted in series with 5-year DFS of 85% to 92%,
particularly for the T1b group. Combination chemotherapy
plus trastuzumab with or without endocrine therapy has
resulted in extremely few recurrences; however, most patients are therefore over-treated. This treatment commits
the patient to myelosuppression, acute chemotherapy toxicities, increased cardiac morbidity, peripheral neuropathy
that can affect balance (particularly in the elderly), and
possible leukemia. Moreover, intravenous therapy is for a
year. Alternative approaches are beginning to be studied. A
17
ANTHONY D. ELIAS
phase II clinical trial, conducted by the Dana-Farber Cancer

Institute, collaborators, and Genentech, administered
single-agent paclitaxel with or without trastuzumab and
recently completed accrual but is not yet reported. The
Japanese RESPECT study is comparing trastuzumab with
or without chemotherapy in older patients with early-stage
HER2-positive BC.
There is no reported experience with anti-HER2 therapy
alone with or without endocrine therapy, although targeted
therapy alone is an attractive option that should be studied
prospectively. A legitimate concern would be some, but low,
cardiac risk. Another more concerning one would be the
observation that single-agent trastuzumab may not be as
effective in the absence of synergy with chemotherapy. For
example, sequential trastuzumab reduced relative relapse
risk by 0%, 14%, and 40% in adjuvant trials compared with
concurrent chemotherapy plus trastuzumab with risk reductions consistently above 50%.39-42 More recently, there is
strong data to suggest that dual anti-HER2 therapy in the
absence of chemotherapy can be highly effective.43,44 Thus,
combinations of trastuzumab/lapatinib, trastuzumab/pertuzumab, and T-DM1 with or without pertuzumab would
make sense. A phase III study in this population would
require an international effort.45
TN BC. At present, TN BC is defined by the absence of
predictive markers for all but chemotherapy, given its high
proliferative thrust. Thus, chemotherapy, particularly for
the T1b group older than age 50, could be considered. I
would use a regimen such as TC for four cycles, without
anthracyclines. Once predictive biomarkers for specific targeted treatments are established in the TN subgroup, the
discussion may read just like the previous paragraph, and
these patients (especially the T1b subset) should be included

in the definitive trials.
The prognosis of T1a and T1mic is also variable. Some
articles, but not all, find that size matters. Because T1mic
and T1a tumors frequently represent small areas of invasion
within a sea of DCIS, careful histologic sampling is required
to avoid missing areas of greater invasion. The very young,
especially younger than 35, could be considered for adjuvant
chemotherapy.
Local-regional control is clearly affected by tumor biology.
IBR is higher in the HER2-positive and TN cancers, thus
local-regional treatment recommendations would be the
same as for T1cN0 tumors.3,6 Sentinel node sampling is still
indicated in these small tumors. The incidence of axillary
node involvement, while lower in smaller primary tumors, is
still 5% to 15%. In the future, once the RxSPONDER trial is
reportedwhich is testing the utility of chemotherapy in
node-positive, low to intermediate RS ER-positive BCs
perhaps nodal stage will become less important. Currently,
nodal involvement still affects absolute risk of distant metastasis, and much of the decision making has to do with the
absolute benefit of given interventions relative to their risks.
Conclusion
Small BCs are increasing in frequency. Although their

risk of disseminated disease is low following local-regional
treatment, certain subtypes are at greater risk. These patients have generally not been included in prospective clinical trials; a fact reflected by the heterogeneous approaches
to systemic adjuvant therapy in the community of clinical
oncologists. Optimal management remains controversial
and will not be clarified until dedicated clinical trials are
conducted.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Anthony D. Elias*
REFERENCES
1. Kennedy T, Stewart AK, Bilimoria KY, et al. Treatment trends and
factors associated with survival in T1aN0 and T1bN0 breast cancer patients.
Ann Surg Oncol. 2007;14:2918-2927.
2. Livi L, Meattini I, Saieva C, et al. Prognostic value of positive human
epidermal growth factor receptor 2 status and negative hormone status in
patients with T1a/T1b, lymph node-negative breast cancer. Cancer. Epub
2011 October 25.
3. Wong FY, Yip CSP, Chua ET. Implications of HER2 amplification in
small, node-negative breast cancers: Do Asians differ? World J Surg. Epub
2011 November 22.
4. Kwon JH, Kim YJ, Lee K-W, et al. Triple negativity and young age as
prognostic factors in lymph node-negative invasive ductal carcinoma of 1 cm
or less. BMC Cancer. 2010;10:557.
5. Tanaka K, Kawaguchi H, Nakamura Y, et al. Effect of HER2 status on
risk of recurrence in women with small, node-negative breast tumours. Br J
Surg. 2011;98:1561-1565.
6. McArthur HL, Mahoney KM, Morris PG, et al. Adjuvant trastuzumab
with chemotherapy is effective in women with small, node-negative, HER2positive breast cancer. Cancer. 2011;117:5461-5468.
7. Horio A, Fujita T, Hayashi H, et al. High recurrence risk and use of
adjuvant trastuzumab in patients with small, HER2-positive, node-negative
breast cancers. Int J Clin Oncol. Epub 2011 June 18.
8. Cancello G, Maisonneuve P, Rotmensz N, et al. Prognosis in women with
18
small (T1mic, T1a, T1b) node-negative operable breast cancer by immunohistochemically selected subtypes. Breast Cancer Res Treat. 2011;127:713-720.
9. Theriault RL, Litton JK, Mittendorf EA, et al. Age and survival estimates in patients who have node-negative T1ab breast cancer by breast
cancer subtype. Clin Breast Cancer. 2011;11:325-331.
10. Rouanet P, Roger P, Daures JP, et al. HER2 expression is the major
risk factor for recurrence in pT1a-b, N0 breast cancer: Clinical implications
from a French regional population-based study of 703 patients. Proc SABCS.
2011;P2:12-16.
11. Habel LA, Shak S, Jacobs MK, et al. A population-based study of tumor
gene expression and risk of breast cancer death among lymph node-negative
patients. Breast Cancer Res. Epub 2006 May 31.
12. Rosen PP, Groshen S, Saigo PE, et al. A long-term follow-up study of
survival in stage I T1N0M0) and stage II (T1N1M0) breast carcinoma. J Clin
Oncol. 1989;7:355-366.
13. Moon TE, Jones SE, Bonadonna G, et al. Development and use of a
natural history data base of breast cancer studies. Am J Clin Oncol.
1987;10:396-403.
14. Stierer M, Rosen HR, Weber R, et al. Long term analysis of factors
influencing the outcome in carcinoma of the breast smaller than one centimeter. Surg Gynecol Obstet. 1992;175:151-160.
15. Hanrahan EO, Valero V, Gonzalez-Angulo AM, et al. Prognosis and
management of patients with node-negative invasive breast carcinoma that is

1 cm or smaller in size (stage 1; T1a,bN0M0): a review of the literature. J Clin
Oncol. 2006;24:2133-2122.
16. Fisher B, Dignam J, Tan-Chiu E, et al. Prognosis and treatment of
patients with breast tumors of one centimeter or less and negative lymph
nodes. J Natl Cancer Inst. 2001;93:112-120.
17. Wood WC, Anderson M, Lyles RH, et al. Can we select which patients
with small breast cancers should receive adjuvant chemotherapy? Ann Surg.
2002;235:859-862.
18. Joensuu H, Isola J, Lundin M, et al. Amplification of erbB2 and erbB2
expression are superior to estrogen receptor status as risk factors for distant
recurrence in pT1N0M0 breast cancer: A nationwide population-based study.
Clin Cancer Res. 2003;9:923-930.
19. Fisher B, Bryant J, Dignam JJ, et al. Tamoxifen, radiation therapy, or
both for prevention of ipsilateral breast tumor recurrence after lumpectomy in
women with invasive breast cancers of one centimeter or less. J Clin Oncol.
2002;20:4141-4149.
20. Fisher ER, Costantino JP, Leon ME, et al. Pathobiology of small
invasive breast cancers without metastases (T1a/b, N0, M0). NSABP B-21.
Cancer. 2007;110:1929-1936.
21. Chia S, Norris B, Speers C, et al. Human epidermal growth factor
receptor 2 overexpression as a prognositic factor in a large tissue microarray
series of node-negative breast cancers. J Clin Oncol. 2008;26:5697-5704.
22. Gonzalez-Angulo AM, Litton JK, Broglio KR, et al. High risk of
recurrence for patients with breast cancer who have human epidermal growth
factor receptor 2-positive, node-negative tumors 1 cm or smaller. J Clin Oncol.
2009;27:5700-5706.
23. Curigliano G, Viale G, Bagnardi V, et al. Clinical relevance of Her2
overexpression/amplification in patients with small tumor size and nodenegative breast cancer. J Clin Oncol. 2009;27:5693-5699.
24. Meattini I, Livi L, Saieva C, et al. Prognostic value of HER2 positivity
and negative hormonal status in patients with small tumor (1cm) and
node-negative breast cancer. Proc SABCS. 2011;P2-12-14.
25. Colleoni M, Rotmensz N, Peruzzotti G, et al. Minimal and small size
invasive breast cancer with no axillary lymph node involvement: The need for
tailored adjuvant therapies. Ann Oncol. 2004;15:1633-1639.
26. Hanrahan EO, Gonzalez-Angulo AM, Giordano SH, et al. Overall
survival and cause-specific mortality of patients with stage T1a,bN0M0
breast carcinoma. J Clin Oncol. 2007;25:4952-4960.
27. Park YH, Kim ST, Cho EY, et al. A risk stratification by hormonal
receptors (ER, PgR) and HER2 status in small (1cm) invasive breast cancer:
Who might be possible candidates for adjuvant treatment. Breast Cancer Res
Treat. 2010;119:653-661.
28. Lai HW, Kuo SJ, Chen LS, et al. Prognostic significance of triple
negative breast cancer at tumor size 1 cm and smaller. Eur J Surg Oncol.
2011;37:18-24.
29. Fehrenbacher L, Shiraz P, Sattavat M, et al. T1abN0M0 HER2
invasive breast cancer recurrence: Population based cohort of 17,000 consecutive breast cancers 2000-2006 at Kaiser Permanente Northern California, KPNC. J Clin Oncol. 2011;29: (suppl; abstr 551).
30. Amar S, McCullough AE, Tan W, et al. Prognosis and outcome of small
(1cm), node-negative breast cancer on the basis of hormonal and Her2

status. Oncologist. 2010;15:1043-1049.
31. Peron J, Frenel JS, Vano Y, et al. Systemic adjuvant treatment of T1a
and T1b N0M0 HER2 breast carcinomas; an AERIO/UNICANCER study.
Proc SABCS. 2011;P2-18-03.
32. Shao T, Boolbol SK, Boachi-Adjei K, Klein P. Clinical significance of
HER2 and triple-negative status in patients with tumor size 1 cm and
node negative breast cancer. Proc SABCS. 2011;4-09-03.
33. Cserni G, Bianchi S, Vezzosi V, et al. Sentinel lymph node biopsy in
staging small (up to 15 mm) breast carcinomas. Results from a European
multi-institutional study. Path Oncol Res. 2007;13:5-14.
34. Rivadeneira DE, Simmons RM, Christos PJ, et al. Predictive factors
associated with axillary lymph node metastases in T1a and T1b breast
carcinomas: Analysis in more than 900 patients. J Am Coll Surg. 2000;191:
1-8.
35. Khair TA, Boolbol SK, Boachi-Adjei K, Klein P. Factors affecting the
development of axillary lymph node metastases in T1a-T1b breast cancers.
Proc SABCS. 2011;P4-09-16.
36. Lyons TR, OBrien J, Borges VF, et al. Postpartum mammary gland
involution drives progression of ductal carcinoma in situ through collagen and
COX-2. Nat Med. 2011;17:1109-1115.
37. Goldhirsch A, Wood WC, Coates AS, et al. Strategies for subtypes
dealing with the diversity of breast cancer: Highlights of the St. Gallen
International Expert Consensus on the primary therapy of early breast
cancer 2011. Ann Oncol. 2011;22:1736-1747.
38. Kelly CM, Pritchard KI, Trudeau M, et al. Coping with uncertainty:
T1a,bN0M0 HER2-positive breast cancer, do we have a treatment threshold?
Ann Oncol. 2011;22:2387-2393.
39. Spielmann M, Roche H, Delozier T, et al. Trastuzumab for patients
with axillary-node positive breast cancer: Results of the FNCLCC-PACS04
trial. J Clin Oncol. 2009;27:6129-6134.
40. Perez EA, Romond EH, Suman VJ. Updated results of the combined
analysis of NCCTG N9831 and NSABP B-31 adjuvant chemotherapy with/
without trastuzumab in patients with HER2-positive breast cancer. J Clin
Oncol. 2007;25 (suppl 18s; abstr 512).
2005;353:1659-1672.
42. Untch M, Gelber RD, Jackisch C, et al. Estimating the magnitude of
trastuzumab effects within patient subgroups in the HERA trial. Ann Oncol.
2008;19:1090-1096.
43. Blackwell KL, Burstein HJ, Stroniolo AM, et al. Randomized study of
Lapatinib alone or in combination with trastuzumab in women with ErbB2positive, trastuzumab-refractory metastatic breast cancer. J Clin Oncol.
2010;28:1124-1130.
44. Baselga J, Gelmon KA, Verma S, et al. Phase II trial of pertuzumab and
trastuzumab in patients with human epidermal growth factor receptor
2-positive metastatic breast cancer that progressed during prior trastuzumab
therapy. J Clin Oncol. 2010;28:1138-1144.
45. Constantinidou A, Smith I. Is there a case for anti-HER2 therapy
without chemotherapy in early breast cancer? Breast. 2011;20:S158-S161.
19
Controversies in Adjuvant Endocrine Therapy

for Breast Cancer
By Stephen R. Johnston, PhD, MA
Overview: Adjuvant endocrine therapy for early-stage breast

cancer has had the single biggest impact on improving survival from the diseasewith tamoxifen alone contributing to
saving many thousands of lives. In postmenopausal women,
additional progress has been made by the incorporation of
aromatase inhibitors into the treatment of early-stage, estrogen receptor (ER)positive breast cancer, as several large
well-conducted trials have established either up-front or
switch strategies that are now widely used. To date, both
have been shown to be beneficial when compared with tamoxifen alone, although controversy exists as to which approach
is superior. Increasingly, extended adjuvant therapy is being
considered, as longer may be better for some women who
have an ongoing risk of recurrence beyond 5 years. However,
controversy remains as to how long adjuvant endocrine therapy should be given for; in clinical practice, clinicians balance
T IS well established that for patients with early-stage

ER-positive breast cancer, adjuvant endocrine therapy
given for 5 years after primary surgery delays local and
distant relapse and prolongs overall survival.1 In addition,
adjuvant therapy substantially reduces the incidence of
contralateral breast cancer in patients with primary breast
cancer. Endocrine responsiveness, for the most part, is
dependent on the presence of a functional estrogen receptor
(ER), a protein that can be detected in approximately 70% to
80% of primary breast cancers. Hormonal manipulation is
either achieved at a cellular level by using antiestrogens
such as tamoxifen to compete for ER in the breast tumor,
or by lowering systemic estrogen levels in premenopausal
women with the use of LHRH agonists, and in postmenopausal women by the use of aromatase inhibitors that block
estrogen biosynthesis in nonovarian tissues. All of these
approaches have been extensively investigated within the
context of large-scale international trials of adjuvant endocrine therapy over the past two decades. However, several
key issues and unanswered questions remain, namely the
best choice of endocrine agents and the optimal strategies
in the postmenopausal setting, the appropriate duration
of endocrine therapy, and the true role of ovarian suppression in premenopausal women. In addition, the molecular
classification of breast cancer has now allowed us to better
estimate endocrine responsiveness in ER-positive breast
cancer, introducing another tool in addition to established
guidelines for decision making in the adjuvant setting. In
particular, this has allowed us to address the issue of
whether endocrine therapy alone is enough for many patients.
The Benefit of Tamoxifen
In early-stage breast cancer, tamoxifen has been the gold

standard of adjuvant endocrine therapy for both premenopausal and postmenopausal breast cancer for more than
three decades. In an overview of the effects of chemotherapy
and hormone therapy for early-stage breast cancer by
the Early Breast Cancer Trialists Collaborative Group
(EBCTG), a 50% reduction in the risk of relapse and a 31%
reduction in the annual breast cancer death rate was re-
20
the level of risk for individual patients versus any ongoing

toxicity concerns. For premenopausal women, with ERpositive breast cancer, tamoxifen remains the gold standard
with uncertainty in the added overall benefit of ovarian suppression. Important clinical trials have recently been completed that may help answers this question, including whether
complete estrogen deprivation using a luteinizing hormone
releasing hormone (LHRH) agonist plus aromatase inhibitors
(AIs) is of added benefit. In recent years, molecular profiling
of ER-positive breast cancer has started to distinguish those
women with a low risk of recurrence on endocrine therapy who
may not need chemotherapy. Thus, with more therapy options
and greater tumour stratification, modern, adjuvant endocrine
therapy is becoming increasingly personalised to suit each
individual patients risk.
ported for 5 years of adjuvant tamoxifen at 15 years of

follow-up.1 Although the risk of distant recurrence is greatest during the first decade, it can still continue through the
second decade,2 raising the question about the magnitude
of carry-over effect from initial therapy, or the need for
extended adjuvant therapy beyond 5 years in those women
at greater risk. In 2011, the EBCTG updated their metaanalysis of long-term outcomes in 21,457 women with earlystage breast cancer from 20 randomized trials of 5 years of
tamoxifen compared with observation or placebo.3 They
showed that 5 years of tamoxifen reduced the recurrence
rate substantially in years 0 4 during therapy (rate ratio
[RR] 0.53), and also in years 59 (RR 0.68) and throughout
the first 15 years (RR 0.70, p 0.00001). Most importantly,
the relapse curves do not converge after year 10, with a
continued annual absolute gain from the annual reduction
in breast cancer mortality through to year 15. Furthermore,
the benefit only occurs in those with ER-positive tumors,
being maximal in those with rich expression of the receptor.
As such, 5 years of tamoxifen probably cures many patients
rather than simply delays an inevitable recurrence.
These mature data from the meta-analysis regarding a
well-established treatment such as tamoxifen give both
clinicians and patients confidence in the beneficial effects
of endocrine therapy on improving overall survival from
breast cancerso how long do patients need to take tamoxifen? Studies of duration such as NSABP-14 have compared
5 to 10 years of tamoxifen and shown no advantage beyond
5 years, indeed perhaps a slight disadvantage in terms of
disease-free survival.4 However, this study only examined
node-negative patients, and in two further trials (aTTOm
and ATLAS), preliminary results have suggested a small
From the Department of Medicine, Royal Marsden NHS Foundation Trust, Chelsea,
London, United Kingdom.
Authors disclosure of potential conflicts of interest are found at the end of this article.
Address reprint requests to Stephen R.D. Johnston, MA, PhD, Department of Medicine,
Royal Marsden NHS Foundation Trust, Fulham Road, Chelsea, London, SW3 6JJ, UK;
email: stephen.johnston@rmh.nhs.uk.
1092-9118/10/1-10
ADJUVANT ENDOCRINE THERAPY
reduction in breast cancer recurrence for those randomized

to continue tamoxifen, although no significant difference
was observed for breast cancer specific or overall mortality.5,6 Further follow-up of these studies is required to
reliably assess the longer term effects on recurrence and
overall effects, if any, on mortality.
Although tamoxifen remains the most appropriate sole
treatment option for many premenopausal and perimenopausal women, the issue for postmenopausal women remains whether there are any better options than tamoxifen,
especially given that some women will still relapse despite
therapy, and that for others, tamoxifen has either unacceptable short-term vasomotor toxicities or increased long-term
risks such as venous thromboses and endometrial cancer.
The incorporation of AIs into the treatment of postmenopausal early-stage ER-positive breast cancer has now led
to further improvements in outcomes over tamoxifen. The
key issue is how AIs should be incorporated as adjuvant
therapy in early-stage breast cancer, and whether a role for
tamoxifen still remains in postmenopausal breast cancer.
AIs: How Should We Use Them?
In contrast to tamoxifen, which antagonizes estrogen at

the ER, the oral AIs such as anastrozole, letrozole and
exemestane all reduce serum estrogen levels in postmenopausal women by preventing the conversion of adrenal
androgens (androstenedione and testosterone) into estradiol
(E1) and estrone (E2) by the cytochrome P450 enzyme
KEY POINTS
Aromatase inhibitors are recommended to be part of

adjuvant endocrine therapy for most postmenopausal
women with estrogen receptor (ER)-positive breast
cancer, either as up-front therapy or as a switch
strategy following initial tamoxifen therapy. Both
approaches are equally effective, and both are superior to tamoxifen alone for 5 years.
The optimal duration for aromatase inhibitors given
as adjuvant therapy is 5 years based on current
safety and efficacy data.
Extended adjuvant therapy with aromatase inhibitors after 5 years of tamoxifen is an additional accepted strategy for reducing ongoing risk of
recurrence, especially for those at greater risk.
The added benefit of ovarian suppression in premenopausal women with ER-positive early-stage breast
cancer over and above tamoxifen remains unknown.
In women younger than 40 years, added benefit may
exist, but must be weighed up against both short and
long-term tolerability.
Molecular profiling of ER-positive breast cancer can
identify those subtypes at low risk of recurrence for
whom the addition of adjuvant chemotherapy over
and above endocrine therapy is not indicated. To date
biomarkers, although clearly of prognostic value, have
not been shown to identify those more likely to benefit
from aromatase inhibitors rather than tamoxifen.
Table 1. Comparative Efficacy of Up-Front 5 Years Aromatase

Inhibitors Versus 5 Years Tamoxifen in Early Breast Cancer
Study (reference)
ATAC11
BIG-19816
Number of patients
Median follow up
Disease-free survival
Five-year disease-free survival difference
Time to distant recurrence
Overall survival
6,241
100 mo
HR 0.90 p .025
2.8%
HR 0.86 p .022
HR 1.00 p .99
4,922
76 mo
HR 0.88 p .03
2.9%
HR 0.85 p .05
HR 0.87 p .08
Abbreviations: HR, hazard ratio.

p .05 significant.
aromatase.7 Although estrogens are primarily synthesized

in the ovary in premenopausal women under the control of
stimulatory effects of luteinizing hormone (LH) and follicle
stimulating hormone (FSH), following menopause, mean
plasma E2 levels fall from about 400 600 pmol/L to around
2550 pmol/L. These residual estrogens come solely from
peripheral aromatase conversion, particularly in muscle and
subcutaneous fat.
The establishment of the efficacy and tolerability of AIs in
advanced breast cancer encouraged the development of a
number of trials examining their use in the adjuvant setting.7 In general, these therapeutic studies have either
compared primary up-front AI therapy for 5 years with 5
years of tamoxifen, or a so-called switch strategy of initial
tamoxifen for 23 years followed either by an AI for 23
years of continued tamoxifen through to year 5. The absolute
benefits of these approaches have been assessed in a recent
meta-analysis of all the adjuvant trials of AIs compared with
the previous standard of care, namely, 5 years of tamoxifen.8
The key message is that AIs produce significantly lower
recurrence rates compared with tamoxifen, either as initial
monotherapy or following 23 years of tamoxifen, although
the true effect on long-term survival is less clear at this
stage. The recent 2010 American Society of Clinical Oncology (ASCO) guidelines on adjuvant endocrine therapy for
women with ER-positive breast cancer recommend incorporating AI therapy at some point during adjuvant treatment,9
while recognizing that the optimal timing and duration of
therapy remain unresolved questions. So what are the
benefits, and what has been learned from all the trials of
adjuvant AIs to date?
Up-Front AIs
Two large studies have assessed the efficacy of AIs compared with tamoxifen as up-front adjuvant endocrine
therapy in postmenopausal women with early-stage breast
cancer (Table 1).10-14 The Arimidex, Tamoxifen, Alone or
in Combination (ATAC) trial was the first large study to
investigate the role of AIs as adjuvant therapy for earlystage breast cancer. Over 4 years, 9,366 postmenopausal
women from 21 countries were enrolled. The hypotheses
tested were that anastrozole was noninferior or superior to
tamoxifen, or that the combination of anastrozole and tamoxifen was superior to tamoxifen alone. The combination
treatment was discontinued following the initial analysis
because it showed no efficacy or tolerability benefits over
tamoxifen alone.10 Following a median follow-up of 100
months,11 anastrozole compared with tamoxifen was associated with a significantly improved disease-free survival
(DFS) in hormone-receptor-positive patients (hazard ratio
[HR] 0.85, p 0.003). In the most recent 10-year analysis,
21
STEPHEN R. JOHNSTON
the absolute differences in time to recurrence (TTR) in

hormone-receptor-positive patients increased over time,
being 2.7% at 5 years and 4.3% at 10 years.12 In addition,
recurrence rates remained significantly lower on anastrozole
compared with tamoxifen after treatment completion (HR
0.81, p 0.03), although the carry-over effect was smaller
after 8 years. However, these differences in preventing/
delaying disease recurrence did not result in a difference in
overall survival between the two treatments (HR 0.95, 95%
CI 0.84 1.06).12 Differences in toxicity profile demonstrated
a higher incidence of thromboembolic and cardiovascular
events with tamoxifen, and more musculoskeletal events
and fractures with anastrozole.13 A bone sub-study of the
main trial confirmed that AI therapy was associated with
accelerated bone loss during the 5-year treatment period,
although no patients with normal bone mineral density at
baseline became osteoporotic at 5 years.14
The BIG 198 trial assessed both the efficacy of up-front
AI therapy with letrozole for 5 years compared with tamoxifen, as well as switching strategies. It recruited 8,028
women who were randomized to one of four treatment arms.
In 2005, the first analysis at a median follow-up of 25.8
months reported that 5 years of letrozole demonstrated a
significant improvement in DFS over tamoxifen (HR 0.81,
p 0.003) and distant disease-free survival (DDFS) (HR
0.73 p 0.001).15 These results led to the unblinding of
the tamoxifen alone arm, and 25.2% of patients selectively
crossed over to letrozole, which has complicated subsequent
intention to treat (ITT) analyses of the monotherapy arms.
The updated report at a median follow-up of 76 months16
included both an ITT analysis and a censored analysis at
the time of cross-over, and still demonstrated a statistically
significant improvement in DFS and DDFS in favor of
letrozole over tamoxifen, with a nonsignificant improvement
in overall survival (HR 0.87, 95% CI 0.751.02, p 0.08). In
terms of toxicity, endometrial cancer, vaginal bleeding, and
thrombo-embolism were more common with tamoxifen,
while musculoskeletal events and hypercholesterolemia
were higher with letrozole.
The meta-analysis of both of these studies showed an
absolute 5-year reduction in recurrence of 2.9% (9.6% for AI
vs 12.6% for tamoxifen, p0.00001), with an absolute 3.9%
reduction at 8 years.8 This was associated with only a 1.1%
reduction in breast cancer mortality at 5 years, which was
nonsignificant.
Switching from Tamoxifen to AIs Compared with Tamoxifen Alone
In terms of a switching strategy, several trials have

evaluated a switch to an aromatase inhibitor after 23 of
tamoxifen compared with 5 years tamoxifen alone, in an
attempt to pre-empt the development of resistance to tamoxifen, and to minimize long-term side effects of either therapy
if given alone for 5 years (Table 2).17-21 The largest of these
studies was the Intergroup Exemestane Study (IES), which
compared switching to exemestane after 23 years of tamoxifen with tamoxifen for 5 years. This demonstrated not only
a statistically significant improvement in disease-free survival (HR 0.68, p 0.001),17 but also in overall survival in
an updated analysis of those with ER-positive disease (HR
0.86, p 0.04).18 These findings were also confirmed in a
meta-analysis of three separate trials with the tamoxifenanastrozole switch compared with tamoxifen alone (the
Austrian Breast and Colorectal Cancer Study Group
22
Table 2. Comparative Efficacy of 23 Years Tamoxifen Followed

by a Switch to 23 Years Aromatase Inhibitor Versus 5 Years
Tamoxifen Alone
Study (reference)
IES18
ARNO 9519
ITA20
ABCSG 819
Number of patients
Median follow up
4,724
55.7 mo
HR 0.76
p .0001
HR 0.83*
p .05
979
30.1 mo
HR 0.66
p .49
HR 0.53
p .045
448
64 mo
HR 0.56
p .01
HR 0.56
p .1
3,714
72 mo
HR 0.79
p .038
HR 0.77
p .025
Overall survival
Abbreviations: HR, hazard ratio.

* In a subset of estrogen-receptor positive patients only p .05 significant.
(ABCSG 8), Arimidex-Nolvadex (ARNO 95) and Italian

Tamoxifen Anastrozole (ITA) studies19,20). The metaanalysis included more than 4,000 patients with a mean
follow-up of 30 months, which is shorter follow-up than in
the IES study. Despite this, the combined analysis still
showed that switching to anastrozole resulted in a significant improvement in disease-free survival (HR 0.59, p
0.0001), and was also associated with a statistically significant reduction in breast cancer mortality and death from
any cause (HR 0.71, p 0.037).21
These individual studies certainly all seem to suggest
significantly better outcomes in terms of disease-free and
overall survival by a switching strategy when compared
with tamoxfen alonein addition to less long-term toxicities
with tamoxifen, there appears to be a substantial gain in
efficacy, presumably by preventing relapses that were destined to occur during continued tamoxifen by utilizing a
noncross-resistant therapy. However, it has been questioned
by some whether early relapses on tamoxifen could be
prevented by an up-front AI strategy, and as such it has
been unclear which approach (tamoxifen switch to an AI, or
AI upfront) would yield better efficacy outcomes and which
would be better tolerated overall. Two large trials (BIG 198
and TEAM) have now given us direct data on these comparisons.
Switching Tamoxifen to AIs Compared with AI Upfront
The Breast International Group (BIG) 198 study was the

first clinical trial to report a direct comparison between the
use of an upfront AI and a switching strategy.16 At a median
follow-up of 76 months, there was no significant difference
in disease-free recurrence, overall survival or time to distant
recurrence between the switching (tamoxifen to letrozole)
and the letrozole monotherapy arms. However, there were
fewer early relapses among women who were assigned to
letrozole alone compared with the switch of tamoxifen followed by letrozolethis trend was greatest in node-positive
patients with a 1.4% difference in breast cancer recurrence
in node-negative patients (3.5% vs. 4.9%), and a 2.3% absolute difference in node-positive patients (12.4% vs. 14.7%) at
5 years. Interestingly, there was no significant difference
in breast cancer recurrence between the letrozole followed
by a reverse-switch to tamoxifen and the letrozole monotherapy arm. The clinical significance of this is that patients
who commence an AI but experience side effects or toxicity
may be safely switched over to tamoxifen to complete their
adjuvant endocrine therapy, without compromising on the
efficacy provided by the AIs.
The second study to provide a direct comparison was the
Tamoxifen Exemestane Adjuvant Multicenter (TEAM) trial,
which randomized 9,779 postmenopausal women with ERpositive early-stage breast cancer to either exemestane for
5 years, or tamoxifen for 2.53 years followed by exemestane.22 At median follow-up of 5.1 years, there was no
difference in either disease-free or overall survival, a result
very similar to that observed in the BIG 198 study.
Who Benefits from Which StrategyUp-Front or Switch?
So what are we to conclude from these large, wellconducted studies as to which strategy is the best? From the
meta-analysis of all these studies the addition of an AI in
either strategy (up-front or switch) clearly improves diseasefree survival compared with 5 years of tamoxifen, although
delaying relapse does not appear to translate into a substantial survival advantage, apart a modest benefit that was
seen in the analysis of all the switching studies, which
yielded an absolute gain in overall survival of 1.7% at 8
years.8 Biomarker studies in many of these trials have
attempted to see whether conventional indicators of hormone responsiveness, such as absolute ER and PgR
levels,23-25 coexpression of HER2,23 proliferation as assessed
by Ki-67 labeling index,26 or the 21-gene recurrence score,27
can identify those who benefit more from an AI than tamoxifen. Although confirming the prognostic role of all these
different biomarkers in each individual study, none have
been shown to identify patients with a differential response
between tamoxifen and AI therapy. At the present time, the
two direct comparisons between up-front AI or switch strategy (BIG 198 and TEAM) have shown no significant differences between these approaches; although in the letrozole
study, for higher-risk node-positive patients, fewer recurrences were seen with AI up-front compared with switching.16 As such, tumor burden in addition to other inherent
risk factors related to the biology of the disease (higher
grade or proliferation score, lower levels of ER or PgR,
coexpression of HER2) might be reasonable factors to use in
clinical practice to favor an up-front use of AIs as opposed to
a switch strategy.
In addition, patient-related factors and toxicity concerns
should always be considered when making a decision with a
patient between either up-front AI use or a switch strategy.
In general, the third generation AIs are well tolerated, and
in the clinical trial setting, the different side effect profiles of
tamoxifen and aromatase inhibitors do not appear to affect
patients quality of life.28 Assessing an individuals risk or
history of venous thrombosis may influence whether tamoxifen is indicated, and assessing joint arthropathies or the
risk of osteoporosis may determine the optimal timing of
starting an AI. Current ASCO guidelines recommend that
postmenopausal women who receive an AI should have their
bone mineral density evaluated, with calcium and vitamin D
supplementation or bisphosphonate use dependent on the
result.29 As such, clinicians and patients have a choice on
when and how to use an AI, dependent on the level of risk
for the cancer and the general health issues for each individual womanwith often only marginal differences in efficacy outcomes (in particular survival), consideration of all
these issues is important.
Which AI Is Best?
The only reported comparison to date is the MA.17 study,

where 5 years of adjuvant exemestane yielded similar results to 5 years of adjuvant anastrozole.30 A trial compar-
Table 3. Comparative Efficacy of Extended Adjuvant Therapy of

5 Years Tamoxifen Followed by 35 Years Aromatase Inhibitor
Versus 5 Years Tamoxifen Alone
Study (reference)
MA-1732
NSABP B-3336
ABCSG-6a35
Number of patients
Median follow up
5,170
64 mo
HR 0.68
p .0001
HR 0.98
p .853
1,562
30 mo
HR 0.68
p .07
NR
852
62 mo
HR 0.62
p .031
HR 0.89
p .57
Overall survival
Abbreviations: HR, hazard ratio; NR, not recorded.
ing the two nonsteroidal aromatase inhibitors letrozole and

anastrozole (FACE) has completed recruitment, and results
are awaited. Tolerability profiles between the different AIs
are very similar, so, to date, there is no evidence to suggest
that one agent is substantially better than another.
Adjuvant Endocrine TherapyIs There an
Optimal Duration?
ER-positive breast cancer has a chronic relapsing nature,

with the risk of recurrence continuing indefinitely. In the
tamoxifen overview, approximately half of all recurrences
occurred between 5 and 15 years after surgery despite 5
years of adjuvant tamoxifen treatment.1,3 There is, therefore, a clear rationale for considering extended adjuvant
endocrine therapy beyond 5 years in an attempt to reduce
long-term risk of recurrence further. The MA.27 study31 was
a double-blind, placebo-controlled trial designed to test
whether 5 years of letrozole therapy in more than 5,000
postmenopausal women who had completed 5 years of adjuvant tamoxifen could lead to a further improvement in
disease-free survival between years 5 and 10 from diagnosis.
The trial demonstrated a significant improvement in
disease-free survival in all patients who continued onto
letrozole after completion of 5 years of tamoxifen, with a
4-year disease-free survival rate (i.e., year 9 since initial
diagnosis) of 93% compared with 87%.31 Furthermore, after
a median follow-up of 30 months among higher risk lymph
node-positive patients, overall survival was statistically significantly improved with letrozole (HR 0.61, 95% CI,
0.38 0.98; p 0.04).32 Subsequent analyses of this trial
have suggested that this benefit was greatest in those with
ER-positive PgR-positive tumors,33 and that benefit may
also have occurred if the switch to extended adjuvant letrozole occurred late following a period of a few years since
completion of tamoxifen.34 Two additional, smaller separate
studies with either anastrozole35 or exemestane36 as extended adjuvant therapy after completion of 5 years of
tamoxifen have shown similar quantitative results in terms
of reducing risk of recurrence, with hazard ratios of 0.62 and
0.68, respectively (Table 3).
These data confirm that risk of recurrence in hormonesensitive breast cancer does indeed persist on an ongoing
basis well beyond completion of 5 years of adjuvant tamoxifen. As such, careful consideration should be given to
offering longer durations of adjuvant therapy with AIs for
those deemed to be at greatest risk. Although all of these
studies looked at treatment following 5 years of tamoxifen,
increasingly, clinical practice is to either use AI therapy
up-front in higher risk node-positive patients, or a switch
approach of AIs following an initial 23 years of tamoxifen.
23
STEPHEN R. JOHNSTON
So how do we translate the data from the extended adjuvant

therapy studies into these settings of how AIs are being
used today? Can we safely use AIs beyond a total duration
of 5 years without, for example, compromising bone health?
The answer at present is unknown, although all the safety
data relating to the optimal duration of AIs indicate that
therapy should be given for a maximum of 5 years. The
ASCO guidelines recommend that AI therapy should not
exceed 5 years outside the setting of clinical trials and that
in the sequential setting, patients should receive an AI after
23 years of tamoxifen for a total of 5 years of adjuvant
endocrine therapy. The guidelines recognize that this may
yield an unfamiliar pattern of different durations of adjuvant treatment based on the treatment strategy used, and
that neither 5 years up-front AI or sequential switch strategies (23 years tamoxifen followed by 23 years of an AI)
have been compared against the longer overall duration of
the extended adjuvant therapy regimens where treatment
duration was up to 8 10 years. Two trials (MA.17R and
NSABP B-42) are addressing whether longer durations of AI
therapy improve outcomes without compromising safety, but
results are not yet available.
permanent amenorrhea. Optimum duration of the use of

reversible ovarian suppression is unknown, although studies in general have utilized 23 years of LHRH agonist with
5 years of tamoxifen.
Although the standard of care for ER-positive premenopausal breast cancer remains tamoxifen alone for 5 years,
prospective trials to address the added role (if any) of
ovarian suppression compared with tamoxifen alone have
been undertaken. The SOFT (Suppression of Ovarian Function Trial) randomized trial will assess the role of ovarian
suppression/ablation in combination with the aromatase
inhibitor exemestane, compared with either ovarian suppression plus tamoxifen or tamoxifen alone. More than 3,000
women were randomized into this study, which completed
accrual in January 2011. The TEXT trial (Tamoxifen and
EXemestane Trial) assesses an LHRH agonist with the
addition of either tamoxifen or exemestane for 5 years
(chemotherapy is optional), and accrual of more than 2,600
women was completed in March 2011. It is hoped that both
these trials will provide important additional information
about the optimal endocrine therapy for premenopausal
women with ER-positive early-stage breast cancer.
Ovarian SuppressionIs It Necessary for

Premenopausal Women?
Adjuvant Endocrine TherapyIs It Enough on Its Own?
LHRH agonists, which initially stimulate and then exhaust the LHRH receptors in the pituitary, cause reversible
suppression of ovarian function, and are currently used as
an alternative to ovarian ablation for treatment of advanced
breast cancer in premenopausal women. The Early Breast
Cancer Trialists Collaborative Group (EBCTG)1 reviewed
trials involving almost 8,000 women with ER-positive or
ER-unknown early-stage breast cancer who were randomized to ovarian ablation by surgery or irradiation or ovarian
suppression with an LHRH agonist. Overall, there was a
definite beneficial effect of ovarian ablation/suppression
both on recurrence and breast cancer mortality. However,
the effects of ovarian treatment appear smaller in the trials
where both groups got chemotherapy than in the trials
where neither did, probably because chemotherapy-induced
amenorrhea attenuated any additional effect of ovarian
suppression. This was best demonstrated in the Intergroup
0101 trial, where in premenopausal women with ER-positive
node-positive early-stage breast cancer, the addition of the
LHRH agonist goserelin to CAF chemotherapy appeared
to have a greater effect in improving disease-free survival
in women younger than 40 compared with those older than
40 years.37 Likewise, the IBCSG trial VIII randomized
premenopausal women with node-negative ER-positive
early-stage breast cancer to either adjuvant CMF chemotherapy, goserelin for 2 years, or CMF followed by goserelin
for 18 months.38 The addition of goserelin overall only had
a minimal effect on 5-year disease-free survival, and again
the benefit was maximal in women younger than 40 years
with a hazard ratio of 0.34.
A more recent meta-analysis looked at only trials where
ER status was known and that used LHRH agonists as the
method of ovarian suppression.39 The primary endpoints
were any recurrence or death after recurrence, with a
median follow-up of 6.8 years. In particular, a benefit was
observed when LHRH agonists were used after chemotherapy (either alone or with tamoxifen) in women younger than
40 years in whom chemotherapy is less likely to induce
24
Perhaps the hottest topic for current debate in the management of ER-positive early-stage breast cancer is not the
choice of agent (tamoxifen or aromatase inhibitor, or the role
of ovarian suppression), or indeed the optimal duration of
therapy, but rather the threshold for using adjuvant chemotherapy in addition to adjuvant endocrine therapy. As we
have come to better understand the various intrinsic subtypes of breast cancer and their differential prognoses,40
oncologists have come to debate the relative benefit of
adjuvant chemotherapy compared with endocrine therapy,
and ask whether for many patients with endocrine responsive breast cancer, a hormonal approach alone will be
sufficient to maximize a patients chance of cure. This has
been reflected in the recent 2011 St. Galen Consensus
guidelines on the management of early-stage breast cancer41previous guidelines in 2007 and 2009 had started to
characterize what might be considered truly endocrineresponsive breast cancer based on levels of receptor expression for both ER and PgR with absence of proliferation
markers. In these so-called endocrine responsive tumors,
the addition of adjuvant chemotherapy to endocrine therapy
was only recommended if other risk features related to
tumor burden were present (i.e., tumor size above 5 cm,
greater than 4 nodes, extensive vascular invasion).
The ability to sub-classify ER-positive breast cancer into
Luminal A or Luminal B subtypes based on either clinicopathologic determination (i.e., ER, PgR, HER2 and Ki-67)
or gene expression profiling is likely to identify ER-positive
cancers with different prognoses and altered levels of endocrine responsiveness. In addition, the 21-gene signature
(Oncotype DX) may also be used to predict chemotherapy
benefit in a patient with an ER-positive breast cancer based
on similar biologic and pathologic features, and in clinical
practice is increasingly being used to identify good prognosis
ER-positive breast cancer with a low recurrence score,
where the addition of chemotherapy to endocrine treatment
may yield no additional benefit.42 This is now being tested
prospectively in the Trial Assigning IndividuaLized Options
for Treatment (Rx) (TAILORx) where the Oncotype DX
assay is being used in more than 10,000 patients to guide

treatment decisions. As we come to understand breast cancer subtypes better, it is hoped that these molecular tools
will allow us to stratify women with ER-positive early-stage
breast cancer much more effectively into those where chemotherapy has a real role to play, and those with truly
endocrine-sensitive disease where the optimal strategy for
endocrine therapy alone discussed above is all that is required to cure the disease.
Conclusion
Endocrine therapy for early-stage breast cancer has had

the biggest single effect on enhancing survival from the
disease, with tamoxifen alone contributing to saving many
thousands of lives. In postmenopausal women, enormous
progress has been made by the incorporation of aromatase

inhibitors into the treatment of early-stage ER-positive
breast cancer, and large well-conducted trials have established up-front or switch strategies that are now widely
used in clinical practice. Increasingly, extended adjuvant
therapy is being considered, as longer may be better for
some women who have an ongoing risk of recurrence beyond
year 5. For others less may be more in terms of molecular
profiling informing us of those women who do not need
chemotherapy. As such, we are refining how to use our
therapies beyond the one strategy fits all approach of old.
Endocrine therapy will continue to evolve, and current
research is now exploring novel approaches to enhance
endocrine responsiveness even further through combination
approaches with novel targeted therapies.
Authors Disclosure of Potential Conflicts of Interest
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Stephen R. Johnston
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
AstraZeneca;
GlaxoSmithKline
REFERENCES
1. Early Breast Cancer Trialists Collaborative Group (EBCTCG). Effects of
chemotherapy and hormonal therapy for early breast cancer on recurrence
and 15-year survival: An overview of the randomised trials. Lancet. 2005;365:
1687-1717.
2. Dignam JJ, Dukic V, Anderson SJ, et al. Hazard of recurrence and
adjuvant treatment effects over time in lymph node negative breast cancer.
Breast Cancer Res Treat. 2009;116:595-602.
3. Early Breast Cancer Trialists Collaborative Group (EBCTCG). Relevance of breast cancer hormone receptors and other factors to the efficacy of
adjuvant tamoxifen: Patient-level meta-analysis of randomised trials. Lancet.
2011;378:771-784.
4. Fisher B, Dignam J, Bryant J, Wolmark N. Five versus more than five
years of tamoxifen for lymph node-negative breast cancer: Updated findings
from the National Surgical Adjuvant Breast and Bowel Project B-14 randomized trial. J Natl Cancer Inst. 2001;93:684-690.
5. Peto R. ATLAS (Adjuvant Tamoxifen, Longer Against Shorter); international rndomised trial of 10 versus 5 years of adjuvant tamoxifen among 11
500 women - preliminary results. Presented at: 2007 San Antonio Breast
Cancer Symposium; Late breaking abstract 48.
6. Gray RG, Handley A, et al. aTTom (adjuvant tamoxifen- to offer more?)
: Randomised trial of 10 versus 5 years of adjuvant tamoxifen among 6,934
women with etrogen receptor positive (ER) or ER untested breast cancerpreliminary results. J Clin Oncol. 2008;26:15S (suppl; abstr 513).
7. Smith IE, Dowsett M. Aromatase inhibitors in breast cancer. N Engl
J Med. 2003;348:2431-2442.
8. Dowsett M, Cuzick J, Ingle J, et al. Meta-analysis of breast cancer
outcomes in adjuvant trials or aromatase inhibitors vs tamoxifen. J Clin
Oncol. 2010;28:509-518.
9. Burstein HJ, Prestrud AA, Seidenfeld J, et al. American Society of
Clinical Oncology Clinical Practice Guideline: Update on adjuvant endocrine
therapy for women with hormone-receptor-positive breast cancer. J Clin
Oncol. 2010;28:3784-3796.
10. The ATAC Trialists Group. Anastrozole alone or in combination with
tamoxifen versus tamoxifen alone for adjuvant treatmnet of postmenopausal
women with early breast cancer: First results of the ATAC randomised trial.
Lancet. 2002;359:2131-2139.
11. The Arimidex, Tamoxifen, Alone or in combination (ATAC) Trialists
Group. Effect of anastrozole and tamoxifen as adjuvant treatment for earlystage breast cancer: 100-month analysis of the ATAC trial. Lancet Oncol.
2008;9:45-53.
12. Cuzick J, Sestak I, Baum M, et al. Effect of anastrozole and tamoxifen
as adjuvant treatment for early stage breast cancer: 10-year analysis of the
ATAC trial. Lancet Oncol. 2010;11:1135-1141.
13. The ATAC (Arimidex, Tamoxifen, Alone or in Combination) Trialists
Group. Anastrozole alone or in combination with Tamoxifen versus Tamox-
ifen alone for adjuvant treatment of postmenopausal women with early stage
breast cancer. Cancer. 2003;98:1802-1810.
14. Eastell R, Adams JE, Coleman RE, et al. Effect of anastrozole on bone
mineral density; 5-year resulst from the anastrozole, tamoxifen, alone or in
combination trial 18233230. J Clin Oncol.2008;26:1051-1058.
15. The Breast International Group (BIG) 1-98 Collaborative Group. A
comparison of letrozole and tamoxifen in postmenopausal women with early
16. The BIG 1-98 Collaborative Group. Letrozole therapy alone or in
sequence with tamoxifen in women with breast cancer. N Engl J Med.
2009;361:766-776.
17. Coombes RC, Hall E, Gibson LJ, et al. A randomised trial of exemestane after two to three years of tamoxifen therapy in postmenopausal women
18. Coombes RC, Kilburn LS, Snowdon CF, et al. Survival and safety of
exemestane versus tamoxifen after 2-3 years tamoxifen treatment (Intergroup Exemestane Study): a randomised controlled trial. Lancet. 2007;369:
559-570.
19. Jackesz R, Jonat W, Gnant M, et al. Switching of postmenopausal
women with endocrine-responsive early breast cancer to anastrozole after 2
years adjuvant tamoxifen; combined resulst of ABCSG trial 8 and ARNO 95
trial. Lancet. 2005;366:455-462.
20. Boccardo F, Rubagotti A, Puntoni M, et al. Switching to anastrozole
versus continued tamoxifen treatment of early breast cancer. preliminary
results of the Italian Tamoxifen Anastrozole (ITA) trial. J Clin Oncol.
2005;23:5138-5147.
21. Jonat W, Gnant M, Boccardo F, et al. Effectiveness of switching form
adjuvant tamoxifen to anastrozole in postmenopausal women with hormonesensitive early-stage breast cancer: A meta-analysis. Lancet Oncol. 2006;7:
991-996.
22. Van de Velde CJH, Rea D, Seynaeve C, et al. Adjuvant tamoxifen and
exemestane in early breast cancer (TEAM): a randomised phase 3 trial.
Lancet. 2011;377:321-331.
23. Dowsett M, Allred C, Knox J, et al. Relationship between quantitative
estrogen and progesterone receptor expression and human epidermal growth
factor 2 (HER-2) status with recurrence in the Arimidex, Tamoxifen, Alone or
in Combination Trial. J Clin Oncol. 2008; 26:1059-1065.
24. Viale G, Regan MM, Maiorano E, et al. Prognostic and predictive value
of centrally reviewed expression of estrogen and progesterone receptors in a
randomised trial comparing letrozole and tamoxifen adjuvant therapy for
postmenopausal early breast cancer: BIG 1-98. J Clin Oncol. 2007;25:38463852.
25. Bartlett JMS, Brookes CL, Billingham LJ, et al. A prospectively
planned pathology study within the TEAM trial confirms that progesterone
receptor expression is prognostic, but is not predictive for differential re-
25
STEPHEN R. JOHNSTON
sponse to exemestane vs tamoxifen. Presented at: 2008 San Antonio Breast
Cancer Symposium.
26. Viale G, Giobbie-Hurder A, Regan MM, et al. Prognostic and predictive
value of centrally reviewed Ki-67 labelling index in postmenopausal women
with endocrine-responsive breast cancer: Results form Breast International
Group trial 1-98 comparing adjuvant tamoxifen with letrozole. J Clin Oncol.
2008;26:5569-5575.
27. Dowsett M, Cuzick J, Wale C, et al. Prediction of risk of distant
recurrence using the 21-gene recurrence score in node-negative and nodepositive postmenopausal patients with breast cancer treated with anastrozole
or tamoxifen: A TransATAC study. J Clin Oncol. 2010;28:1829-1834.
28. Buzdar A, Howell A, Cuzick J, et al. Comprehensive side-effect profile
of anastrozole and tamoxifen as adjuvant treatment for early-stage breast
cancer: Long-term safety analysis of the ATAC trial. Lancet Oncol. 2006;7:
633-643.
29. Hillner BE, Ingle JN, Chlebowski RT, et al. American Society of
Clinical Oncology 2003 update on the role of bisphosphonates and bone health
issues in women with breast cancer. J Clin Oncol. 2003;21:4042-4057.
30. Goss PE, Ingle JN, Chapman J-AW, et al. Final analysis of NCIC
MA.27; a randomised phase III trial of exemestane versus anastrozole in
postmenopausal women with hormone receptor positive primary breast
cancer. Presented at San Antonio Breast Cancer Symposium 2010, Abstract
S1-1.
32. Goss PE, Ingle JN, Martino S, et al. Randomised trial of letrozole
following tamoxifen as extended adjuvant therapy in receptor-positive breast
cancer: Updated findings from NCIC CTG MA. 17. J Natl Cancer Inst.
2005;97:1262-1271.
33. Goss PE, Ingle JN, Martino S, et al. Efficacy of letrozole extended
adjuavnt therapy according to estrogen receptor and progesterone receptor
status of the primary tumour: National Cancer Institute of Canada Trials
Group MA. 17. J Clin Oncol. 2006;25:2006-2011.
26
34. Goss PE, Ingle JN, Pater JL, et al. Late extended adjuvant treatment
with letrozole improved outcome in women with early stage breast cancer who
completes 5 years of tamoxifen. J Clin Oncol. 2008;26:1948-1955.
35. Jakesz R, Greil R, Gnant M, et al. Extended adjuvant therapy with
anastrozole among postmenopausal breast cancer patients; results from the
randomised Austrian Breast and Colorectal Cancer Study Group Trial 6a.
J Natl Cancer Inst. 2007;99:1845-1853.
36. Mamounas EP, Jeong J-H, Wickerham L, et al. Benefit from exemestane as extended adjuvant therapy after 5 years of adjuvant tamoxifen;
intention-to-treat analysis of the National Surgical Adjuvant Breast and
Bowel Project B-33 Trial. J Clin Oncol. 2008;26:1965-1971.
37. Davidson NE, ONeill AM, Vukov AM, et al. Chemoendocrine therapy
for pre-menopausal women with axillary lymph-node positive, steroid hormone receptor positive breast cancer: Results from INT 0101 (E5188). J Clin
Oncol. 2005;23:2973-2982
38. International Breast Cancer Study Group. Adjuvant chemotherapy
followed by goserelin versus either modality alone for pre-menopausal lymph
node negative breast cancera randomised trial. J Natl Cancer Inst. 2003;
95:1833-1846.
39. Cuzick J, Ambroisine L, Davidson N, et al. Use of luteinising-hormonereleasing hormone agonists as adjuvant treatment in premenopausal patients
with hormone-receptor-positive breast cancer: A meta-analysis of individual
patient data from randomised adjuvant trials. Lancet. 2007;369:1711-1723.
40. Sotiriou C and Pusztai L. Gene expression signatures in breast cancer.
N Engl J Med. 2009;360:790-800.
41. Goldhirsch A, Wood WC, Coates RD, et al. Strategies for subtypes dealing with the diversity of breast cancer; highlights of the St Galen
International Expert Consensus on the Primary Therapy of Early Breast
Cancer 2011. Ann Oncol. 2011;22:1736-1747.
42. Albain K, Barlow WE, Shak S, et al. Prognostic and predictive value of
the 21-gene recurrence score assay in postmenopausal women with nodepositive, oestrogen-receptor positive breast cancer on chemotherapy; a retrospective analysis of a randomised trial. Lancet Oncol. 2010;11:55-65.
A DICKENS TALE OF THE TREATMENT OF

ADVANCED BREAST CANCER:
THE PAST, THE PRESENT, AND THE FUTURE
CHAIR
George W. Sledge Jr., MD
Indiana University Simon Cancer Center
Indianapolis, IN
SPEAKERS
Fatima Cardoso, MD
Champalimaud Cancer Center
Lisbon, Portugal
Martine J. Piccart, MD, PhD
Jules Bordet Institute
Brussels, Belgium
Eric P. Winer, MD
Boston, MA
A Dickens Tale of the Treatment of Advanced

Breast Cancer: The Past, the Present, and
the Future
By George W. Sledge Jr., MD, Fatima Cardoso, MD, Eric P. Winer, MD,
and Martine J. Piccart, MD
Overview: Metastatic breast cancer (MBC), a usually incurable disease, continues to vex physicians and patients. Recent
decades have seen great improvements in the treatment of
MBC, based on the availability of novel targeted therapeutics
and more standard chemotherapeutic agents. This article
describes the goals of therapy for MBC, the progress made

against MBC in recent decades, the current standard of care,
and the ongoing efforts of basic and translational researchers
to transfer the fruits of modern scientific discovery to patients
in the clinic.
Palliation of Symptoms and Maintenance of

Quality of Life
OTENTIAL GOALS of care in MBC include cure, prolongation of survival, palliation of symptoms and maintenance of quality of life, the development of new treatment
options, and what might be termed a good death for
patients with advanced disease.
Cure
Metastatic breast cancer is usually incurable. Nevertheless, there are patients who are long-term survivors of the
disease (whether the word cure should be applied to such
patients is contentious). Though the overall percentage of
patients is small (1% to 2% of patients in the largest
reported series), their existence is inarguable and suggests
that long-term survival represents a possible goal, if a rare
one.1
Prolongation of Survival
If cure is rare, prolongation of survival nevertheless represents a reasonable goal. How much does therapy improve
survival in metastatic breast cancer? We lack trials comparing active therapy with best supportive care. Studies suggesting an improvement over time in median survival2
imply that this improvement is related to therapy, but are
potentially flawed because of earlier detection of metastatic
disease. A recent study by the Eastern Cooperative Oncology
Group has suggested that, adjusted for disease relapse-free
interval, overall survival (OS) of patients with metastatic
disease has not improved.3
The best evidence for improvements in survival comes
from studies comparing active therapies. These include
comparisons of anthracycline- with nonanthracycline-based
therapies, of chemotherapy with chemotherapy plus HER2targeted therapy, of aromatase inhibitor therapy with tamoxifen in estrogen receptor-positive disease, and of
eribulin with doctors best choice.4-7 In each of these cases,
the improvement associated with the new intervention is
measurable in months. The cumulative effect of such improvements is hard to quantify.
As current patients are more heavily pretreated (because
of increasingly intensive adjuvant therapies) than their
predecessors, it may become increasingly difficult to demonstrate improvements in OS. The availability of multiple
lines of systemic therapy in the metastatic setting may also
dilute or hide the benefit of individual new agents, whether
in the front-line or refractory disease settings.8
28
Metastatic breast cancer impairs quality of life through its

symptom-producing effects on organ-specific function (e.g.,
bone metastasis-related fractures and pain) and general
effects on quality of life (both physical and psychologic). An
important part of the physicians role is to maintain patient
quality of life and palliate cancer-related symptoms.
The tools available for this task have expanded in recent
decades. These include appropriate pain control, treatment
of isolated metastases (e.g., brain metastases and epidural
cord compression), antiemetic therapy, bone maintenance
therapy (with bisphosphonates and denosumab), and psychosocial and dietary interventions. Appropriate pain control in particular is critical, and prompt referral to pain
specialists and palliative care experts is valuable in complex
cases.
Systemic therapies obviously play a role in maintenance of
quality of life and palliation of symptoms. Measuring the
effects of systemic therapy on health-related quality of life in
the metastatic setting has been difficult because of the
general lack of placebo-controlled trials, the inherent difficulty of measuring health-related quality of life over time,
and the competing toxicities of systemic therapy.
Psychosocial support, often for the patient and the family,
is equally important, and symptoms of insomnia, anxiety,
and depression often require treatment. In the optimal
situation, palliative and psychosocial care should be seamlessly integrated into the medical care provided by the
primary oncologist after discussion with a multidisciplinary
team.
Developing New Agents
The metastatic setting has played an important role in the

cure of micrometastatic breast cancer through the development of novel agents applicable to early disease. Physicians,
patients, and society in general owe much to the altruism of
patients with MBC. All patients with MBC should be offered
clinical trials of novel agents as an appropriate treatment
From the Indiana University Simon Cancer Center, Indianapolis, IN; Breast Unit,
Champalimaud Cancer Center, Lisbon, Portugal; Dana-Farber Cancer Institute, Boston,
MA; Institut Jules Bordet, Universite Libre de Bruxelles, Belgium.
Address reprint requests to George Sledge, Jr., MD, Indiana Cancer Pavilion, 535
Barnhill Dr., RT-473, Indianapolis, IN 46202; email: gsledge@iupui.edu.
1092-9118/10/1-10
TREATMENT OF ADVANCED BREAST CANCER
option. New trials represent an important source of hope

and offer potential therapeutic benefitfor patients with
advanced breast cancer. The National Institutes of Healths
clinicaltrials.gov website represents an invaluable resource
in this regard.
A Good Death
Because MBC is generally incurable, it is important for

physicians to be honest with patients regarding their prognosis, particularly as treatment options dwindle. Advanced
care planning should be part of a physicians discussion with
patients. Appropriate end-of-life care includes the discussion
of hospice care as a therapeutic option. The American
Society of Clinical Oncologys Advanced Care Planning booklet, available freely through its Cancer.net website, is a good
starting point for patients with MBC.
Treatment Past: Lessons Learned, but Lets Move On
One of the major criticisms raised nowadays of the older

clinical trials is the fact that most of them were run in
all-comers without a biologic-based patient selection. However, it was in the metastatic setting where the first targeted
trials were conducted. Our oldest form of targeted therapy
endocrine therapywas first used in MBC patients. The
understanding of the crucial predictive role of the estrogen
receptor (ER) led to selection based on ER status. The first
studies of the anti-HER2 agent trastuzumab were performed in patients with MBC selected by HER2 receptor
status. In the trials of cytotoxic agents, patient selection
according to biology is only a recent phenomenon.
Multidisciplinary and Locoregional Treatments
An important change in oncology occurred with the understanding of the crucial role of a multidisciplinary approach
to cancer treatment, with active cooperation among all
specialists involved in the care of these patients. Unfortunately, this multidisciplinary approach is often forgotten in
the advanced setting. With the development of efficacious
locoregional treatments to several types of metastases
(mainly brain, bone, and liver) the multimodal approach has
become more prominent.
Many unanswered questions remainthe role of locoregional therapy of the primary cancer for patients diagnosed
with stage IV breast cancer (a survival benefit is suggested
KEY POINTS
The goals of care in metastatic breast cancer (MBC)

include cure, prolongation of survival, palliation of
symptoms, development of new agents, and a good
death.
The roots of targeted therapy lie in past decades of
clinical trials.
These trials also did much to elucidate the proper role
of systemic chemotherapy regimens in MBC.
Current treatment approaches are based on the
proper use of knowledge regarding growth factor
receptors (estrogen receptor and HER2).
Novel treatments for MBC attack increasing portions
of the hallmarks of cancer.
by retrospective series,9 and there are three ongoing trials);

the best candidates, timing, and approach (surgery, radiofrequency, a combination of both) for liver and lung metastases; and the best sequence of therapies for brain
metastases (surgery, radiotherapy, radio-surgery, combination with a systemic therapy).
Systemic Therapy: Best Endpoints, Survival Benefits
Though many new agents have been developed and incorporated into the treatment of MBC, very few provided a
survival benefit,10 and when they did, it was almost exclusively as first-line therapy (doxorubicin paclitaxel vs.
5-FU doxorubicin cyclophosphamide; letrozole vs. tamoxifen) or when compared with very old and abandoned
drugs (i.e., melphalan or mitomycin plus vinblastine) or in
trials that compared combination regimens with single
agents, although neglecting planned cross-over (docetaxel
capecitabine vs. docetaxel alone or paclitaxel gemcitabine
vs. paclitaxel alone).
Although OS benefit is undoubtedly the most desired
outcome, progression-free survival (PFS) has been the most
widely used endpoint. However, it is not a good surrogate for
OS benefit,11 and no good surrogate has yet been developed.
Heated discussions regarding the merits of OS or PFS as the
most adequate endpoint are ongoing, as is the incorporation
of validated quality-of-life measurements and patientreported outcomes.
Endocrine Therapy (ET): Optimal Agents and Optimal Sequence
of Therapies
For ER-positive premenopausal women, some relatively

small trials and a meta-analysis found statistically significant survival benefits with third-generation aromatase inhibitors (AIs) (letrozole, exemestane, and anastrozole)
(relative hazard (RH) 0.87, CI 95%: 0.82 to 0.93; p
0.001). In first-line trials, in which AIs were compared with
tamoxifen, the survival benefit was of 11% RH reduction
(95% CI: 1% to 19%; p 0.03). Their benefit in second and
subsequent line trials, in which they were compared with
other treatments, was similar.12 However, these trials have
been performed in a era when tamoxifen was the gold
standard adjuvant ET, and currently the great majority of
postmenopausal women receives an AI in the adjuvant
setting. Trials in MBC patients pretreated with AIs have
shown that another AI, fulvestrant, and tamoxifen are
viable options at progression.
For ER-positive premenopausal women, tamoxifen with or
without ovarian suppression/ablation (OS/OA) has remained
the standard adjuvant therapy, and some relatively small
trials and a meta-analysis provide evidence that for first-line
therapy for MBC ER-positive premenopausal patients, the
combination of OS/OA with tamoxifen is superior to tamoxifen alone. Although frequently used in the clinic, the
combination OS/OA with an AI has not been properly
evaluated in the metastatic setting, and, unfortunately,
randomized trials of fulvestrant for MBC have not included
premenopausal women.
Chemotherapy: Optimal Drug/Regimen for First, Second, and More
Lines of Therapy; Combination Versus Sequential Monotherapy;
Optimal Sequence of Drugs
One of the most challenging and controversial issues in

MBC treatment relates to chemotherapy use. Almost all
29
SLEDGE, CARDOSO, WINER, AND PICCART
available data comes from an era when taxane adjuvant use

was not yet standard, and even anthracycline-based regimens were not always used. These data may not apply to our
current MBC patient population.
Available trials have produced the following conclusions:
for taxane-naive and anthracycline-naive/minimally exposed patients, single-agent anthracycline or single-agent
taxane yield similar results; for taxane-naive and anthracycline-resistant/refractory patients, single-agent taxane led
to better outcomes than single-agent anthracyclines; combinations of anthracyclines plus taxanes in the metastatic
setting consistently led to higher response rates, sometimes
higher time to progression or PFS, higher toxicity, but not
better OS. Because of this, sequential single agent therapy
is not only appropriate, but preferred for most patients. A
2008 meta-analysis of individual patient data concludes that
taxanes do not improve survival when compared with anthracyclines, either as single agents or in anthracycline
combinations and that taxanes in combination with anthracyclines modestly improve response rate (RR) and PFS but
not OS.13
For patients pretreated with both anthracyclines and
taxanes, the most consistent data concerns capecitabine use.
Vinorelbine has been compared head-to-head with docetaxel
and yielded similar efficacy and significantly less toxicity.
Given that both these agents are available as oral formulations and are associated with much less toxicity including
alopecia, they are attractive options. Although there is also
some data supporting rechallenging with taxanes as firstline therapy, the wealth of other available options and
toxicity concerns make this a less appealing option.
A common question relates to the use of combination of
cytotoxic agents compared with their sequential use as
monotherapy. Available evidence14 suggests monotherapy
as the preferred option (even in the presence of visceral
metastases) in all cases except those with a rapid progressive or highly symptomatic disease.
It is important to know that there are no data to support
an optimal sequence of therapies and that only eribulin led
to an OS benefit. Therefore, treatment decisions must be
individualized, taking into account many other factors beyond efficacy.
Chemotherapy: Optimal Duration and Number of Lines of Therapy
Many studies have looked into the question of optimal

duration of systemic therapy, mainly chemotherapy, for
which the balance between efficacy and toxicity is perhaps
more difficult to achieve. Recently, a meta-analysis of published trials15 has looked into this issue and concluded that
longer first-line chemotherapy duration is associated with
marginally longer OS and a substantially longer PFS, and
treatment should be prescribed until progression or unacceptable toxicity.
Treatment Present: Metastatic Breast Cancer in 2012
Todays systemic treatment decisions are guided by a

range of factors: 1) the subtype of the breast cancer (defined
by ER status and HER2) 2) sites of disease 3) disease tempo
4) presence and extent of symptoms 5) prior therapy 6)
availability of effective local therapy and 7) patient preferences.
30
Hormone Receptor-Positive MBC
Hormone therapy is the preferred first-line treatment for

most. Several options are available, including the AIs
tamoxifen and fulvestrant (not U.S. Food and Drug Administration (FDA)-approved for first-line use). In postmenopausal patients who present de novo with metastatic disease
or those who have only received tamoxifen in the adjuvant
setting, treatment with an AI is standard of care.16 A
substantial number of these patients remain on an AI with
disease control for prolonged periods. In a minimally symptomatic patient without an extensive disease burden, it is
reasonable to try a second-line agent, even if there has been
no or little benefit with the initial treatment. Among patients who develop disease progression after an objective
response or prolonged disease stabilization on first-line
treatment, another endocrine agent is generally preferred.
Treatment with fulvestrant, a steroidal AI (assuming a
nonsteroidal agent was used initially) or tamoxifen (unless
there had been prior progression on tamoxifen) are reasonable.17 Unfortunately, clinical benefit from second-line therapy in the metastatic setting is less impressive than with
first-line treatment. Some patients will continue to derive
benefit from hormone therapy and can go on to receive thirdand fourth-line therapy.
Many patients with a new diagnosis of MBC will have
already progressed on an adjuvant aromatase inhibitor. For
these, options are more limited. In general, fulvestrant is
used as the first-line metastatic treatment in such patients.
A recent study conducted by the Southwest Oncology Group
that compared anastrozole alone with anastrozole plus fulvestrant in the first-line setting18 suggested an advantage
both in terms of PFS and OS for the combination.
In premenopausal women with MBC, tamoxifen remains
the standard treatment, often administered in conjunction
with a luteinizing hormone releasing hormone (LHRH) agonist.19 For premenopausal women who develop metastatic
disease while on tamoxifen, the usual approach is a combination of an LHRH agonist and an AI. AIs are ineffective in
women with functioning ovaries.
Ultimately, women with hormone receptor-positive disease become refractory to endocrine therapy. At that point,
chemotherapy is administered. A wide variety of choices are
available, and treatment options are similar to those that
are available for women with triple-negative disease. Both
the taxanes and capecitabine are commonly given in the
first-line setting. Although capecitabine does not have FDA
approval for first-line therapy, several small studies comparing capecitabine with other agents have suggested that
capecitabine is a reasonable initial approach.20
Triple-Negative Breast Cancer
Options are limited to chemotherapy in patients with

triple-negative breast cancer. Most patients have received
adjuvant chemotherapy, often with a short disease-free
interval. For these reasons, chemotherapy options in the
metastatic setting are often limited. Survival from the time
of metastatic disease tends to be relatively short, with a
median of approximately 1 year.21
Treatment options for metastatic triple-negative disease
include all of the available chemotherapeutic agents and
regimens. Given the widespread use of anthracyclines in the
adjuvant setting, these agents are not commonly used in the

Table 1. Targeting Breast Cancer Cells
Compound
Molecular Target
Clinical Trials
Pharmaceutical
HER2
Pertuzumab
HER2
-Phase I pertuzumab, T-DM1 and paclitaxel in HER2 MBC (NCT00951665)

-Phase I/II T-DM1 with pertuzumab in trastuzumab pretreated HER2 MBC (NCT00875979)
-Phase II Pertuzumab, trastuzuman and weekly paclitaxel in HER2 MBC (NCT01276041)
-Phase III pertuzumab T-DM1 versus T-DM1placebo versus trastuzumab taxane in
HER2 MBC (MARIANNE) (NCT01120184)
-Phase II trastuzumab capecitabine pertuzumab in HER2 MBC (PHEREXA)
(NCT01026142)
-Phase II pertuzumab with trastuzumab and AI in HRHER2 MBC (NCT01491737)
-Phase II of pertuzumab and trastuzuman with neoadjuvant chemotherapy in HER2 BC
(NCT00976989)
-Phase III, trastuzumab chemotherapy pertuzumab in early stage HER2 BC (APHINITY)
(NCT01358877)
Genentech
Trastuzumab-DM1
HER2
-Phase I pertuzumab, T-DM1 and paclitaxel in HER2 MBC (NCT00951665)

-Phase I T-DM1 monotherapy in HER2 MBC in patients with normal or reduced hepatic
function (NCT01513083)
-Phase I with T-DM1 GDC0941 versus trastuzumab GDC0941 in trastuzumab pretreated
HER2 MBC (NCT00928330)
-Phase I T-DM1 docetaxel in HER2MBC (NCT00934856)
-Phase I/II T-DM1 with pertuzumab in trastuzumab pretreated HER2 MBC (NCT00875979)
-Phase II T-DM1 sequentially with anthracycline-based chemotherapy, as adjuvant or
neoadjuvant therapy in HER2 BC (NCT01196052)
-Phase III pertuzumab T-DM1 versus T-DM1 placebo versus trastuzumab taxane in
HER2 MBC (MARIANNE) (NCT01120184)
-Phase III T-DM1 versus lapatinib capecitabine in HER2 MBC (EMILIA) (NCT00829166)
Genentech
HER2/HER3 bispecific
moAb
-Phase I with cisplatin, capecitabine, trastuzumab or lapatinib, trastuzumab or paclitaxel,

trastuzumab in HER2 MBC (NCT01304784)
-Phase I with trastuzumab in HER2 MBC (NCT01097460)
-Phase I monotherapy in HER2 MBC (NCT00911898)
Merrimack
MM-121
HER3
-Phase
-Phase
-Phase
-Phase
Merrimack
Neratinib
EGFR, HER2 and HER4

TKI
-Phase I neratinib with trastuzumab and weekly paclitaxel in HER2 MBC (NCT01423123)
-Phase I/II neratinib with temsirolimus in HER2 or TN MBC (NCT01111825)
-Phase II neratinib with paclitaxel in HER2 MBC (NCT00445458)
-Phase I/II neratinib with trastuzumab in HER2 MBC (NCT00398567)
-Phase I/II neratinib with vinorelbine in HER2 MBC (NCT00706030)
-Phase II neratinib with capecitabine in HER2 MBC (NCT00741260)
-Phase II neratinib versus lapatinib with capecitabine in HER2 MBC
-Phase II monotherapy in HER2 MBC pts with CNS mets (NCT01494662)
-Phase II of neoadjuvant chemotherapy with neratinib or trastuzumab followed by
postoperative trastuzumab in HER2 BC (NCT01008150)
-Phase II of neratinib plus paclitaxel versus trastuzumab plus paclitaxel in HER2 MBC
(NEFERTT) (NCT00915018)
-Phase III of neratinib after adjuvant trastuzumab in HER2 BC (ExteNET) (NCT00878709)
Pfizer
Afatinib
EGFR, HER2 TKI
-Phase I afatinib with vinorelbine in EGFR and/or HER2 overexpressing advanced solid
tumors (NCT00906698)
-Phase II afatinib or afatinib with chemotherapy (paclitaxel or vinorelbine weekly) in HER2
MBC (LUX-Breast 2) (NCT01271725)
-Phase II Afatinib in HER2 inflammatory breast cancer (NCT01325428)
-Phase II afatinib alone or in combination with vinorelbine in HER2 MBC with CNS mets
(Lux-Breast 3) (NCT01441596)
-Phase III afatinib plus vinorelbine versus trastuzumab plus vinorelbine in HER2 MBC after
trastuzumab failure (Lux-Breast 1) (NCT01125566)
Boehringer Ingelheim
HGF/cMET
Foretinib
VEGFR2/cMET
-Phase I/II with lapatinib in HER2MBC (NCT01138384)

-Phase II monotherapy in met TNBC (NCT01147484)
GSK
Cabozantinib
c-Met/VEGFR2/AXL/
KIT/TIE2/FLT3/RET
-Phase II monotherapy in HR MBC (NCT01441947)
Exelixis
Onartuzumab (MetMAb)
c-Met
-Phase II of onartuzumab and/or bevacizumab in combination with paclitaxel in met TNBC

(NCT01186991)
Genentech
IGFR
Ganitumab (AMG 479)
IGF-1R
-Phase I/II with trastuzumab in HER2 MBC ( NCT01479179)

-Phase II with exemestane or fulvestrant in HR MBC (NCT00626106)
Takeda with Amgen
IGF-1R and InsR dual

Inhibition
-Phase II letrozole in HR MBC (NCT01225172)

-Phase I/II with trastuzumab in HER2MBC (NCT00788333)
Bristol Myers Squibb
HER Family
MM-111
BMS-754807
II of MM121 exemestane in HER2- MBC (NCT01151046)

II with paclitaxel in HER2- MBC (NCT01421472)
I with paclitaxel in HER2- MBC (NCT01209195)
I with XL147 (SAR245408) in advanced solid tumors (NCT01436565)
OSI-906
IGF-1R
-Phase II with OSI-906 letrozole/goserelin erlotinib in HR MBC (NCT01205685)
Astellas
Cixutumumab (IMC-A12)
IGF-1R
-Phase I/II with temsirolimus in advanced or MBC (NCT00699491)

-capecitabine and lapatinib Cixutumumab in HER2 locally advanced or MBC (stages IIIB
IV) (NCT00684983)
-Phase II cixutumumab antiestrogens in HR MBC (NCT00728949)
Lilly
Dalotuzumab
IGF1R
-Phase I monotherapy in the neoadjuvant setting (NCT00759785)

-Phase II dalotuzumab ridaforolimus (MK-8669) in HR MBC (NCT01234857)
Merck
31

Table 1. Targeting Breast Cancer Cells (Contd)
Compound
Molecular Target
Clinical Trials
Pharmaceutical
BMS-754807
IGF-1R/InsR TKI
-Phase II letrozole in HR MBC (NCT01225172)

-Phase I/II with trastuzumab in HER2 MBC (NCT00788333)
Bristol Myers Squibb
MEDI-573
IGF anti-ligand mAb
-Phase Ib/II with AI versus AI monotherapy in HR MBC (NCT01446159)
MedImmune
FGFR
Dovitinib (TKI 258)
FGF-R and VEGFR
-Phase I with AI in HR MBC (NCT01484041)

-Phase II salvage monotherapy in inflammatory MBC (NCT01262027)
Novartis
BGJ398
Pan-FGF-R TKI
-Phase I trial for advanced solid tumors bearing FGFR1 or FGFR2 amplification or FGFR3
mutation (NCT01004224)
Novartis
PI3K
XL147 (SAR245408)
Pan-PI3K inhibition
-Phase I/II with HER or HERTaxol in MBC 2nd line ( NCT01042925)

-Phase I with letrozole (NCT01082068)
-Phase I with MSC1936369B (MEK inhibition) in advanced solid tumors (NCT01357330)
Exelixis with Sanofi-Aventis
GDC-0941
Pan-PI3K Inhibition
-Phase I with paclitaxel and bevacizumab (NCT00960960)

-Phase Ib with trastuzumab or T-DM1 in MBC (NCT00928330)
-Phase II with fulvestrant versus fulvestrant (NCT01437566) in hormone resistant MBC
Genentech/Roche
BKM120
Pan-PI3K Inhibition
-Phase I with fulvestrant in HR MBC ( NCT01339442)

-Phase I with letrozole in HR MBC (NCT01248494)
-Phase I/II with trastuzumab in TzRes MBC ( NCT01132664)
Novartis
GDC-0032
PI3Ka
-Phase I monotherapy in advanced solid tumors ( NCT01296555)
Genentech
GDC-0980
PI3K-mTOR Inhibition
-Phase Ib with paclitaxel, bevacizumab and trastuzumab in MBC (NCT01254526)

-Phase II with fulvestrant versus fulvestrant (NCT01437566) in hormone resistant MBC
Genentech
BEZ235
-Phase
-Phase
-Phase
-Phase
-Phase
-Phase
-Phase
Novartis
I/II in advanced solid tumors enriched for MBC (NCT00620594)

I with everolimus in advanced solid tumors enriched for MBC (NCT01482156)
I/II with trastuzumab versus lapatinib capecitabine in MBC (NCT01471847)
I with paclitaxel trastuzumab in HER2 MBC (NCT01285466)
I with letrozole in MBC (NCT01248494)
I/II with paclitaxel in HER2- MBC (NCT01495247)
I/II monotherapy in HER2-/HR MBC (NCT01288092)
XL765
-Phase I with letrozole ( NCT01082068)
Exelixis/Aventis
PF04691502
-Phase Ib with letrozole in MBC (NCT01430585)
Pfizer
mTOR
INK128
Pan-mTOR
-Phase I with paclitaxel and trastuzumab in HER2 MBC (NCT01351350)
Intellikine
Akt
MK-2206
Akt
-Phase I with paclitaxel in MBC (NCT01263145)CTCs analysiscleaved caspase 3 analysis

-Phase I with paclitaxel and trastuzumab in HER2 advanced solid tumors (NCT01235897)
-Phase II monotherapy in MBC with mutPI3K or mutAkt and/or PTEN loss (NCT01277757)
-Phase I with trastuzumab lapatinib in HER2 advanced solid tumors
(NCT00963547)completed
-Phase I with AI or fulvestrant in HR MBC (NCT01344031)
-Phase I with lapatinib in HER2 MBC (NCT01245205)
Merck
Ras/MEK/ERK
AZD6244 (ARRY-142886)
MEK/ERK
-Phase I with docetaxel in MBC (NCT00600496)

-Phase II monotherapy in stage Ic-III ER- BC to induce ER expression (NCT01313039)
-Phase II fulvestrant AZD6244 in HR MBC ( NCT01160718)
AstraZeneca
GSK1120212
MEK1/2
- Defining the triple negative breast cancer kinome response to GSK1120212 in stage Ic-III
TNBC (NCT01467310)
GlaxoSmithKline
-Phase I/II letrozole PD 0332991 in HR MBC (NCT00721409)

-Phase I PD 0332991 with paclitaxel in MBC (NCT01320592)
Pfizer
CDK (Cyclin Dependent Kinase)

PD 0332991
CDK4, 6
Dinaciclib (SCH 727965)
CDK1, 2, 5, 9
-Phase I veliparib and dinaciclib carboplatin in advanced solid tumors (BRCA1,2 mutation
carriers included) (NCT01434316)
Merck
Seliciclib (CYC202)
CDK2, 7, 9
-Phase I seliciclib with sapacitabine in advanced solid tumors (NCT00999401)

-Phase I liposomal doxorubicin with seliciclib in met TNBC (NCT01333423)
Cyclacel
Aurora Kinases
Alisertib
Aurora-A Kinase
-Phase I/II momotherapy in advanced solid tumors (NCT01045421)

-Phase I alisertib with paclitaxel in metastatic ovarian and breast cancer (NCT01091428)
Millennium Pharmaceuticals
AMG 900
Aurora-A, -B, -C Kinase
-Phase I monotherapy in advanced solid tumors (NCT00858377)
Amgen
Apoptosis
Navitoclax (ABT-263)
Bcl-2, Bcl-XL, and Bcl-w
-Phase I with paclitaxel in advanced solid tumors (NCT00891605)
Abbott
MDM2 Antagonist
Roche
HDAC
-Phase
-Phase
-Phase
-Phase
-Phase
Merck
RG7112
HDAC
Vorinostat
32
II with lapatinib in HER2 MBC (NCT01118975)

I with ixabepilone in MBC (NCT01084057)
II with tamoxifen in stage I-III BC (NCT01194427)
II with tamoxifen in HR MBC (NCT00365599)
II with AI in HR MBC (NCT01153672)

Table 1. Targeting Breast Cancer Cells (Contd)
Compound
Molecular Target
Clinical Trials
Pharmaceutical
Entinostat (SNDX-275 or
MS-275)
HDAC class I
-Phase II with lapatinib in trastuzumab resistant HER2 MBC (NCT01434303)

-Phase II with azacitidine in MBC (NCT01349959)
-Phase II exemestane entinostat in HR MBC ( NCT00676663)
-Phase II trial completed (ENCORE 301) of continuing AI upon progression of HR MBC with the
addition of entinostat (NCT00828854)
Syndax
HSP90Geldanamycin
derivative
-Phase II with trastuzumab in HER2 MBC (NCT00817362) completed (02.01.2012)

-Phase II with trastuzumab in trastuzumab pretreated HER2 MBC (NCT00773344) completed
Infinity
AUY922
HSP90 (nongeldanamycin
derivative)
-Phase I/II trial monotherapy in HER2 or ER MBC (NCT00526045)

-Phase I/II with trastuzumab in HER2 MBC (NCT01271920)
-Phase I/II trial with lapatinib and letrozole in HER2 MBC (NCT0136194)
Novartis
Retaspimycin (IPI-504)
HSP90
-Phase II with trastuzumab in trastuzumab pretreated HER2 MBC (NCT00817362) completed
Infinity
AR
Bicalutamide
AR
-Phase II monotherapy in ARER-PR- MBC (NCT00468715)
AstraZeneca
-Phase II abiraterone exemestane in ER MBC after AI failure (NCT01381874)
Johnson & Johnson
-Phase I monotherapy in PRL MBC (NCT01338831)
Novartis
HSP90
Tanespimycin (17-AAG)
Abiraterone
PRL
LFA 102
PRL-R
Abbreviations: MBC, metastatic breast cancer; AI, aromatase inhibitor; BC, breast cancer; MoAb, monoclonal antibody; EGFR, epidermal growth factor receptor; CNS,
central nervous system; GSK, GlaxoSmithKline; PTEN, phosphatase and tensin homolog; HDAC, histone deacetylase; AR, androgen receptor; FGF, fibroblast growth
factor; VEGFR, vascular endothelial growth factor receptor; PI3K, phosphatidylinositol 3-kinase; CDK, cyclin-dependent kinase; ER, estrogen receptor; PR,
progesterone receptor; IGF, insulin-like growth factor.
metastatic setting. If not previously used, an anthracycline

is an appropriate option. The taxanes (particularly weekly
paclitaxel) are a mainstay of treatment and one of the most
commonly employed first-line regimens, particularly as a
backbone for new treatment approaches with combined
biologic agents. Other active agents include capecitabine,
gemcitabine, eribulin, and vinorelbine. The platinum salts,
both carboplatin and cisplatin, are also used in patients with
metastatic triple-negative disease.22
A variety of biologic agents have been evaluated in patients with triple-negative disease in conjunction with chemotherapy. Bevacizumab was commonly administered with
a taxane for metastatic triple-negative disease, but the
benefits of this approach have been called into question
based on the results of recent trials and a meta-analysis, and
the FDA withdrew approval for bevacizumab as a treatment
for MBC.23,24 Although there is great interest in the PARP
inhibitors, and an initial randomized phase II clinical trial
suggested a substantial benefit when iniparib was combined
with carboplatin-based chemotherapy;25 a subsequent phase
III trial failed to support these preliminary results.26
HER2-Positive Breast Cancer
Trastuzumab, a humanized monoclonal antibody directed

against HER2, plus chemotherapy represents the standard
first-line approach for patients who have ER-negative disease or have had progression on endocrine therapy.4 In
initial trials, trastuzumab was combined with a taxane, but
more recent studies have suggested that other chemotherapeutic agents, such as vinorelbine, yield similar results.27
Recent evidence suggests that continued suppression of
HER2 with trastuzumab is useful even after disease progression.28
Lapatinib, a small molecule tyrosine kinase inhibitor
directed against EGFR and HER2, is also approved for
the treatment of HER2-positive breast cancer in combination with capecitabine.29 Many patients receive a combination of lapatinib and capecitabine at some point in their
treatment course, after the first or second trastuzumabbased regimen, and subsequently return to trastuzumab. Of
interest, the combination of trastuzumab and lapatinib
without chemotherapy has also been shown to be an active

regimen and substantially more effective than lapatinib
alone.30
The development of central nervous system (CNS) disease
is a major challenge that is faced by more than a third of all
women with HER2-positive metastatic disease. As women
live longer with HER2-positive metastatic disease, the incidence of CNS disease increases. Treatment options include
whole brain irradiation, stereotactic radiosurgery, surgery,
and a number of systemic approaches. Of the available
medical therapies, single-agent lapatinib results in rare
objective responses, but a combination of capecitabine plus
lapatinib appears to be more active.31
The development of trastuzumab (and, subsequently,
lapatinib) for MBC is an example of the problem of leaving
unanswered questions in the metastatic setting once therapy has moved to the adjuvant setting. It took almost a
decade to gather evidence about the best dose and regimen
(weekly and 3-weekly regimens with a loading dose are
similar), the best timing of introduction, the best agent to
combine it with (several options now known to exist), and
the efficacy of this treatment beyond progression. For patients with ER-positive, HER2-positive disease, the combination of aromatase inhibitor therapy with HER2-targeted
therapy is superior to aromatase inhibitor therapy alone.
This combination may be considered as an alternative to a
chemotherapy plus HER2-targeted therapy in selected patients.
Treatment Future: Newer Targeted Agents in
Metastatic Breast Cancer
Gene-expression profiling analysis studies32 have fundamentally changed the way we think about breast cancer,
which is now considered a group of diseases with distinct
genetic and epigenetic alterations. The translation of this
improved molecular knowledge into effective molecularlytargeted agents is slower than expected. The advent of
next-generation sequencing technologies has generated new
enthusiasm,33 with the promise of better identification of the
molecular defects responsible for carcinogenesis.
33
RTK
PIK3CA gene mutations

(exon 9 or 20)
PIK3CA gene
amplification
Truncating mutations
(Deletions)
Epigenetic silencing
PTEN
PAN-PI3K inhib.
PI3K
Dual PI3KmTor
inhibitors
PI3K inhibitors
AKT1 gene mutations

AKT2 gene amplification
BEZ 235
BGT 226
SF1126
GSK 1059615
GDC-0980
XL765
PF-4691502
mTORC2
BKM120
GDC-0941
XL147
PKI-587
AKT
mTORC1
Isoform specific
PI3K inhib.
AKT inhibitors
GSK 690693
MK 2206
mTOR CATALYTIC SITE

inhibitors
AZD 8055
OSI 027
INK 128
mTORC1 inhibitors
Everolimus
Fig. 1.
Genetic aberrations of the PI3K pathway in BC and PI3K pathway inhibitors.
Targeting Breast Cancer Cells
Lessons learned from the recent randomized testing of new

strategies for treating MBC. Recent months have witnessed
three striking advances and one failure with the investigation of novel strategies for MBC: dual receptor blockage, use
of an antireceptor antibody-drug conjugate, use of drugs
targeting the PI3K/AKT/mTOR pathway, and exploitation of
genome instability.
Dual receptor blockage was first explored in heavily pretreated patients with HER2-positive disease, using the combination of trastuzumab and lapatinib, which was superior
to lapatinib alone in a relatively small, although provocative, randomized trial.30 The recently published phase III
CLEOPATRA study34 testing dual HER2 blockade with
trastuzumab and pertuzumab along with docetaxel compared with docetaxel plus trastuzumab, represents a major
advance in the treatment of HER2-positive MBC. The combined strategy achieved an impressive PFS prolongation of
6.1 months and is likely to become a new standard of care.
Others are building on this concept to design cocktails of
monoclonal antibodies directed at several HER family members in pairs, such as two antibodies targeting two different
domains of each receptor,35 or at the design of bispecific
monoclonal antibodies, which target two HER family members simultaneously (e.g., HER2 and HER3 simultaneously)
(Table 1).
The hope is to induce downregulation of surface membrane receptors and reduce activation of compensatory pathways. It is unclear whether this strategy will be tolerable
34
and superior to the use of small molecules that potently

inhibit the tyrosine kinase of HER1-HER2 (afatinib) or
HER1-HER2-HER4 (neratinib). These oral agents are currently in phase III trials for HER2-positive MBC.
Three receptors are attracting growing interest as therapeutic targets: HER3, cMET, and FGFR. HER3 is not only a
key player in driving the growth of HER2-positive breast
cancer through the formation of the potent HER2-HER3
heterodimer (able to stimulate growth in a ligand-dependent
or independent manner), but also as a mediator of resistance
to antiestrogen therapies.36 Pertuzumab does not prevent
dimerization of HER3 with HER1 or other potential partners. Several anti-HER3 monoclonal antibodies in early
clinical development (Table 1) will be tested in HER2positive and luminal MBC.
C-MET and its stromal ligandthe hepatocyte growth
factorplay a key role in cell survival, mobility, and invasion. They have been proposed as drivers of resistance to
EGFR kinase inhibitors37 and first-line trastuzumabtreatment.38 New agents targeting C-MET alone (anti C-MET
MAb) or together with an angiogenic receptor such as
VEGFR2 (TKIs) are currently being explored in MBC (Table
1).
FGFR overexpression is robustly associated with FGFR1
amplification, and the latter occurs in triple-negative39 and
in luminal B-type40 breast cancer, where it is observed in
16% to 27% of the cases. FGFR1 overexpression promotes
endocrine therapy resistance. Two FGFR inhibitors have
entered the clinic, and a recently reported phase II study of

Table 2. Targeting Cancer Stem Cells
Compound
Molecular Target
Preclinical/Clinical Data
Pharmaceutical
Hedgehog
Sarigedib (IPI-926)
Smo Antagonist
-Phase I trials in pancreatic cancer and HNSCC
Infinity
XL139 (BMS-833923)
Smo Antagonist
Exelixis
LDE225
Smo Antagonist
-Phase I and II trials for pancreatic AdenoCa, medulloblastoma and basal cell Ca
Novartis
PF-04449913
Smo Antagonist
Pfizer
Notch
MK-0752
Gamma Secretase Inhibition
-Phase I with letrozole or tamoxifen in the neoadjuvant setting (NCT00756717)

-Phase I with ridaforolimus in advanced solid tumors (NCT01295632)
-Phase I/II with docetaxel in MBC (NCT00645333)
Merck
RO4929097
Gamma Secretase Inhibition
-Phase
-Phase
-Phase
-Phase
-Phase
Roche
Wnt
PRI-724
CBP--catenin
I/II with exemestane in HR MBC (NCT01149356)

I with paclitaxel and carboplatin in the neoadjuvant TNBC (NCT01238133)
I with capecitabine in advanced solid tumors (NCT01158274)
I with cediranib in advanced solid tumors (NCT01131234)
II monotherapy in met TNBC (NCT01151449)
Prism Biolab
Abbreviations: HNSCC, head and neck squamous cell carcinoma; AdenoCa, adenocarcinoma; Ca, carcinoma.
dovitinib (a dual FGFR1 and VEGFR TKI) demonstrated

antitumor activity in heavily pretreated MBC patients with
FGFR1 amplified tumors.41
Finally, there is a strong rationale for blocking IGF1R
signaling, which plays an important role in MBC growth and
is hyperactivated as a result of the release of a negative
feedback loop observed with mTORC1 inhibition. Agents
blocking IGF1R signaling are in clinical development with
early signs of clinical activity seen in a phase I trial combining ridaforolimus (a rapalog) and dalotuzumab (an antiIGF1R monoclonal antibody). A randomized phase II of this
combination compared with exemestane is ongoing.
The use of an antireceptor antibody-drug conjugate is
another striking advance in strategies for MBC.
Trastuzumab-DM1 (T-DM1) consists of trastuzumab covalently bound via a linker to DM1, a derivative of the
antimicrotubule cytotoxic drug maytansine. This elegant
way of localizing an active chemotherapy inside HER2positive tumor cells minimizes harm to normal cells, while
maintaining the full properties of the monoclonal antibody
(such as ADCC). This translates not only into improved
antitumor activity in a randomized phase II trial42 compared with the standard combination of docetaxel and trastuzumab, but also into a better toxicity profile as well as
improved patient reported outcomes in terms of quality of
life.
The use of drugs targeting the PI3K/AKT/mTOR pathway
is a novel strategy for MBC with future ramifications. The
central role of PI3K signaling in several cellular processes
critical for cancer progression has been clearly established.
Genetic alterations in several components of the PI3K pathway (Fig. 1) lead to aberrant pathway activation and mediate resistance to endocrine and anti-HER2 drugs. Among
the rapidly growing number of potential therapeutics targeting the PI3K signaling cascade, mTORC1 inhibitors are
the most advanced.
The BOLERO-2 phase III randomized clinical trial compared the efficacy of exemestane combined with the
mTORC1 inhibitor everolimus with exemestane alone in the
setting of hormone receptor-positive MBC in patients refractory to non-steroidal AIs.43 The biologic rationale involved
ligand-independent activation of ER through a cross-talk
Table 3. Targeting Tumor Stroma

Compound
Angiogenesis Inhibition
MEGF0444A/RG7414
Molecular Target
EGFL7
Preclinical/Clinical Data
Pharmaceutical
-Phase I with bevacizumab paclitaxel in advanced solid tumors (NCT01075464)
Genentech
MNRP1685/RG7347
NRP1
-Phase I with bevacizumab paclitaxel in advanced solid tumors (NCT00954642)
Genentech
TB-403/RG7334
PlGF
-Phase I monotherapy in advanced solid tumors (NCT00702494) completed
BioInvent International AB
with Genentech
Integrin Inhibitors
Cilengitide
v3
-Phase I with paclitaxel in MBC (NCT01276496)
Merck
PF-04605412
51 Integrin
Pfizer
IMGN388
av
ImmunoGen, Inc.
Adhesion Signalling Molecules

Removab (catumaxomab)
Ep-CAM and CD3
Trion
Hypoxia
Aminoflavone (AFP464)
HIF-1 mRNA
Tigris
EZN-2968
HIF-1 mRNA
-Phase I monotherapy in advanced solid tumors with hepatic mets (NCT01120288)
Enzon
Abbreviation: mets, metastases; HIF, hypoxia-inducible factor; Ep-CAM, epithelial cell adhesion membrane; NRP1, neuroplin1; PIGF, placental growth factor.
35

44
with the mTOR pathway. The combined treatment led to

an impressive 6.5 month extension of PFS.43
There are four additional classes of PI3K pathway inhibitors in clinical development: dual PI3K-mTOR inhibitors,
PI3K inhibitors that do not inhibit mTOR, AKT inhibitors,
and mTOR catalytic inhibitors (also called pan-mTOR inhibitors because they inhibit mTORC2 in addition to mTORC1)
(Fig. 1). The future will tell if these newer agents are
effective and tolerable pathway inhibitors or are able to
overcome feedback inhibition normally observed with
mTORC1 inhibitors.
Of critical importance will be the identification of populations in which the benefits of these PI3K pathway inhibitors
will justify their manageable but real toxicitythe current
assumption is that PI3KCA-mutated cancers and PTENdeficient cancers will be at the top of the list. This would
suggest that this family of compounds may find a role in all
three types of breast cancer.
Recently, there has been growing interest in combining
PI3K with MEK pathway inhibitors. Many cancers sensitive
to PI3K pathway inhibitors show tumor stasis rather than
regression. The MEK pathway can represent an escape
route in the presence of effective downregulation of the PI3K
pathway. Therapeutic inhibition of both pathways is under
clinical investigation with the use of MEK and PI3K inhibitors (Table 1).
Instability of the genome is inherent to the great majority
of human cancer cells and is instrumental for tumor progression. It has been proposed as a fundamental hallmark of
cancer.45
Although PARP inhibitors have clearly demonstrated
striking antitumor activity in BRCA mutation carriers with
advanced breast cancer, their activity in sporadic triplenegative breast cancer, alone or in combination with chemotherapy, has been less consistent. Myelosuppression has
been an obstacle to their development. It remains uncertain
whether PARP inhibitors will find an application beyond the
treatment of BRCA-mutated tumors.
Strategies directed at hallmarks of cancer in early development for MBC. Cyclin-dependent kinase inhibitors target
evasion from growth suppressors by directly interfering with
the cell cycle, downstream of the receptor tyrosine kinase
signaling cascades. Several CDK inhibitors are in phase I or
II trials, in combinations with either endocrine therapy in
luminal cancers or chemotherapy for basal-like cancers. A
randomized trial of letrozole with or without a CDK-4, -6
inhibitor25 has generated promising results.46
A similar anticancer strategy is inhibition of Aurora
kinases. These serine/threonine kinases play an essential
role in cellular proliferation, through control over chromatid
segregation. Agents targeting Aurora kinases have entered
early clinical trials.
Proapoptotic agents aim at antagonizing resistance to
cell death, and a number of them are in phase I clinical
trials (Table 1). Triple-negative breast cancer seems to be a
priority given the high prevalence of p53 mutations seen in
this subtype.
Strategies directed at epigenetic aberrations or at downregulation of key intracellular targets. As our understanding
of the deregulation of epigenetic mechanisms evolves, growing interest in targeting these aberrations has emerged.
Histone deacetylase inhibitors are progressing from phase I
to phase II trials with interesting results seen when these
36
agents are combined with endocrine agents or anti-HER2

drugs.
Heat shock protein 90 (HSP90) represents another interesting anticancer target in MBC.47 A molecular chaperone,
HSP90 plays a vital role in the intact function of several
oncogenic proteins and, thus, promotes key molecular events
in malignant progression. Several agents have been tested
with promising results in HER2-positive MBC.
Antihormone strategies. The androgen and prolactin receptors are being explored in advanced breast cancer. The
androgen receptor (AR) has recently been shown to play a
role in HER2-positive/estrogen receptor-negative breast
cancers, which can get stimulated by testosterone and also
show upregulation of the Wnt signaling pathway.48 Moreover, AR positivity has been found in approximately 20% of
triple-negative breast cancers,49 and a phase II study with
bicalutamide for ER-negative breast cancer expressing AR
has recently been performed.
Targeting Breast Cancer Stem Cells
The identification of a cellular subpopulation in breast

cancer with a specific immunophenotype (i.e., CD44
CD24-)50 bearing a tumor-initiating capacity was the first
identification of cancer stem cells (CSCs) in the setting of a
solid tumor. Ever since, a detailed exploration of this cellular compartment of breast cancer has been undertaken, and
this expanding knowledge provides novel therapeutic targets51 for this therapy-resistant cellular population.
One of the fundamental aspects of CSCs is their capacity
for self-renewal, mediated through well-conserved developmental signaling pathways52 (the Notch, Hedgehog, and
Wnt pathways) amenable to therapeutic targeting. The
Notch signaling pathway, frequently deregulated in breast
cancer,53 can be targeted by gammasecretase inhibitors
(Table 2). These small molecules block the translocation of
the intracellular part of the Notch receptor to the nucleus
and corresponding signaling pathway activation. The
Hedgehog signaling pathway has also been shown to promote the tumor-initiating ability of breast CSCs.54 A class of
compounds called Smo-antagonists has entered clinical trials (Table 2).
A second approach targeting breast CSCs is based on
high-throughput screening methods. A population of breast
cancer cells enriched for CSCs are screened either for
druggable targets through shRNA/siRNA technology or
through the testing of extended small molecule compound
collections. Based on the latter approach, Gupta and colleagues55 recently discovered that salinomycin is able to
eradicate breast CSCs.
Targeting the Tumor Microenvironment
Tumor cells are integrated into their microenvironment,

constantly interacting with it.56 These components are
appealing therapeutic targets, with antiangiogenic factors
being the most clinically advanced. The failure of bevacizumab, a humanized monoclonal antibody targeting
VEGF-A, to extend OS of MBC patients23,57 should not be
regarded as a proof of failure of the concept of antiangiogenic
therapy in breast cancer. The assumption that bevacizumab
administration potentiates the delivery of higher intratumor
chemotherapy concentrations proved incorrect.58 Moreover,
the redundancy of different angiogenic mechanisms and
proangiogenic factors could explain this clinical failure, as
inhibition of VEGF-A rapidly leads to compensatory activation of alternative pathways.59,60 Other compounds targeting the tumor microenvironment can be found in Table 3.
subtyping continues to advance therapy. The twin revolutions of cancer genomics and computational chemistry
should lead to many new treatment options for patients in
coming years.
Conclusion
Acknowledgments
MBC represents a continuing challenge to physicians and

patients. An improved understanding of breast cancer biology has improved prognosis, and progress based on biologic
M. Piccart would like to thank Jose Baselga, MD, PhD for

useful advice and Dimitrios Zardavas, MD, for great help in
preparation of the tables listing newer targets and drugs.
Author
George W. Sledge Jr.*
Fatima Cardoso
Eric P. Winer
Martine J. Piccart
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Abraxis
BioScience;
AstraZeneca;
Celgene; Eisai;
GlaxoSmithKline;
Johnson &
Johnson;
Novartis; Pfizer;
Roche
Novartis; Roche/
Genentech (U)
Amgen; Bayer
Schering
Pharma;
Boehringer
Ingelheim;
Bristol-Myers
Squibb;
GlaxoSmithKline;
PharmaMar;
Roche; Sanofi
Honoraria
Abraxis
BioScience;
AstraZeneca;
Celgene; Eisai;
GlaxoSmithKline;
Johnson &
Johnson;
Novartis; Pfizer;
Roche
Research
Funding
Expert
Testimony
Other
Remuneration
AstraZeneca;
Novartis
Roche/Genentech
Amgen; Bayer
Schering
Pharma;
Boehringer
Ingelheim;
Bristol-Myers
Squibb;
GlaxoSmithKline;
PharmaMar;
Roche; Sanofi
Bristol-Myers
Squibb; Pfizer;
Roche
REFERENCES
1. Greenberg P, Hortobagyi G, Smith T, et al. Long-term follow-up of
patients with complete remission following combination chemotherapy for
metastatic breast cancer. J Clin Oncol. 1996;14:2197-2205.
2. Giordano S, Buzdar A, Smith T, et al. Is breast cancer survival improving? Cancer. 2004;100(1):44-52.
3. Tevaarwerk A, Gray R, Schneider B, et al. Survival in Metastatic Breast
Cancer (MBC): No Evidence for Improved Survival Following Distant Recurrence after Adjuvant Chemotherapy. Presented at: San Antonio Breast
Cancer Symposium. 2011; San Antonio, TX.
4. Slamon D, et al. Use of chemotherapy plus a monoclonal antibody
against HER2 for metastatic breast cancer that overexpresses HER2. N Engl
J Med. 2001;344:783-792.
5. Cortes J, OShaughnessy J, Loesch D, et al. Eribulin monotherapy
versus treatment of physicians choice in patients with metastatic breast
cancer (EMBRACE): a phase 3 open-label randomised study. Lancet. 2011;
377:914-923.
6. AHern R, Smith I, Ebbs S. Chemotherapy and survival in advanced
breast cancer: The inclusion of doxorubicin in Cooper type regimens. Br J
Cancer. 1993;67:801-805.
7. Gibson L, Lawrence D, Dawson C, et al. Aromatase inhibitors for
treatment of advanced breast cancer in postmenopausal women. Cochrane
Database Syst Rev. 2009; CD003370.
8. Broglio K, Berry D. Detecting an overall survival benefit that is derived
from progression-free survival. J Natl Cancer Inst. 2009;101:1642-1649.
9. Pagani O, Senkus E, Wood W, et al. International guidelines for
management of metastatic breast cancer (MBC) from the European School of
Oncology (ESO)-MBC Task Force: Can metastatic breast cancer be cured?
J Natl Cancer Inst. 2010;102(7):456-463.
10. Cardoso F, Di Leo A, Lohrisch C, et al. Second and subsequent lines of
chemotherapy for metastatic breast cancer: What did we learn in the last two
decades? Ann Oncol. 2002;13:197-207.
11. Burzykowski T, Buyse M, Piccart-Gebhart M, et al. Evaluation of
tumor response, disease control, progression-free survival, and time to pro-
gression as potential surrogate end points in metastatic breast cancer. J Clin

Oncol. 2008;26:1987-1992.
12. Mauri D, Pavlidis N, Polyzos N, et al. Survival with aromatase
inhibitors and inactivators versus standard hormonal therapy in advanced
breast cancer: Meta-analysis. J Natl Cancer Inst. 2006;98:1285-1291.
13. Piccart-Gebhart M, Burzykowski T, Buyse M, et al. Taxanes alone or in
combination with anthracyclines as first-line therapy of patients with metastatic breast cancer. J Clin Oncol. 2008;26:1980-1986.
14. Cardoso F, Bedard P, Winer E, et al. International guidelines for
management of metastatic breast cancer: Combination vs. sequential singleagent chemotherapy. J Natl Cancer Inst. 2009;101:1174-1181.
15. Gennari A, Stockler M, Puntoni M, et al. Duration of chemotherapy for
metastatic breast cancer: A systematic review and meta-analysis of randomized clinical trials. J Clin Oncol. 2011;29:2144-2149.
16. Riemsma R, Forbes C, Kessels A, et al. Systematic review of aromatase
inhibitors in the first-line treatment for hormone sensitive advanced or
metastatic breast cancer. Breast Cancer Res Treat. 2010;123:9-24.
17. Chia S, Gradishar W, Mauriac L, et al. Double-blind, randomized
placebo controlled trial of fulvestrant compared with exemestane after prior
nonsteroidal aromatase inhibitor therapy in postmenopausal women with
hormone receptor-positive, advanced breast cancer: Results from EFECT.
J Clin Oncol. 2008;26:1664-1670.
18. Mehta R, Barlow W, Albain K, et al. A Phase III Randomized Trial of
Anastrozole Versus Anastrozole and Fulvestrant as First-Line Therapy for
Postmenopausal Women with Metastatic Breast Cancer: SWOG S0226.
Presented at: San Antonio Breast Cancer Symposium; 2011; San Antonio, TX.
19. Klijn J, Blarney R, Boccardo F, et al. Combination LHRH-agonist plus
tamoxifen treatment is superior to medical castration alone in premenopausal
metastatic breast cancer. Breast Cancer Res Treat. 1998;50:227.
20. Talbot D, Moiseyenko V, Van Belle S, et al. Randomised, phase II trial
comparing oral capecitabine (Xeloda) with paclitaxel in patients with metastatic/advanced breast cancer pretreated with anthracyclines. Br J Cancer.
2002;86:1367-1372.
21. Lin N, Claus E, Sohl J, et al. Sites of distant recurrence and clinical
37

outcomes in patients with metastatic triple-negative breast cancer: High
incidence of central nervous system metastates. Cancer. 2008;113:2638-2645.
22. Isakoff S, Goss P, Mayer E, et al. TBCRC009: A multicenter phase II
study of cisplatin or carboplatin for metastatic triple-negative breast cancer
and evaluation of p63/p73 as a biomarker of response. J Clin Oncol. 2011;29
(suppl; abstr 1025).
23. Miller K, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus
paclitaxel alone for metastatic breast cancer. N Engl J Med. 2007;357:26662676.
24. Miles D, Chan A, Dirix L, et al. Phase III study of bevacizumab plus
docetaxel compared with placebo plus docetaxel for the first-line treatment of
human epidermal growth factor receptor 2-negative metastatic breast cancer.
J Clin Oncol. 2010;28:3239-3247.
25. OShaughnessy J, Osborne C, Pippen J, et al. Iniparib plus chemotherapy in metastatic triple negative breast cancer. N Engl J Med. 2011;364:205214.
26. OShaughnessy J, Schwartzberg L, Danso M, et al. A randomized phase
III study of iniparib (BSI-201) in combination with gemcitabine/carboplatin
(G/C) in metastatic triple-negative breast cancer (TNBC). J Clin Oncol.
2011;29:(suppl; Abstr 1007).
27. Andersson M, Lidbrink E, Bjerre K, et al. Phase III randomized study
comparing docetaxel plus trastuzumab with vinorelbine plus trastuzumab as
first-line therapy of metastatic or locally advanced human epidermal growth
factor receptor 2-positive breast cancer: The HERNATA study. J Clin Oncol.
2011;29:264-271.
28. Von Minckwitz G, du Bois A, Schmidt M, et al. Trastuzumab beyond
progression in human epidermal growth factor receptor 2-positive advanced
breast cancer: A German Breast Group 26/Breast International Group 03-05
study. J Clin Oncol. 2009;27:1999-2006.
29. Geyer C, Forster J, Lindquis TD, et al. Lapatinib plus capecitabine for
HER2-positive advanced breast cancer. N Engl J Med. 2006;355:2733-2744.
30. Blackwell K, Burstein H, Storniolo A, et al. Randomized study of
Lapatinib alone or in combination with trastuzumab in women with ErbB2positive, trastuzumab-refractory metastatic breast cancer. J Clin Oncol.
2010;28:1124-1130.
31. Lin N, Dieras V, Paul D, et al. Multicenter phase II study of lapatinib
in patients with brain metastases from HER2 breast cancer. Clin Cancer
Res. 2009;15:1452-1459.
32. Perou C, Sorlie T, Eisen M, et al. Molecular portraits of human breast
tumours. Nature. 2000;406:747-752.
33. Metzker M. Sequencing technologies - the next generation. Nat Rev
Genet. 2010;11:31-46.
34. Baselga J, Cortes J, Kim S, et al. Pertuzumab plus trastuzumab plus
docetaxel for metastatic breast cancer. N Engl J Med. 2012;366:109-119.
35. Pedersen M, et al. Sym004: a novel synergistic anti-epidermal growth
factor receptor amtibody mixture with superior anticancer efficacy. Cancer
Res. 2010;70:588-597.
36. Baselga J, Swain S. Novel anticancer targets: Revisiting ERBB2 and
discovering ERBB3. Nat Rev Cancer. 2009;9:463-475.
37. Engelman JA, et al. MET amplification leads to gefitinib resistance in
lung cancer by activating ERBB3 signaling. Science. 2007;316:1039-1043.
38. Liu L, et al. Synergistic effects of foretinib with HER-targeted agents in
MET and HER1- or HER2-coactivated tumor cells. Mol. Cancer Ther. 2011;
10:518-530.
39. Turner N, Lambros M, Horlings H, et al. Integrative molecular profiling of triple negative breast cancers identifies amplicon drivers and potential
therapeutic targets. Oncogene. 2010;29:2013-2023.
40. Turner N, Pearson A, Sharpe R, et al. FGFR1 amplification drives
endocrine therapy resistance and is a therapeutic target in breast cancer.
Cancer Res. 2010;70:2085-2094.
38
41. Andre F, Bachelot T, Campone M, et al. A multicenter, open-label phase

II trial of dovitinib, an FGFR1 inhibitor, in FGFR1 amplified and nonamplified metastatic breast cancer. J Clin Oncol. 2011;29(suppl; abstr 508).
42. Bianchi G, Kocsis J, Dirix L, et al. Patient-Reported Outcomes From a
Randomized Phase II Study (TDM4450g/BO21976) of Trastuzumab
Emtansine vs Trastuzumab Plus Docetaxel in Previously Untreated Human
Epidermal Growth Factor Receptor 2-Positive Metastatic Breast Cancer.
Presented at: San Antonio Breast Cancer Symposium; 2011; San Antonio, TX.
43. Baselga J, Campone M, Piccart M, et al. Everolimus in Postmenopausal
Hormone-Receptor-Positive Advanced Breast Cancer. N Engl J Med. 2012;
366:520-529.
44. Yamnik R, Holz M. mTOR/S6K1 and MAPK/RSK signaling pathways
coordinately regulate estrogen receptor alpha serine 167 phosphorylation.
FEBS Lett. 2010;584:124-128.
45. Hanahan D, Weinberg R. Hallmarks of cancer: The next generation.
Cell. 2011;144:646-674.
46. Finn R, Boer K, Lang I, et al. A randomized phase II study of PD
0332991, cyclin-dependent kinase (CDK) 4/6 inhibitor, in combination with
letrozole for first-line treatment of patients with postmenopausal, estrogen
receptor (ER)-positive, human epidermal growth factor receptor 2 (HER2)negative advanced breast cancer. J Clin Oncol. 2011;29 (suppl; abstr
TPS100).
47. Neckers L, Workman P. Hsp90 Molecular Chaperone Inhibitors: Are
We There Yet? Clin Cancer Res. 2012;16:64-76.
48. Ni M, Chen Y, Lim E, et al. Targeting androgen receptor in estrogen
receptor-negative breast cancer. Cancer Cell. 2011;20:119-131.
49. Doane A, Danso M, Lal P, et al. An estrogen-receptor negative breast
cancer subset characterized by a hormonally regulated transcriptional program and response to androgen. Oncogene. 2006;25:3994-4008.
50. Al-Hajj M, Wicha M, Benito-Hernandez A, et al. Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A.
2003;100:3983-3988.
51. McDermott S, Wicha M. Targeting breast cancer stem cells. Mol Oncol.
2010;4:404-419.
52. Izrailit J, Reedijk M. Developmental pathways in breast cancer and
breast tumor-initiating cells: Therapeutic Implications. Cancer Lett. 2012;
317:115-126.
53. Stylianou S, Clarke R, Brennan K. Aberrant activation of notch
signaling in human breast cancer. Cancer Res. 2006;66:1517-1525.
54. Kasper M, Jaks V, Fiaschi M, et al. Hedgehog signaling in breast
cancer. Carcinogenesis. 2009;30:903-911.
55. Gupta P, Onder T, Jiang G, et al. Identification of selective inhibitors of
cancer stem cells by high-throughput screening. Cell. 2009;138:645-659.
56. Joyce J, Pollard J. Microenvironmental regulation of metastasis. Nature Rev Cancer. 2009;9:239-252.
57. Robert N, Dieras V, Glaspy J, et al. RIBBON-1: randomized, doubleblind, placebo-controlled, phase III trial of chemotherapy with or without
bevacizumab for first-line treatment of human epidermal growth factor
receptor 2-negative, locally recurrent or metastatic breast cancer. J Clin
Oncol. 2011;29:1252-1260.
58. Van der Veldt A, Lubberink M, Bahce I, et al. Rapid Decrease in
Delivery of Chemotherapy to Tumors after Anti-VEGF Therapy: Implications
for Scheduling of Anti-Angiogenic Drugs. Cancer Cell. 2012;21:82-91.
59. Ebos J, Lee C, Kerbel RS. Tumor and host-mediated pathways of
resistance and disease progression in response to antiangiogenic therapy.
Clin Cancer Res. 2009;15:5020-5025.
60. Ebos J, Lee C, Christensen J, et al. Multiple circulating proangiogenic
factors induced by sunitinib malate are tumor-independent and correlate with
antitumor efficacy. Proc Natl Acad Sci U S A. 2007;104:17069-17074.
A FRESH LOOK AT DUCTAL CARCINOMA IN SITU:

BASIC SCIENCE TO CLINICAL MANAGEMENT
CHAIR
Eun-Sil S. Hwang, MD, MPH
Durham, NC
SPEAKERS
Victoria Seewaldt, MD
Durham, NC
Beth A. Virnig, PhD
University of Minnesota, School of Public Health
Minneapolis, MN
Ductal Carcinoma In Situ: Challenges,

Opportunities, and Uncharted Waters
By Abigail W. Hoffman, MD, Catherine Ibarra-Drendall, PhD, Virginia Espina, MS,
Lance Liotta, MD, PhD, and Victoria Seewaldt, MD
Overview: Ductal carcinoma in situ (DCIS) is a heterogeneous group of diseases that differ in biology and clinical
behavior. Until 1980, DCIS represented less than 1% of all
breast cancer cases. With the increased utilization of mammography, DCIS now accounts for 15% to 25% of newly
diagnosed breast cancer cases in the United States. Although
our ability to detect DCIS has radically improved, our understanding of the pathophysiology and factors involved in its
progression to invasive carcinoma is still poorly defined. In
many patients, DCIS will never progress to invasive breast
cancer and these women are overtreated. In contrast, some
DCIS cases are clinically aggressive and the women may be
undertreated. We are able to define some of the predictors of
aggressive DCIS compared with DCIS of low malignant poten-
UCTAL CARCINOMA in situ is defined as a breast

lesion in which neoplastic mammary epithelial cells
are confined to the mammary ductal-lobular system without
invasion into the surrounding stroma. DCIS accounts for
approximately 15% to 25% of all female breast cancers. The
primary clinical goal in the management of DCIS is to
prevent the development of invasive breast cancer. In order
to individualize treatment of DCIS, there is a need for
biomarkers to prospectively identify DCIS with aggressive
biologic potential.
DCIS is highly variable in its clinical presentation, pathology, genetic/epigenetic alterations, and biologic potential.1,2
Although DCIS is a heterogeneous disease, until recently,
treatment of DCIS was relatively uniform. There is emerging evidence that DCIS has a wide range of biologic phenotypes; however, we lack biomarkers to definitively identify
DCIS that will progress to invasive disease. As a result,
there is considerable debate regarding how best to manage
patients with DCIS. The current primary management
options for women with DCIS include lumpectomy plus
radiation therapy; total mastectomy, with or without sentinel node biopsy, with or without reconstruction; or lumpectomy alone followed by clinical observation. Mastectomy is
an effective therapy for most patients; however, for many
women it represents overtreatment, particularly those with
minimal lesions identified by mammography. Randomized
clinical trials comparing breast-conserving surgery and radiation versus breast-conserving surgery alone show that
radiation therapy reduces local recurrence by approximately
50%.3,4 Conversely, it is possible that a subgroup of patients
with DCIS may not benefit from radiation following breastconserving surgery.
Women with biologically aggressive DCIS are at high risk
of local-regional recurrence or progression to invasive breast
cancer and are logically best served by mastectomy. However, it is also clear that there are women with DCIS at a low
risk for local-regional recurrence and/or progression who
might be adequately treated by breast-conserving surgery
alone, without radiation therapy. Emerging evidence suggests that there are also women whose DCIS is at such low
risk for recurrence or progression to invasive breast cancer
40
tial. However, our ability to risk-stratify DCIS is still in its

infancy. Clinical risk factors that predict aggressive disease
and increased risk of local recurrence include young age at
diagnosis, large lesion size, high nuclear grade, comedo
necrosis, and involved margins. Treatment factors such as
wider surgical margins and radiation therapy reduce the risk
of local recurrence. DCIS represents a key intermediate in the
stepwise progression to malignancy, but not all aggressive
breast cancers appear to have a DCIS intermediate, notably
within triple-negative breast cancer. Ongoing studies of the
genetic and epigenetic alterations in precancerous breast
lesions (atypia and DCIS) as well as the breast microenvironment are important for developing effective early detection
and individualized targeted prevention.
that they might be observed following a diagnostic biopsy

and undergo watchful waiting.5 However, at this time, we
cannot accurately predict the course of each patients DCIS.
Herein, we review our current understanding of 1) clinical
risk factors that predict local recurrence in patients with
DCIS treated with breast-conserving therapy, 2) evidence
for invasive breast cancers that potentially emerge from
intraepithelial neoplasia of lower grade than DCIS, and
3) molecular studies that aim to improve our understanding
of the biologic potential and diversity of DCIS lesions.
Clinical, Treatment, and Pathologic Risk
Factors That Predict Local Recurrence and
Invasion in Women with DCIS
Diagnosis and Prognosis
Until 1980, DCIS represented less than 1% of all breast

cancer cases. With the increased utilization of mammography during the 1980s, DCIS now accounts for 15% to 25% of
newly diagnosed breast cancer cases in the United States.6
Currently over 90% of DCIS is diagnosed by mammographic
examination alone; however, mammograms detect calcifications and frequently do not identify the neoplasm size
accurately.7 Precise preoperative evaluation is required to
determine the size and extent of the disease to optimize
surgical management. The recommended work-up and staging of DCIS includes a history and physical examination,
bilateral diagnostic mammography, pathology review, and
tumor estrogen receptor (ER) determination. Breast Magnetic Resonance Imaging (MRI) is increasingly being used
to evaluate DCIS; however, the exact use of breast MRI in
the preoperative management of DCIS remains a matter of
debate.8
With a follow-up period of 25 years, it is estimated that
14% to 60% of women with untreated DCIS will develop
From Duke University, Durham, NC; George Mason University, Manassas, VA.
Address reprint requests to Victoria Seewaldt, MD, Duke University, Box 2628, Room
221A MSRB, Durham, NC 27710; email: seewa001@mc.duke.edu.
1092-9118/10/1-10
DCIS: OPPORTUNITIES AND UNCHARTED WATERS
invasive breast cancer.1 In women who receive standard

treatment for DCIS, women with microinvasion have a
prognosis almost identical to women with DCIS alone, and
mastectomy is curative in 95% to 100% of women.9 Furthermore, approximately 50% of the local recurrences following
initial treatment for a pure DCIS is invasive cancer.1
Clinical and Treatment Risk Factors
Some of the clinical manifestations and treatment factors

associated with an increased risk of local recurrence following breast-conserving treatment for DCIS have been identified. Young women are at increased risk for local-regional
recurrence and invasive disease.6-10 Women with nonpalpable mammographically detected DCIS have a better prognosis than women with palpable DCIS, which is often
associated with microinvasion and occasionally with axillary
nodal involvement. Women who undergo mastectomy are at
low risk for local-regional recurrence. The use of radiation
therapy following breast-conserving surgery is associated
with approximately 50% reduction in the risk of local
recurrence.2-4 In the National Surgical Adjuvant Breast and
Bowel Project (NSABP) B-24 trial, the addition of tamoxifen
further reduced the risk of local-regional recurrence in
women treated with lumpectomy and radiation therapy.11
However, tamoxifen did not reduce the risk of local-regional
recurrence in a second randomized trial.4 Currently, tamoxifen treatment may be considered as a strategy to reduce the
risk of breast cancer recurrence in women with ER-positive
DCIS.1,5 The benefit of tamoxifen for ER-negative DCIS is
not known.
Prospective and retrospective analyses have identified
pathologic characteristics of DCIS that correlate with an
increased risk of local recurrence following breastconserving therapy. Pathologic features that are associated
with an increased risk of local-regional recurrence include
KEY POINTS
DCIS is a heterogeneous disease with a complex

array of biologic potential, clinical behavior, and
prognosis.
Although the biologic behavior of DCIS is heterogeneous, our clinical management has remained relatively uniform.
Biomarkers are needed to tailor clinical treatment of
DCIS to biologic potential for recurrence and invasion, but the genomic and pathologic nature of the
neoplasia that emerges during local recurrence compared with the original DCIS lesion is unknown.
Triple-negative breast cancers may not appear to
have a DCIS intermediate.
Numerous studies comparing the gene expression,
genetic, and epigenetic profiles of DCIS and invasive
breast carcinomas reported a high degree of similarity between the molecular alterations in the DCIS
and the invasive cancer in the same patient. This
provides evidence that the aggressive phenotype of
the breast cancer may be determined at the level of
the preinvasive lesion.
high nuclear grade, comedo necrosis, larger tumor size, and

involved margins of excision.1-3 The effect of pathologic
factors on the risk of local recurrence in patients with DCIS
varies with treatment as well as length of follow-up.10-13
Studies by Silverstein and colleagues provide evidence that
DCIS size, nuclear grade, and margin status are associated
with the risk of local-regional recurrence, but only in cases
with small surgical margin widths.12,13 For women with
DCIS who undergo surgical excision leaving margin widths
of 10 mm or more, the risk of local recurrence is unchanged
by size, grade, the presence of comedo necrosis, and radiation therapy.12 The effect of grade on local-regional recurrence of DCIS appears to be related to the length of followup.10 The NSABP B17 trial provides evidence that comedo
necrosis in DCIS is associated with an increased risk of
local-regional recurrence in women treated with excision
alone.3 Local recurrence rates in women undergoing excision
alone, after 8 years, was 40% for those with moderate or
marked comedo necrosis compared with 23% for women
without comedo necrosis. In women who underwent surgical
excision and radiation therapy, local recurrence rates were
similar for women with moderate or marked comedo necrosis relative to women without comedo necrosis (14% compared with 13%, respectively).3 In summary, these studies
suggest that there are multifaceted interactions between
pathologic factors and other elements that ultimately determine the risk of local recurrence.1-3,10-13
Evidence That Triple-Negative Breast Cancers May
Lack a DCIS Precursor
Triple-negative breast cancers are defined as tumors that

lack ER, PR, and HER2 expression. These tumors account
for 10% to 15% of all breast carcinomas, depending on the
thresholds used to define ER and PR positivity and the
methods used for HER2 assessment.14-16 The clinical interest in these tumors stems from the lack of targeted therapies
for patients affected by this breast cancer subtype, which is
more prevalent in younger women,14 black women,16 and
BRCA1 mutation carriers,17 and more aggressive than tumors of other molecular subtypes.14 Patients with triplenegative cancers also have a significantly shorter survival
following the first metastatic event when compared with
those with non-triple-negative controls (p 0.0001).14 A
predictive rather than prognostic classification system is
required for the success of targeted therapy, and therefore,
the origins and developmental mechanisms must also be
clearly defined.
The biologic developments of triple-negative breast cancers have debatable beginnings. Triple-negative breast cancers are thought to develop either from cells that are
triple-negative from the early phase of intraductal proliferation, leading to invasion, or cells that have lost hormone
receptor expression before invasion.18 A group of high-grade
DCIS lacking ER, PR ,and HER2, and expressing basal
markers has been identified.18 Bryan and colleagues studied
66 cases of high-grade DCIS and found a small proportion
(four cases, 6%) that exhibited the triple-negative phenotype
with increased invasion.18 In a study by Meijnen and colleagues, only eight out of 163 cases (5%) demonstrated
triple-negative lesions.19 In addition, tumors with BRCA1
germ-line mutations are significantly more likely to be
triple-negative breast cancers (p 0.005) and have early
high-grade DCIS (p 0.0045).17 However, it is important to
41
HOFFMAN ET AL
recognize that there are large discrepancies in the relative

frequency of triple-negative subtype between DCIS and
invasive disease. The prevalence of DCIS is substantially
lower than invasive triple-negative breast cancers, suggesting that the latter often lack an obvious in situ element.
Without supporting evidence or definitive conclusions, many
studies have attributed this discrepancy as a result of
triple-negative breast cancers progressing rapidly from
DCIS to invasive cancer or obliterating the DCIS precursor
from which they arose.17,18
Molecular Studies of the Biologic Potential and
Diversity of DCIS Lesions
Although the transition from DCIS to invasive cancer is

central to understanding the origins of breast cancer, little is
known about the time of onset or the triggering mechanisms
that promote invasion of DCIS. Some DCIS progresses to
invade the stroma, but other DCIS lies dormant. This raises
the question of whether there are subsets of precancerous
lesions that are preprogrammed to invade and metastasize
or whether the ability to invade and metastasize is acquired
late in the process of progression.20
Molecular Markers Predicting Invasion of DCIS
Numerous studies have compared the gene expression,

genetic, and epigenetic profiles of DCIS and invasive breast
carcinomas, showing a high degree of similarity between
DCIS and invasive cancer in the same patient.21 As expected, there is a significant difference in gene expression
between normal mammary epithelial cells and DCIS, but
surprisingly, DCIS and invasive breast cancer of the same
histologic subtype essentially share the same genetic/epigenetic alterations and gene expression patterns. In contrast,
the molecular profiles of breast tumors of distinct subtypes
(e.g., luminal, HER2) are significantly different. The expression and mutation status of numerous tumor suppressor
and oncogenes have been analyzed in DCIS and invasive
breast cancer including TP53, PTEN, PIK3CA, ERBB2,
MYC; differences in the frequency of these changes were
associated with tumor subtype, but none were associated
with invasion.20,22 For example, amplification of ERBB2 is
specific for the HER2 subtype; however, amplification of
ERBB2 is observed in both DCIS and invasive breast cancer.
The expression of several selected candidate genes based
on their biologic function has been analyzed in DCIS.23 Two
recent studies identified molecular markers that hold promise for identifying the risk of recurrence of DCIS.24,25 Gauthier and colleagues demonstrated that high expression of
COX-2 and Ki67 in DCIS correlates with higher risk of local
in situ and invasive recurrence.24 Lu and colleagues identified a molecular synergy between ERBB2 and 14 3-3 that
may increase the risk of invasive progression via epithelialto-mesenchymal transition regulation.25 A major limitation
of both of these studies was the use of small cohorts. Despite
the promise and need of DCIS risk biomarkers, prospective
validation is difficult. Validation of a predictive biomarker
may require a long waiting period to see if a patients DCIS
lesion will progress to invasive cancer. Most patients diagnosed with DCIS do not even want to experience a delay
before surgical therapy. Therefore, patients judged at low
risk may be unwilling to forego treatment. Since the population of DCIS and other preinvasive proliferative lesions is
42
heterogeneous in the same patient, the portion of the DCIS

lesion sampled at the initial biopsy diagnosis, which is used
for the predictive biomarker measurement, may not be the
same region of the lesion that has malignant potential.
There is very little data comparing the molecular genetic
characteristics of the cancers that recur to the original DCIS
lesion.
Evidence for Acquired Invasive Potential and
Intratumor Heterogeneity
The stepwise model of breast tumorigenesis assumes a

stepwise transition from epithelial hyperproliferation, to
atypical hyperplasia, to DCIS, and finally to invasive and
metastatic cancer.22,23 This progression model is supported
by human clinical and epidemiologic data and molecular
clonality studies addressing the relationships between in
situ and invasive regions of the same tumor and between
DCIS and its local invasive recurrence.21 There is emerging
evidence that neoplastic cells with invasive potential arise
frequently within DCIS lesions, but are held in check and
only invade when further molecular changes occur.20
Emerging data suggest a critical role for the microenvironment in promoting invasion of DCIS.
Numerous studies have shown that DCIS exhibits intratumoral heterogeneity. The prevailing model for explaining
this heterogeneity, the clonal evolution model, has recently
been challenged by proponents of the cancer stem cell
hypothesis.26 To investigate this issue, Kornelia Polyaks
group performed combined analyses of markers associated
with cellular differentiation states and genotypic alterations
in DCIS and invasive breast cancer and evaluated diversity
with ecological and evolutionary methods. Such studies
demonstrated a high degree of genetic heterogeneity both
within and between distinct tumor cell populations that
were defined based on markers of cellular phenotypes including stem cell-like characteristics.26 The degree of diversity correlated with clinically relevant breast tumor
subtypes.26 This supports the concept that molecular alterations at the level of DCIS may precede and determine the
breast tumor subtype.
There is emerging evidence that adaptation to hypoxic
and metabolic stress promotes progression of DCIS to invasive breast cancer. It is hypothesized that premalignant and
malignant cells down-regulate proteins that induce apoptosis or senescence and upregulate pro-survival pathways that
protect against hypoxia, DNA damage, metabolic and genotoxic stress.20 Nevertheless, even if a cell can resist programmed cell death or senescence it will still not survive in
a hypoxic, nutrient-deprived environment unless it can find
alternative sources of energy for cellular functions, such as
through autophagy, anaerobic respiration, or increasing the
efficiency of aerobic respiration.20 To appreciate how DCIS
might progress and circumvent stress-induced death or
senescence, and use alternative sources of energy, it is
important to consider the stresses that affect DCIS cells and
molecular signaling pathways that may act to protect
against these stressors.20 Autophagy promotes survival in
the face of hypoxic and nutrient stress and is an emerging
target for cancer therapy.20 Autophagy is upregulated and
colocalizes with areas of hypoxic stress, and immortalized
mammary epithelial cells are more susceptible to cell death
under conditions of metabolic stress.20 DCIS growth or the
growth of any intraductal proliferative lesion is limited
DCIS: OPPORTUNITIES AND UNCHARTED WATERS
because of confinement within the duct and the absence of

a blood supply. Espina and colleagues proposed that autophagy is a major survival mechanism that is used by DCIS
cells to persist and proliferate in the high-stress environment of the intraductal space and can be a main determinant of acquired DCIS cell fate in response to metabolic
stress.20
Evidence for Preprogramming of Precancerous Breast Lesions
Triple-negative breast cancers frequently exhibit aggressive clinical behavior and are many times metastatic at
diagnosis. In addition to being poorly differentiated, as
outlined above, triple-negative breast cancers are thought to
have a lower proportion of associated DCIS compared with
other types of breast carcinoma. These observations challenge whether the stepwise model of mammary carcinogenesis adequately models initiation and progression of triplenegative breast cancer. In contrast to the stepwise model,
aggressive cancers may emerge from low-grade DCIS or
atypical proliferative lesions. There is growing recognition
that the activated (e.g., phosphorylated) state of cellular
protein signaling networks can play a key role in breast
cancer initiation and progression27 and provide information
about the functional state of the cancer cell that cannot be
accurately predicted based on genomic sequencing alone.
Likewise, it has been recently shown that single biomarkers
are not adequate to capture the complex changes in signaling networks activated during breast cancer initiation.28
Ibarra-Drendall et al. are currently testing the hypothesis
that mammary atypia from high-risk women exhibits activation of phospho-protein signaling networks activated in
aggressive triple-negative breast cancer.29 We performed
Reverse-Phase Protein Microarray (RPMA) profiling of cyto-
logic atypia obtained prospectively from unaffected highrisk women in our cohort.30,31 We observed coactivation of
Akt/mTOR/insulin- and IL6/Stat3/vimentin-network signaling in cytologic atypia and found a statistically significant
association between body mass index of 30 kg/m2 and
vimentin expression (p 0.028).30,31 Limited immunohistochemistry studies demonstrated vimentin expression in
atypical mammary epithelial cells,31 which in turn correlates with activation of Stat3 and IL6 levels in patientmatched mammary fluids.32 Both Akt/mTOR and IL6/Stat3
signaling have been implicated in the aggressive biology of
triple-negative breast cancer and epithelial-to-mesenchymal
transition. These studies provide preliminary evidence that
the biologically aggressive atypia may be present before the
development of an invasive triple-negative breast cancer.
Conclusion
Currently, there are many unanswered questions regarding the diagnosis and management of DCIS. Despite the
complexity and heterogeneous nature of DCIS, all women
are treated with a one size fits all scenario that includes
mastectomy versus breast-conserving surgery with radiation therapy. Most recently, Oncotype DX Breast Cancer
Assay for DCIS33 has been developed and is being utilized to
obtain an estimate of 10-year risk of local recurrence, with
the hope of providing guidance to treatment decision. But
the success of this multigene assay in predicting the fate
of DCIS remains to be seen. An essential goal of future
research should focus on the development of accurate riskstratification methods based on a comprehensive understanding of the biologic, pathologic, radiologic, and clinical
factors associated with DCIS.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Abigail W. Hoffman*
Catherine Ibarra-Drendall*
Virginia Espina
Theranostics
Health
Lance Liotta*
Victoria Seewaldt*
REFERENCES
1. Burstein HJ, Polyak K, Wong JS, et al. Ductal carcinoma in situ of the
breast. N Engl J Med. 2004;350:1430-1441.
2. Schnitt SJ. Local outcomes in ductal carcinoma in situ based on patient
and tumor characteristics. J Natl Cancer Inst Monogr. 2010;41:158-161.
3. Fisher ER, Dignam J, Tan-Chiu E, et al. Pathologic findings from the
National Surgical Adjuvant Breast Project (NSABP) eight-year update of
Protocol B-17: intraductal carcinoma. Cancer. 1999;86:429-438.
4. Houghton J, George WD, Cuzick J, et al. Radiotherapy and tamoxifen in
women with completely excised ductal carcinoma in situ of the breast in the
UK, Australia, and New Zealand: randomised controlled trial. Lancet. 2003;
362:95-102.
5. Chen YY, DeVries S, Anderson J, et al. Pathologic and biologic response
to preoperative endocrine therapy in patients with ER-positive ductal carcinoma in situ. BMC Cancer. 2009;9:285.
6. Brinton LA, Sherman ME, Carreon JD, et al. Recent trends in breast
cancer among younger women in the United States. J Natl Cancer Inst.
2008;100:1643-1648.
7. Evans A, Pinder S, Wilson R, et al. Ductal carcinoma in situ of the
breast: correlation between mammographic and pathologic findings. AJR

Am J Roentgenol. 1994;162:1307-1311.
8. Kuhl CK, Schrading S, Bieling HB, et al. MRI for diagnosis of pure
carcinoma in situ: a prospective observational study. Lancet. 2007;370:485492.
9. Silver SA, Tavassoli FA. Mammary ductal carcinoma in situ with
microinvasion. Cancer. 1998;82:2382-2390.
10. Solin LJ, Fourquet A, Vicini FA, et al. Long-term outcome after
breast-conservation treatment with radiation for mammographically detected
ductal carcinoma in situ of the breast. Cancer. 2005;103:1137-1146.
11. Fisher B, Dignam J, Wolmark N, et al. Tamoxifen in treatment of
intraductal breast cancer: National Surgical Adjuvant Breast and Bowel
Project B-24 randomised controlled trial. Lancet. 1999;353:1993-2000.
12. Silverstein MJ, Lagios MD, Craig PH, et al. A prognostic index for
ductal carcinoma in situ of the breast. Cancer. 1996;77:2267-2274.
13. Silverstein MJ, Lagios MD, Groshen S, et al. The influence of margin
width on local control of ductal carcinoma in situ of the breast. N Engl J Med.
1999;340:1455-1461.
43
HOFFMAN ET AL
14. Dent R, Trudeau M, Pritchard KI, et al. Triple-negative breast cancer:
clinical features and patterns of recurrence. Clin Cancer Res. 2007;13:44294434.
15. Haffty BG, Yang Q, Reiss M, et al. Locoregional relapse and distant
metastasis in conservatively managed triple negative early-stage breast
16. Harris LN, Broadwater G, Lin NU, et al. Molecular subtypes of breast
cancer in relation to paclitaxel response and outcomes in women with
metastatic disease: results from CALGB 9342. Breast Cancer Res. 2006;8:R66.
17. Arun B, Vogel KJ, Lopez A, et al. High prevalence of preinvasive lesions
adjacent to BRCA1/2-associated breast cancers. Cancer Prev Res. 2009;2:122127.
18. Bryan BB, Schnitt SJ, Collins LC. Ductal carcinoma in situ with
basal-like phenotype: a possible precursor to invasive basal-like breast
cancer. Mod Pathol. 2006;19:617-621.
19. Meijnen P, Peterse JL, Antonini N, et al. Immunohistochemical categorisation of ductal carcinoma in situ of the breast. Br J Cancer. 2008;98:137142.
20. Espina V, and Liota LA. What is the malignant nature of human ductal
carcinoma in situ? Nature Reviews Cancer. 2011;11:68-75.
21. Polyak K. Molecular markers for the diagnosis and management of
ductal carcinoma in situ. J Natl Cancer Inst Monogr. 2010;41:210-213.
22. Sgroi DC. Preinvasive breast cancer. Annu Rev Pathol. 2010;5:193-221.
23. Kuerer HM, Albarracin CT, Yang WT, et al. Ductal carcinoma in situ:
state of the science and roadmap to advance the field. J Clin Oncol.
2009;27:279-288.
24. Gauthier ML, Berman HK, Miller C, et al. Abrogated response to
44
cellular stress identifies DCIS associated with subsequent tumor events and
defines basal-like breast tumors. Cancer Cell. 2007;12:479-491.
25. Lu J, Guo H, Treekitkarnmongkol W, et al. 14-3-3zeta cooperates with
ErbB2 to promote ductal carcinoma in situ progression to invasive breast
cancer by inducing epithelial-mesenchymal transition. Cancer Cell. 2009;16:
195-207.
26. Polyak K. Breast cancers: origins and evolution. J Clin Invest. 2007;
117:3155-3163.
27. Sachs K, Perez O, Peer D, et al. Causal protein-signaling networks
derived from multiparameter single-cell data. Science. 2005;308:523-529.
28. Chen FL, Xia W, and Spector NL. Acquired resistance to small molecule
ErbB2 tyrosine kinase inhibitors. Clin Cancer Res. 2008;14:6730-6734.
29. Ibarra-Drendall C, Seewaldt V. Evidence for the Warburg Effect in
Mammary Atypia from High-Risk Women. Oral presentation at: Gordon
Conference in Mammary Gland Biology; June 2011; Newport, RI.
30. Ibarra-Drendall C, Troch MM, Barry WT, et al. Pilot and feasibility
study: prospective proteomic profiling of mammary epithelial cells from
high-risk women provides evidence of activation of pro-survival pathways.
Breast Cancer Res Treat. Epub 2011 Jun 7.
31. Pilie PG, Ibarra-Drendall C, Troch MM, et al. Protein microarray
analysis of mammary epithelial cells from obese and nonobese women at high
risk for breast cancer: Feasibility data. Cancer Epidemiol Biomarkers Prev.
2011 20:476-482.
32. Seewaldt V. Evidence for the Warburg Effect in Mammary Atypia from
High-Risk Women. Oral presentation at: Gordon Conference in Mammary
Gland Biology; June 2011; Newport, RI.
33. Oncotype DX Breast Cancer Assay for DCIS (www.oncotypedx.com).
Accessed March 15, 2012.
D u c t a l C a r c i n o m a I n S i t u , a n d t h e I n fl u e n c e
of the Mode of Detection, Population
Characteristics, and Other Risk Factors
By Beth A. Virnig, PhD, MPH, Shi-Yi Wang, MD, MS, and Todd M. Tuttle, MD, MS
Overview: Approximately 25% of breast cancers in the United

States are diagnosed as ductal carcinoma in situ (DCIS). Rates
of DCIS have risen from 5.8 per 100,000 women in the 1970s to
32.5 per 100,000 in 2004. This pattern is generally attributed to
increased use of screening mammography. DCIS is a major
risk factor for invasive breast cancer, and considerable controversy remains about whether DCIS should be considered a
direct precursor of invasive breast cancer. There is, however,
a general consensus that DCIS represents an intermediate
step between normal breast tissue and invasive breast cancer.
Although the majority of major risk factors are similar for DCIS
and invasive breast cancer, prognostic factors including estrogen and progesterone receptor status and HER2 positivity
UCTAL CARCINOMA in situ (DCIS) is noninvasive

breast cancer that encompasses a wide spectrum of
diseases ranging from low-grade lesions that are not life
threatening to high-grade lesions that may harbor foci of
invasive breast cancer. DCIS is typically classified according
to architectural pattern (solid, cribriform, papillary, and
micropapillary), tumor grade (high, intermediate, and low),
and the presence or absence of comedo necrosis. Approximately 25% of breast cancers diagnosed in the United States
are DCIS.1
The fundamental question underlying treatment and detection of DCIS is whether it should be considered a direct
precursor of invasive breast cancer. There is general consensus that DCIS represents an intermediate step between
normal breast tissue and invasive breast cancer. However,
because of general treatment patterns and the recognition
that even a biopsy may remove an important portion of the
tumor, the natural history of untreated DCIS is unknown
and likely is unknowable. This article provides updates on a
report requested by the National Institutes of Health Office
of Medical Applications of Research, which is available at
www.ahrq.gov//clinic/epcix.htm and summarized by Virnig.2
The report addresses key questions about the incidence of
DCIS and the associated risk factors, the effect of magnetic
resonance imaging (MRI) and sentinel lymph node biopsy
(SLNB) on patient outcomes, and the effect of tumor and
patient characteristics, surgery, radiation, and systemic
treatment on outcomes.
The incidence of DCIS has increased dramatically since
the early 1970s from 5.8 per 100,000 in 1975 to 32.5 per
100,000 in 2004. Yet, the rate of DCIS is considerably less
than the invasive breast cancer incidence, estimated to be
124.3 per 100,000 in 2004.1 The incidence of the most
aggressive subtype of DCIS comedo has not increased as
rapidly as the less aggressive noncomedo form. It is not clear
whether these patterns reflect screening picking up more
noncomedo tumors or pathologists using the comedo classification less often.
The increased risk of invasive breast cancer associated
with hormone replacement therapy (HRT) with estrogen
plus progestin is well established. The reported association
between HRT and DCIS is inconsistent across studies.3-5
are less well studied but look to have similar value in both
cases. The use of postdiagnostic MRI, sentinel lymph node
biopsy, surgery, radiation, and endocrine therapy are all
evolving as evidence from randomized and observational
studies continues to accumulate. Treatment of DCIS requires
a balance between risk of overtreatment and undertreatment.
Ongoing studies are focusing on whether partial-breast irradiation is as effective as whole-breast irradiation and whether
treatment with endocrine therapies can reduce the likelihood
of either invasive breast cancer or DCIS recurrence. In general, treatment decisions should take into account the likelihood that an apparent case of DCIS could harbor foci of
invasive disease.
The strongest evidence that the increased incidence of

DCIS can be attributed to the use of screening mammography comes from eight population-based trials of mammography screening. Mammographic screening consistently was
more likely to lead to the diagnosis of invasive breast cancer
than of DCISall trials reported that less than 20% of
screen-detected breast cancers were DCIS. There is considerable evidence that the detection of DCIS is greatest at
baseline screening. For example, the breast cancer surveillance consortium reported DCIS incidence of 150 per 1,000
screening mammograms on first screening and incidence of
0.83 per 1,000 for subsequent screening mammograms.6
Although several trials have assessed the value of tamoxifen or raloxifene for preventing DCIS, the trials, in reality,
were designed to assess the value of the agents for preventing invasive breast cancer, with DCIS as a secondary outcome. Two large, controlled trials showed that tamoxifen
had a protective role on the development of DCIS (and
invasive breast cancer), though the magnitude of effects
varied somewhat.7,8 The Study of Tamoxifen and Raloxifene
(STAR) trial found 40% higher incidence of DCIS for women
treated with raloxifene than with tamoxifen. The study also
found both treatments decreased the risk of invasive breast
cancer by half. The women randomly assigned to raloxifene
had 36% fewer uterine cancers and 29% fewer blood clots
than the women assigned to tamoxifen.9
The Continuing Outcomes Relevant to Evista (CORE)/
Multiple Outcomes of Raloxifene Evaluation (MORE) study
found significantly reduced incidence of invasive breast
cancer associated with raloxifene (hazard ratio [HR] 0.41),
but a nonsignificant increase in the incidence of DCIS
among the treated women (HR 1.78).10 The inconsistent
effect of raloxifene and tamoxifen on DCIS and invasive
From the University of Minnesota School of Public Health and University of Minnesota
Medical School, Minneapolis, MN.
Address reprint requests to Beth A. Virnig, PhD, MPH, University of Minnesota School of
Public Health, 420 Delaware Street SE, MMC 729, Minneapolis, MN 55455; email:
virni001@umn.edu.
1092-9118/10/1-10
45
VIRNIG, WANG, AND TUTTLE
breast cancer incidence deserves further investigation and

may, ultimately, shed light on the biology of DCIS and
invasive breast cancer and the factors that control invasive
progression.
The use of breast MRI for patients with DCIS is not yet
established. MRI can influence treatment recommendations
for some patients by identifying occult disease not visualized
with mammography. Among patients with DCIS, three
studies found that the sensitivity of detecting multicentric
disease is higher with MRI compared with mammography.11,12 Breast MRI can potentially influence treatment
decisions by providing more accurate information on the size
and extent of the known DCIS. Such findings may determine
the choice of breast-conserving surgery (BCS) compared
with mastectomy or the width of excision margins. Studies
have been mixed when comparing whether MRI overestimates or underestimates tumor size relative to mammography. The potential benefits of MRI include fewer re-excisions
after BCS and decreased local recurrence rates after excision. However, no studies to date have reported that MRI
yields these improved patient outcomes. Breast MRI may
also have potential disadvantages such as increased patient
anxiety, costs, and unnecessary use of breast biopsy; for
example, one study of MRI for women with DCIS reported
that 47.6% of the biopsies stemming from a positive MRI
were negative.13 If MRI leads to overestimation of the extent
and size of DCIS in some patients, it points to MRI use
resulting in more mastectomies and, for BCS, wider excisions and their associated less favorable cosmetic outcomes.
SLNB is recommended for patients with invasive breast
cancer to determine prognosis and to guide adjuvant treatment decisions. The risk of SLN metastasis is higher for
patients with a final diagnosis of DCIS with microinvasion
compared with pure DCIS (9.3% vs. 4.8%).14 In addition,
approximately 15% of patients who are initially diagnosed
with DCIS on core needle biopsy have invasive breast cancer
identified in the excision or mastectomy specimen.15 Thus,
some patients may require axillary lymph node staging
after definitive surgical treatment for presumed DCIS. Although SLNB is feasible for most patients after excision,
it is not feasible after mastectomy.16 Thus, some authors
recommend routine SLNB for women with high-risk DCIS
(palpable mass, comedo necrosis) and for those patients
undergoing mastectomy.17
The 10-year breast cancer mortality rate from DCIS is less
KEY POINTS
46
Most of the risk factors for DCIS and invasive breast

cancer are similar.
Rates of DCIS have risen and are associated with
rising use of mammography.
The value of breast MRI for women diagnosed with
DCIS is not clear.
Sentinel lymph node biopsy is thought to be most
valuable for women with high-risk disease or undergoing mastectomy.
Treatment of DCIS includes surgery, radiation, and
endocrine therapy. Studies of the optimal treatment
are ongoing.
18
than 2%. Therefore, the primary outcomes for DCIS are

ipsilateral and contralateral breast cancer recurrence. Many
of the prognostic factors are shared between DCIS and
invasive cancer. The local and recurrence rates for DCIS are
between 10% and 24% after 10 years. Younger age at
diagnosis is a consistent adverse prognostic factor. Women
older than 40 or 50 years consistently have reduced risk of
DCIS or invasive recurrence than younger women. This increased risk may reflect the observation that DCIS in younger
women is more often symptomatic and more extensive.
Studies of racial differences in DCIS recurrence point to a
somewhat complex story. Studies of single treatments that
only adjust for demographic factors alone,19,20 report that
black women with DCIS are more likely than white women
to die from breast cancer (response rate [RR] 1.35) or
experience an invasive recurrence (RR 1.4). However, the
studies that adjust for a more detailed set of tumor factors
find no difference between racial groups and risk of DCIS or
invasive recurrence (RR 1.12). This suggests that there
may be differences in the tumor characteristics between
black and white women. There also may be systematic
differences in aggressiveness of DCIS treatment by race.
Positive surgical margins are consistently associated with
increased DCIS and invasive breast cancer recurrence, although the magnitude of excess risk varies considerably.21
There is considerable debate regarding whether width of a
negative margin (width of a margin negative for tumor cells)
is associated with a decreased risk of recurrence. In general,
larger tumors were associated with higher rates of local
DCIS and invasive recurrence than smaller tumors.18,20
Although somewhat inconsistently labeled, a higher pathologic or nuclear grade (grade 3) was consistently associated
with a higher probability of local DCIS or invasive recurrence than an intermediate or low grade (grade 2 or 1).
Comedo necrosis, a factor unique to DCIS, is strongly and
consistently associated with poorer outcomes and increased
risk of DCIS or invasive recurrence.20
Few of the important markers of tumor aggressiveness in
invasive breast cancer are well studied in DCIS. Rates of
estrogen receptor (ER) testing for women with DCIS are
rising but still lag behind testing for invasive cancer. However, rates of ER positivity as a percentage of tumors tested
are similar for DCIS and invasive breast cancer with 81% to
85% positivity. ER positivity has been linked with a decreased risk of recurrence in several small studies. DCIS
is rarely tested for HER2 positivity, but, nonetheless, several small studies have linked it to increased risk of recurrence (RR 1.53.7) and reduced time to recurrence.22 The
possibility of treating HER2-positive tumors is being studied
in ongoing trials. Ongoing research is also focusing on
developing genetic markers of DCIS that has more or less
favorable biology.
In randomized trials including NSABP-17 and the European Organization for Research and Treatment of Cancer
(EORTC) randomized phase III trial 10853, whole-breast
radiation therapy (RT) following BCS was associated with
reduced local DCIS or invasive carcinoma recurrence. Although statistically significant, the number of events prevented per 1,000 treated women was less than 10%. Despite
the reduced recurrence, RT had no effect on either breast
cancer mortality or total mortality.23-25 Neither randomized
nor observational studies pointed to compelling evidence
that BCS plus radiation has differing relative effectiveness
DCIS AND INFLUENCE OF RISK FACTORS
in the presence or absence of adverse prognostic factors. This

lack of differential effect can be seen for the most important
prognostic factors, including tumor grade and size and
comedo necrosis.18 The key issue then becomes whether the
absolute benefit of RT for low-risk women is small enough to
justify use of BCS alone.
Although outcomes between mastectomy and BCS or BCS
plus RT were not studied in a randomized fashion, several
observational studies reported that women undergoing mastectomy were less likely than women undergoing BCS or
BCS plus RT to experience local DCIS or invasive recurrence.26,27 We found no study showing a mortality reduction
associated with mastectomy over BCS with or without
radiation.
The literature is still evolving regarding the use of accelerated partial-breast irradiation (APBI) delivered via MammoSite or balloon-based brachytherapy28-30 and whether
it is associated with similar levels of control as whole-breast
irradiation.
The NSABP-24 assessed the value of tamoxifen following
DCIS diagnosis and found tamoxifen was associated with
statistically significant reductions in local recurrence (RR
0.60) and contralateral disease (RR 0.56). However, the
absolute risk reduction in ipsilateral and contralateral disease was small. Ongoing studies such as the NSABP-37 are
examining the comparative effectiveness of tamoxifen and
aromatase inhibitors, and the NSABP B-43 is examining use
of trastuzumab for women with HER2-positive tumors.
DCIS and is part of an active policy discussion related to

invasive breast cancer. There is also an aspect of underdiagnosis that must be considered. In some instances, DCIS
may be underdiagnosed invasive cancer for which the pathology sections simply missed the invasive area. Overall,
the arguments for a close relationship between in situ and
invasive breast cancer can be found in the similarity of risk
factors for both the incidence of the diseases and their
similar responses to treatment.
From a clinical perspective, it seems prudent to approach
DCIS and invasive breast cancer as being related. Given the
rate of sampling error in needle biopsies, presumptive DCIS
should be treated as potential invasive cancer until a more
definitive pathologic sample is available. In clinical settings,
efforts should be made to make full use of markers such as
estrogen and progesterone receptor status, and necrosis to
differentiate women at high risk from those at lower risk of
developing invasive disease. Effort should be undertaken to
evaluate whether HER2 status offers any promise for clinical decision-making. Ultimately, these markers would allow
for focusing aggressive treatment on those who have the
greatest probability of benefit.
Note: Because of space limitations, the reference list is
necessarily incomplete. See the article by Virnig and colleagues2 for a more complete review.
Discussion
Funding: Agency for Healthcare Research and Quality, US

Department of Health and Human Services (contract number
290-02-10064-I). The authors of this report are responsible for
its content. Statements in the report should not be construed as
endorsement by the Agency for Healthcare Research and Quality or the U.S. Department of Health and Human Services.
It is clear that many cases of DCIS either would not

develop into invasive disease or would do so much later in
life, perhaps never becoming clinically relevant. This issue
of overdiagnosed DCIS because of screening is not limited to
ACKNOWLEDGMENTS
Author
Beth A. Virnig*
Shi-Yi Wang*
Todd M. Tuttle*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Brinton LA, Sherman ME, Carreon JD, Anderson WF. Recent trends in
breast cancer among younger women in the United States. J Natl Cancer Inst.
2008;100:1643-1648.
2. Virnig BA, Tuttle TM, Shamliyan T, Kane RL. Ductal carcinoma in situ
of the breast: A systematic review of incidence, treatment, and outcomes.
3. Chlebowski RT, Hendrix SL, Langer RD, et al. Influence of estrogen plus
progestin on breast cancer and mammography in healthy postmenopausal
women: The Womens Health Initiative Randomized Trial. JAMA. 2003;289:
3243-3253.
4. Collaborative Group on Hormonal Factors in Breast Cancer. Breast
cancer and hormone replacement therapy: Collaborative reanalysis of data
from 51 epidemiological studies of 52,705 women with breast cancer and
108,411 women without breast cancer. Lancet. 1997;350:1047-1059.
5. Reeves GK, Pirie K, Green J, et al. For the Million Women Study
Collaborators. Comparison of the effects of genetic and environmental risk
factors on in situ and invasive ductal breast cancer. Int J Cancer. Epub ahead
2011 Sept 27.
6. Ernster VL, Ballard-Barbash R, Barlow WE, et al. Detection of ductal
carcinoma in situ in women undergoing screening mammography. J Natl
Cancer Inst. 2002;94:1546-1554.
7. Cuzick J, Forbes JF, Sestak I, et al. Long-term results of tamoxifen

prophylaxis for breast cancer96-month follow-up of the randomized IBIS-I
trial. J Natl Cancer Inst. 2007;99:272-282.
8. Powles TJ, Ashley S, Tidy A, et al. Twenty-year follow-up of the Royal
Marsden randomized, double-blinded tamoxifen breast cancer prevention
trial. J Natl Cancer Inst. 2007;99:283-290.
9. Vogel VG, Costantino JP, Wickerham DL, et al. Effects of tamoxifen vs
raloxifene on the risk of developing invasive breast cancer and other disease
outcomes: The NSABP Study of Tamoxifen and Raloxifene (STAR) P-2 Trial.
JAMA. 2006;295:2727-2741.
10. Martino S, Cauley JA, Barrett-Connor E, et al. Continuing outcomes
relevant to Evista: Breast cancer incidence in postmenopausal osteoporotic
women in a randomized trial of raloxifene. J Natl Cancer Inst. 2004;96:17511761.
11. Hwang ES, Kinkel K, Esserman LJ, et al. Magnetic resonance imaging
in patients diagnosed with ductal carcinoma-in-situ: Value in the diagnosis of
residual disease, occult invasion, and multicentricity. Ann Surg Oncol.
2003;10:381-388.
12. Menell JH, Morris EA, Dershaw DD, et al. Determination of the
presence and extent of pure ductal carcinoma in situ by mammography and
magnetic resonance imaging. Breast J. 2005;11:382-390.
47
VIRNIG, WANG, AND TUTTLE

13. Allen LR, Lago-Toro CE, Hughes JH, et al. Is there a role for MRI in the
pre-operative assessment of patients with DCIS? Ann Surg Oncol. 2010;17:
2395-2400.
14. Katz A, Gage I, Evans S, et al. Sentinel lymph node positivity of
patients with ductal carcinoma in situ or microinvasive breast cancer. Am J
Surg. 2006;191:761-766.
15. Bruening W, Fontarosa J, Tipton K, et al. Systematic review: Comparative effectiveness of core-needle and open surgical biopsy to diagnose breast
lesions. Ann Intern Med. Epub 2009 Dec 15.
16. Wong SL, Edwards MJ, Chao C, et al. The effect of prior breast biopsy
method and concurrent definitive breast procedure on success and accuracy of
sentinel lymph node biopsy. Ann Surg Oncol. 2002;9:272-277.
17. Schneider C, Trocha S, McKinley B, et al. The use of sentinel lymph
node biopsy in ductal carcinoma in situ. Am Surg. 2010;76:943-946.
18. Fisher ER, Dignam J, Tan-Chiu E, et al. Pathologic findings from the
National Surgical Adjuvant Breast Project (NSABP) eight-year update of
Protocol B-17: intraductal carcinoma. Cancer. 1999;86:429-438.
19. Joslyn SA. Ductal carcinoma in situ: Trends in geographic, temporal,
and demographic patterns of care and survival. Breast J. 2006;12:20-27.
20. Li CI, Malone KE, Saltzman BS, Daling JR. Risk of invasive breast
carcinoma among women diagnosed with ductal carcinoma in situ and lobular
carcinoma in situ, 1988-2001. Cancer. 2006;106:2104-2112.
21. Dick AW, Sorbero MS, Ahrendt GM, et al. Comparative effectiveness of
ductal carcinoma in situ management and the roles of margins and surgeons.
22. Holmes P, Lloyd J, Chervoneva I, et al. Prognostic markers and
long-term outcomes in ductal carcinoma in situ of the breast treated with
excision alone. Cancer. 2011;117:3650-3657.
23. Bijker N, Meijnen P, Peterse JL, et al. Breast-conserving treatment
with or without radiotherapy in ductal carcinoma-in-situ: Ten-year results of
48
European Organisation for Research and Treatment of Cancer randomized

phase III trial 10853a study by the EORTC Breast Cancer Cooperative
Group and EORTC Radiotherapy Group. J Clin Oncol. 2006;24:3381-3387.
24. Holmberg L, Garmo H, Granstrand B, et al. Absolute risk reductions for
local recurrence after postoperative radiotherapy after sector resection for
ductal carcinoma in situ of the breast. J Clin Oncol. 2008;26:1247-1252.
25. Wapnir IL, Dignam JJ, Fisher B, et al. Long-term outcomes of invasive
ipsilateral breast tumor recurrences after lumpectomy in NSABP B-17 and B-24
randomized clinical trials for DCIS. J Natl Cancer Inst. 2011;103:478-488.
26. Meijnen P, Oldenburg HS, Peterse JL, et al. Clinical outcome after
selective treatment of patients diagnosed with ductal carcinoma in situ of the
breast. Ann Surg Oncol. 2008;15:235-243.
27. Schouten van der Velden AP, van Vugt R, Van Dijck JA,et al. Local
recurrences after different treatment strategies for ductal carcinoma in situ of
the breast: A population-based study in the East Netherlands. Int J Radiat
Oncol Biol Phys. 2007;69:703-710.
28. Jeruss JS, Kuerer HM, Beitsch PD, et al. Update on DCIS outcomes
from the American Society of Breast Surgeons accelerated partial breast
irradiation registry trial. Ann Surg Oncol. 2011;18:65-71.
29. Israel PZ, Vicini F, Robbins AB, et al. Ductal carcinoma in situ of the
breast treated with accelerated partial breast irradiation using balloon-based
brachytherapy. Ann Surg Oncol. 2010;17:2940-2944.
30. Keisch M, Vicini F, Beitsch P, et al. American Society of Breast
Surgeons MammoSite Radiation Therapy System Registry Trial: Ductal
carcinoma-in-situ subset analysis-4-year data in 194 treated lesions. Am J
Surg. 2009;198:505-507.
31. Fisher B, Land S, Mamounas E, et al. Prevention of invasive breast
cancer in women with ductal carcinoma in situ: An update of the national
surgical adjuvant breast and bowel project experience. Semin Oncol. 2001;28:
400-418.
KEY QUESTIONS IN THE LOCO-REGIONAL

TREATMENT OF BREAST CANCER
CHAIR
Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University
Baltimore, MD
SPEAKERS
Thomas A. Buchholz, MD
University of Texas M. D. Anderson Cancer Center
Houston, TX
Monica Morrow, MD
Memorial Sloan-Kettering Cancer Center
New York, NY
Postmastectomy Radiation and Partial

Breast Irradiation
By Richard C. Zellars, MD
Overview: Between 1997 and 1999, three modern postmastectomy radiation therapy (PMXRT) trials were published.
These trials showed a significant benefit with respect to
local control and survival in women who received adjuvant
radiation after mastectomy. Despite two decades of follow-up,
reanalyses, meta-analyses and vigorous and often acrimonious debate, questions about the benefit of PMXRT in women
with one to three positive lymph nodes (LNs) remain unanswered for many. This persistent debate has limited the use
ETWEEN 1997 and 1999, three randomized prospective trials of postmastectomy radiation therapy
(PMXRT) were published, the Danish Premenopausal and
Postmenopausal trials (82b and 82c, respectively) and the
British Columbia trial.1-3 The Danish trials (19821989),
with approximately 1,400 patients each and the smaller
Canadian trial (1978 1985) with 318 patients randomly
assigned women receiving treatment with mastectomy and
adjuvant systemic therapy to receive or not receive PMXRT.
As seen in the earlier studies, these modern trials showed a
significant difference in locoregional recurrence (LRR) rate
in favor of PMXRT. However, for the first time, all three
trials showed a significant improvement in overall survival
(OS) also in favor of radiation. These trials, in contrast to
their predecessors, benefited from modern standardized radiation therapy techniques, as well as modern chemotherapy.
There was general agreement with the results of the trials
with respect to patients with four or more positive lymph
nodes (LNs). However, the true benefit of PMXRT in patients with one to three positive LNs was called into question. All three trials had LRR rates of 30 to 33% at 10 to 15
years in patients with one to three positive LNs who did not
undergo radiation. This rate is in contrast to a 13% 10-year
LRR rate in similarly staged and treated disease as reported
in retrospective reviews of prospective Eastern Cooperative
Oncology Group (ECOG) and National Surgical Adjuvant
Breast and Bowel Project (NSABP) trials.4,5 Commonly
proposed reasons for the LRR discrepancies center on what
some critics would consider the nonstandard systemic and
local therapy administered in these trials.
The criticisms of systemic therapy concern the use of CMF
(cyclophosphamide, methotrexate, fluorouracil) in two of the
trials, and tamoxifen in the third. One may argue that the
results of these PMXRT trials are not applicable to todays
breastpatient with cancer because 1) CMF is no longer the
most common first-line chemotherapy regimen in breast
cancer and 2) tamoxifen was administered without knowledge of the patients estrogen-(ER) and progesteronereceptor (PR) status and was prescribed for only 1 year as
opposed to todays recommended 5-year course. When one
considers that systemic therapy can influence local control
(e.g., B-06, B-21) the argument that substandard systemic
therapy may affect LRR rate seems wholly reasonable.
With respect to local therapy, some have argued that an
inadequate axillary dissection may be the cause of the high
rate of LRR (30 to 33%) seen in the patients with one to three
50
of PMXRT. A treatment concept certain to be as vigorously

debated as that of PMXRT is that of partial breast irradiation
(PBI). However, unlike PMXRT, the acceptance of PBI, by
community physicians, has been nothing short of meteoric
and represents the first major shift in the local therapy
paradigm in more than 10 years. Herein, information will be
presented that will suggest we are nearing the end of one
debatePMXRT in patients with one to three positive LNs
and beginning anotherappropriateness of PBI.
positive LNs. The median numbers of LNs removed were

seven and 11 in these trials, respectively. In contrast, the
median numbers of LNs removed in the retrospective analyses of ECOG and NSABP trials, both of which reported a
13% LRR rate, were 15 and 16, respectively.4,5 The extent of
axillary surgery has been shown to have a local therapy
benefit.5 With that in mind, some concluded that the
radiation in the modern PMXRT trials compensated for
less-than-ideal surgery, and thus the benefit of radiation is
exaggerated for patients with one to three positive LNs. It is
this belief that prevents many physicians from offering
PMXRT to patients with breast cancer who have one to three
positive LNs.6
Meta-analyses can often provide insight when there are
apparent discrepancies between studies. The Early Breast
Cancer Trialists Cooperative Group (EBCTCG) reviewed
the records of more than 3,000 postmastectomy patients
with breast cancer who had one to three positive LNs and
were randomly assigned to receive or not receive adjuvant
radiation. The majority of the patients received systemic
therapy. In this meta-analysis the authors reported an
absolute reduction in breast cancer mortality and all-cause
mortality with PMXRT of 7.6% (log-rank 2p 0.002) and
5.3% (log-rank 2p 0.05), respectively.7
There is also indirect evidence supporting PMXRT in
patients with one to three positive LNs. In the NCIC trial
MA.20, 1,800 women were randomly assigned after lumpectomy to whole-breast irradiation (WBI) with or without
regional nodal irradiation (RNI).8 The results of this trial
were recently presented at ASCOs 47th Annual Meeting
(June 37, 2011, Chicago, IL). Of those enrolled onto MA.20,
85% had one to three positive LNs. With 5-year median
follow-up, the authors reported an improved locoregional
relapse-free survival (94.5% vs. 96.8%, p 0.02), distant
disease-free survival (87% vs. 92.4%, p 0.002) and OS
(90.7% vs. 92.4%, p 0.07) with the addition of RNI. By
showing that there are substantial benefits with the addition of RNI in women with one to three positive LNs, this
From the Departments of Radiation Oncology and Oncology, Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University, Baltimore, MD.
Address reprint requests to Richard C. Zellars, MD, 401 North Broadway, Suite 1440,
Baltimore, MD 21231-2410; email: zellari@jhmi.edu.
1092-9118/10/1-10
POSTMASTECTOMY RADIATION AND PARTIAL BREAST IRRADIATION
trial indirectly validates the findings of the three modern

PMXRT trials.
It is likely that other factors can aid in predicting LRR
risk after mastectomy. For example, researchers have
shown that the 21-gene recurrence score assay (Oncotype
DX; Genomic Health, Inc., Redwood City, CA), commonly
used to determine the risk of distant recurrence, may also be
predictive of the risk of locoregional recurrence.9 Other
factors reported to be associated with increased locoregional
failure include tumor size, positive margins, extracapsular
extension, lymphovascular invasion, response to neoadjuvant chemotherapy, age, ER/PR status, p53 overexpression,
and breast cancer subtypes.5,8,9 While we await evaluation
of putative biologic, genetic and clinical factors predictive of
LRR such that we can judge a patients need for PMXRT, it
is perhaps prudent to at least seriously consider PMXRT for
women with one to three positive axillary LNs.
PBI
In-breast failures after breast-conserving therapy (defined

here as lumpectomy and WBI) occur in the vicinity of the
original primary tumor approximately 70 to 90% of the time.
This phenomenon gave birth to the concept of PBI. Benefits
of this proposed new treatment paradigm include shorter
overall treatment course, increased efficacy via the larger
biologically effective doses that are possible with PBI, and
decreased toxicity because less tissue is exposed to radiation. PBI can be divided into two general techniques: brachytherapy based (radiation close to the target) and teletherapy
based (external beam). Brachytherapy-based PBI normally
occurs postoperatively and may be delivered by interstitial
(needles) or intracavitary (mammosite, contura, etc.) techniques. Teletherapy-based PBI can be delivered either intraoperatively or postoperatively. Of the three largest
randomized controlled trials comparing various PBI techniques versus standard WBI, only one has been published;
another is closed to accrual but the results have not been
KEY POINTS
Three modern postmastectomy radiation therapy

(PMXRT) trials reported improved overall survival
with adjuvant radiation in patients with one to three
positive lymph nodes.
Meta-analyses support the conclusion of these modern PMXRT trials.
A new trial from the National Cancer Institute of
Canada (NCIC), which reported a survival benefit in
women receiving treatment with whole-breast irradiation (WBI) and regional nodal irradiation when
compared with WBI alone, indirectly supports
PMXRT.
Partial breast irradiation (PBI) has been shown to be
comparable with WBI in a single large randomized
controlled trial using intraoperative PBI.
Although smaller trials and retrospective studies
support PBI, it may be best to await the results of
other large randomized controlled trials evaluating
various PBI techniques.
presented, and a third remains open to accrual. The largest

trial published to date is the TARGIT trial in which more
than 2,000 women were randomly assigned to receive WBI
or intraoperative radiation therapy (IORT). Patients received 20 Gy (photons) to the surface of the lumpectomy
bed.10 Patients enrolled were generally low risk: More than
80% were 50 years of age or older, node negative, ERpositive, and HER2-negative. Interestingly, as a result of
various high-risk features, approximately 14% of the patients randomly assigned to receive PBI also received WBI.
Estimates of ipsilateral breast tumor recurrence rates at 4
years were 1.2% and 1.0% for PBI and WBI, respectively.
Although the results are promising, skepticism has prevented wide acceptance. There are concerns that the
follow-up period is too short, the prescribed dose is too low,
and delivery of IORT is logistically too difficult. Nonetheless,
the TARGIT trial supports the concept of PBI.
The second large randomized controlled trial of PBI is
from Milan. Similar to the TARGIT trial, PBI consists of 20
Gy (electrons) delivered to the lumpectomy bed intraoperatively. The accrual goal of approximately 1,306 women was
reached some time ago. So far, one of the investigators has
stated that the there is no difference in survival between the
two arms at 10 years.11 We eagerly await the results of this
important study.
The largest trial is that of NSABP B-39/Radiation Therapy Oncology Group (RTOG) 0413.12 This trial is unique not
only for its accrual goal of 4,300 women but also because
three separate PBI techniques are permitted. Women randomly assigned to receive PBI may be offered, depending on
the hosting institution, interstitial or intracavitary brachytherapy or postoperative teletherapy. Interestingly, with
more than 4,100 patients enrolled thus far (and on target to
complete accrual in 2012), more than 70% of the patients
receiving treatment with PBI received it via the postoperative teletherapy technique. The trial completed accrual of
low-risk patients, and now remains open to accrue patients
with a higher risk of recurrence (e.g., ER-negative, LNpositive, or age younger than 50 years).
Another potential benefit of PBI is that it could be combined with systemic therapy, thereby not only shortening
the course of radiation but shortening the the overall
treatment course. For example in a small feasibility study,
patients received treatment with PBI (2.7 Gy twice daily
for 15 days) and concurrent dose-dense doxorubicin and
cyclophosphamide.13 The authors reported that systemic
toxicity was minimal, radiation dermatitis was far less than
that seen in standard WBI without concurrent chemotherapy, and cosmetic outcome was good to excellent in more
than 90% of the patients. This small pilot study deserves
validation in larger trials but may hint to a new era of
combined-modality therapy in the management of breast
cancer.
Given the general public interest in PBI, and the fact that
smaller randomized controlled trials and retrospective
studies support its use, the American Society for Therapeutic Radiology and Oncology (ASTRO) published guidelines
defining patient characteristics most appropriate for this
new treatment option. Many practitioners have accepted
these guidelines and use them in daily practice. However, a
recent presentation of a study comparing brachytherapybased PBI versus WBI may have placed a pall over the
51
RICHARD C. ZELLARS
passionate pursuit of PBI. This study, presented at the

34th Annual San Antonio Breast Cancer Symposium (December 6 10, San Antonio, TX), reviewed the Medicare
records of 130,535 patients who received treatment with
brachytherapy-based PBI or standard WBI between 2001
and 2007.14 The authors reported that women receiving
treatment with brachytherapy-based PBI were almost twice
as likely to subsequently require a mastectomy, be hospital-
ized for complications, and acquire a treatment-induced

infection compared with those who received the more standard WBI. One must note that this study is retrospective
and covers a period in which new PBI brachytherapy techniques were just being introduced. Nonetheless, it is perhaps most prudent for practitioners to reserve PBI for the
protocol setting, or to wait for longer-term results from
randomized controlled trials.
Author
Richard C. Zellars*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Overgaard M, Hansen PS, Overgaard J, et al. Postoperative radiotherapy in high-risk premenopausal women with breast cancer who receive
adjuvant chemotherapy. Danish Breast Cancer Cooperative Group 82b Trial.
N Engl J Med. 1997;337:949-955.
2. Overgaard M. Overview of randomized trials in high risk breast cancer
patients treated with adjuvant systemic therapy with or without postmastectomy irradiation. Semin Radiat Oncol. 1999;9:292-299.
3. Ragaz J, Olivotto IA, Spinelli JJ, et al. Locoregional radiation therapy in
patients with high-risk breast cancer receiving adjuvant chemotherapy:
20-year results of the British Columbia randomized trial. J Natl Cancer Inst.
2005;97:116-126.
4. Taghian A, Jeong JH, Mamounas E, et al. Patterns of locoregional
failure in patients with operable breast cancer treated by mastectomy and
adjuvant chemotherapy with or without tamoxifen and without radiotherapy:
results from five National Surgical Adjuvant Breast and Bowel Project
randomized clinical trials. J Clin Oncol. 2004;22:4247-4254.
5. Recht A, Gray R, Davidson NE, et al. Locoregional failure 10 years after
mastectomy and adjuvant chemotherapy with or without tamoxifen without
irradiation: experience of the Eastern Cooperative Oncology Group. J Clin
Oncol. 1999;17:1689-1700.
6. Ceilley E, Jagsi R, Goldberg S, et al. RADIOTHERAPY for invasive
breast cancer in north America and Europe: results of a survey. Int J Radiat
Oncol Biol Phys. 2005;61:365-373.
7. McGale P. The 2006 worldwide overview of the effects of local treatments
for early breast cancer on long-term outcome: meta-analysis of the randomized trials of radiotherapy after mastectomy with axillary clearance. Pre-
52
sented at: 48th Annual Meeting of the American Society of Therapeutic

Radiology and Oncology; November 2006; Philadelphia, PA.
8. Whelan T. NCIC-CTG MA.20: an intergroup trial of regional nodal
irradiation in early breast cancer. J Clin Oncol. 2011;29 (suppl; abstr
LBA1003).
9. Voogd AC, Nielsen M, Peterse JL, et al. Differences in risk factors for
local and distant recurrence after breast-conserving therapy or mastectomy
for stage I and II breast cancer: pooled results of two large European
randomized trials. J Clin Oncol. 2001;19:1688-1697.
10. Vaidya JS, Joseph DJ, Tobias JS, et al. Targeted intraoperative
radiotherapy versus whole breast radiotherapy for breast cancer (TARGIT-A):
an international, prospective, randomized, noninferiority phase 3 trial. Lancet. 2010;376:91-102.
11. Oncology Times UK. June 2011 Volume 8.
12. Julian T. Early toxicity results with 3-D conformal external beam
therapy from the NSABP B-39/RTOG 0413 Accelerated Partial Breast Irradiation (APBI) Trial. Presented at: 53rd Annual Meeting of the American
Society of Therapeutic Radiology and Oncology; October 2011; Philadelphia,
PA.
13. Zellars RC, Stearns V, Frassica D, et al. Feasibility trial of PBI with
concurrent dose-dense doxorubicin and cyclophosphamide in early-stage
breast cancer. J Clin Oncol 2009;27:2816-2822.
14. Smith GL, Buchholz TA, Giordano SH et al. Partial breast brachytherapy is associated with inferior effectiveness and increased toxicity compared
with whole breast irradiation in older patients. Presented at: 34th Annual San
Antonio Breast Cancer Symposium; December 2011; San Antonio, TX.
The Appropriate Extent of Surgery for

Early-Stage Breast Cancer
By Monica Morrow, MD
Overview: Attitudes regarding the appropriate extent of surgery for breast cancer and the effect of surgery on breast
cancerspecific survival have varied over time. Failure to
maintain local control is associated with decreased survival,
but the extent of surgery necessary for local control has
decreased as other treatment modalities, such as radiotherapy and systemic therapy, have become more widely used.
Both endocrine therapy and chemotherapy considerably re-
ERCEPTIONS OF the role of surgery in contributing to

the cure of breast cancer have varied over time. Until
the mid-1970s, surgery was performed with the intent of
curing breast cancer, and bigger was better. In the 1980s
and 1990s the popularization of the systemic disease
hypothesis by Dr. Bernard Fisher opened the door to smaller
surgical procedures, the widespread use of systemic therapy,
and the thought that surgery was primarily a staging procedure. In 2005 the Early Breast Cancer Trialists Collaborative
Group (EBCTCG) overview analysis1 demonstrated that failure to achieve local control in both node-positive and nodenegative women at 5 years after either lumpectomy or
mastectomy decreased breast cancer specific survival at 15
years, indicating that local therapy does affect survival. This
observation renewed interest in the appropriate extent of
surgery to maximize local control in the index breast and the
axilla, as well as the role of contralateral prophylactic
mastectomy (CPM) in breast cancer management.
Extent of Surgery in the Index Breast: Margins
The selection criteria for breast-conserving surgery (BCS)

developed in the 1980s and in use today are related to the
extent of the tumor burden in the breast and the ability to
safely deliver radiotherapy. Biologic tumor features such as
estrogen receptor (ER), histologic grade, and node status are
not used as selection criteria for lumpectomy compared with
mastectomy.2 The most important determinant of tumor
burden in the breast with a unicentric cancer is margin
status. Although it is clear that patients with positive
margins, defined as tumor cells touching ink, have an
increased risk of local recurrence compared with those with
negative margins;3 evidence that more widely clear margins
further reduce local recurrence is lacking.
In the prospective, randomized trials that established the
safety of BCS, only the National Surgical Adjuvant Breast
and Bowel Project (NSABP) B06 study required a microscopically negative margin, defined as tumor cells not touching
ink.4 After 20 years of follow-up, patients undergoing
lumpectomy and radiation therapy (RT) in this study had a
survival outcome equivalent to those having mastectomy,
and only 14% developed a local recurrence (LR).
Since the time of this study, the rates of LR in NSABP
trials have steadily declined, although the required margin
has not changed.5 This can largely be attributed to the
widespread use of adjuvant systemic therapy, although
improvements in pathologic evaluation and the quality of
mammography have also contributed to this trend. Data
from prospective randomized trials of adjuvant systemic
duce rates of local recurrence in the breast, as well as the

incidence of contralateral breast cancer, and as efficacy in
reducing metastatic disease increases, so does the benefit in
reducing local recurrence. The excellent rates of local control
in the ACOSOG Z11 trial after elimination of axillary dissection
in patients with positive sentinel nodes receiving whole-breast
irradiation and systemic therapy are a model for reducing
surgical morbidity in the era of multimodality therapy.
chemotherapy and endocrine therapy demonstrate reductions in LR rates from 13% to 15% in patients randomized to
no adjuvant therapy compared with 3% to 4% in those
receiving adjuvant therapy.6,7
As the efficacy of systemic therapy improves, LR is further
reduced. In the randomized trials of the addition of trastuzumab to chemotherapy in patients with HER2 overexpressing tumors, the addition of trastuzumab resulted in an
approximately 40% relative reduction in the risk of LR
compared with treatment with chemotherapy alone.8 Rates
of LR less than 5% at 8 10 years are regularly seen in
patients undergoing BCS in more recent time periods.5,9
Despite these favorable outcomes, there appears to be confusion regarding what constitutes an adequate margin of
resection as evidenced by the results of a population-based
survey of 318 surgeons in which only 11% endorsed tumor
not touching ink as an adequate margin, 42% endorsed an
adequate margin of greater than 12 mm, 28% endorsed an
adequate margin of greater than 5 mm, and 19% endorsed
an adequate margin greater than 1 cm.10 A similar lack of
consensus has been documented among radiation oncologists.11 This lack of consensus has made re-excision a
common procedure,12 and a study of 2,206 patients with
invasive breast cancer documented that although 23% of
patients had re-excision, almost half (47.5%) of the reexcisions were performed for margins where tumor was not
touching ink.13 A meta-analysis of 14,571 patients with
1,026 local recurrences found no difference in rates of local
recurrence for any negative margin width greater than
tumor not touching ink after adjusting for factors such as
the use of a boost dose of RT and endocrine therapy.3 Other
strategies, such as magnetic resonance imaging, that are
intended to reduce the subclinical tumor burden (the goal of
more widely clear margins) have also not been shown to
decrease LR.9,14 In addition, there is an increasing body of
evidence indicating that patients at high risk of LR after
BCS are also at high risk of LR after mastectomy.15,16 Two
studies examining LR after BCS and mastectomy for patients with triple-negative breast cancers have found no
From the Breast Service, Department of Surgery, Memorial Sloan-Kettering Cancer

Center, New York, NY.
Address reprint requests to Monica Morrow, MD, Breast Service, Department of Surgery,
Memorial Sloan-Kettering Cancer Center, 300 East 66th St., New York, NY 10065; email:
morrowm@mskcc.org.
1092-9118/10/1-10
53
MONICA MORROW
17,18
advantage for mastectomy

in this patient subset at high
risk of distant relapse. In a retrospective study of the relationship of the 21-gene recurrence score (OncotypeDx) to the risk
of LR, the use of appropriate systemic therapy as predicted
by the 21-gene recurrence score, significantly reduced LR
(p 0.0001), and in a multivariate model, type of surgery
(lumpectomy compared with mastectomy) was not significantly
associated with LR after controlling for biologic variables.16
In aggregate, these findings indicate that local control is a
complex interaction between disease burden, tumor biology,
and the effectiveness of systemic therapy. In the current era
where multimodality treatment is routinely employed, there
is little evidence that larger surgical resections improve
patient outcomes.
Axilla
The multidisciplinary nature of breast cancer therapy

today has also affected our approach to the axilla. Axillary
dissection has traditionally been performed for staging and
local control. Its role in contributing to the cure of breast
cancer has been controversial since the NSABP B04 trial
demonstrated no difference in survival in patients treated
with and without axillary dissection in the prechemotherapy
era.19 Sentinel node biopsy is a reliable method of identifying axillary node involvement with a lower morbidity than
axillary dissection,20 and biologic tumor features, rather
than number of axillary lymph nodes involved with tumor,
are increasingly used to select systemic therapy. In patients
undergoing BCS, the tangent field RT, which is a standard
part of treatment, covers some of the axilla,21 and systemic
therapy also contributes to locoregional control.6-8 These
observations resulted in the American College of Surgeons
Group (ACOSOG) Z11 study in which clinically nodenegative women with T1 and T2 breast cancers undergoing
BCS and found to have hematoxylin and eosin-detected
metastasis in 3 sentinel nodes were randomized to completion axillary dissection or no further axillary treatment.22,23 All patients received tangent breast RT and
systemic therapy. After a median follow-up of 6.3 years, the
KEY POINTS
54
Rates of local recurrence after breast-conserving surgery and radiotherapy have decreased over time.
Evidence that lumpectomy margins more widely
clear further reduce local recurrence than tumor not
touching ink is lacking.
Local control can be achieved with removal of the
sentinel nodes and no further axillary treatment for
patients with involvement of one or two sentinel
nodes treated with whole-breast irradiation and systemic therapy.
The incidence of contralateral breast cancer has been
declining since the mid-1980s because of the increased use of adjuvant systemic therapy.
Systemic therapy has a major effect on local control,
offering the opportunity to decrease the extent and
morbidity of surgery as the effectiveness of systemic
therapy increases.
Table 1. Patient Groups for Whom Axillary Dissection Remains

Standard Management of a Positive Sentinel Node
Palpable Node Disease

Locally Advanced (Stage III) Breast Cancer
Treatment with Mastectomy
Treatment with Partial Breast Irradiation
Treatment with Neoadjuvant Therapy
Involvement of 3 Sentinel Nodes with Metastases
Gross Extranodal Tumor Extension in Sentinel Nodes
regional recurrence rate in the sentinel node only arm was

0.9% compared with 0.4% in the axillary dissection arm,
despite the fact that 27.4% of patients treated with axillary
dissection had additional positive nodes removed beyond the
sentinel node.23 No survival differences were seen between
groups,22 and morbidity was significantly lower in the sentinel node-only group.22,23 This study clearly indicates that
axillary dissection can be eliminated for patients meeting
the ACOSOG Z11 eligibility criteria, but does not apply to a
considerable number of women with breast cancer (Table 1).
There have been concerns expressed that because most of
the patients in Z11 were postmenopausal with ER-positive
tumors, these results cannot be extrapolated to younger
women and those with ER-negative tumors. However, large
studies of regional node recurrence do not identify age or ER
status as predictors,24,25 suggesting that they should not be
used to exclude women from this approach. Implementing
the results of Z11 necessitates some changes in practice. For
example, the identification of a single metastatic axillary
node with ultrasound and fine-needle aspiration no longer
changes the operative approach because the sentinel node
still must be surgically removed, and it should be abandoned. Frozen section of the sentinel node also is not necessary or appropriate in these cases because the finding of tumor
in a single node will not lead to dissection, and because
examination of all nodal tissue is necessary to determine the
number of nodes involved and the presence of extranodal
extension. Finally, based on the results of ACOSOG Z10 and
NSABP B32 indicating that micrometastases in the sentinel
node have minimal prognostic impact, the routine use of
immunohistochemistry to look for small tumor deposits can
be abandoned.
Contralateral Breast
The use of contralateral prophylactic mastectomy (CPM)

is increasing,26 with the most notable increases seen in
women less than 50 years of age and those treated in
hospitals managing a higher volume of patients with breast
cancer.27 This is somewhat surprising, considering that the
clinical incidence of contralateral breast cancer has been
declining since 1985 because of the widespread use of
adjuvant therapy.28 This decrease is most evident among
women with ER-positive cancers, where even for those in
their 20s and 30s at initial diagnosis, 10-year rates of
contralateral cancer are 2.5% 4.5%.28 With such a low
incidence rate, it is not surprising that evidence indicating
that CPM improves breast cancer mortality is lacking.29
Evidence suggests that level of anxiety, regardless of the
presence or absence of risk factors for contralateral breast
cancer, is a major predictor of undergoing CPM.30
SURGERY FOR EARLY-STAGE BREAST CANCER

Conclusion
Current risks for LR with lumpectomy to a margin of

tumor not touching ink, and in the axilla after management
of a positive node with sentinel node biopsy alone in patients
undergoing BCS, emphasize the powerful effect of systemic
therapy on local control. As the options for targeted therapy
expand and our ability to tailor the extent of systemic
therapy to an individual patients level of risk improves, not
only does survival improve, but local recurrence decreases as

well. These findings have important implications for the
future of local therapy. More aggressive surgical strategies
are unlikely to improve local outcomes for poor-prognosis
cancers in the absence of improved systemic therapies.
Strategies to reduce the extent of surgery in patient subsets
where systemic therapy is available should be sought to
reduce the morbidity of treatment.
Author
Monica Morrow*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Clarke M, Collins R, Darby S, et al. Effects of radiotherapy and of
differences in the extent of surgery for early breast cancer on local recurrence
and 15-year survival: An overview of the randomised trials. Lancet. 2005;
366(9503):2087-2106.
2. Morrow M, Harris JR. Practice guideline for breast conservation therapy
in the management of invasive breast cancer. J Am Coll Surg. 2007;205:362376.
3. Houssami N, Macaskill P, Marinovich ML, et al. Meta-analysis of the
impact of surgical margins on local recurrence in women with early-stage
invasive breast cancer treated with breast-conserving therapy. Eur J Cancer.
2010;46(18):3219-3232.
4. Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a
randomized trial comparing total mastectomy, lumpectomy, and lumpectomy
plus irradiation for the treatment of invasive breast cancer. N Engl J Med.
2002;347(16):1233-1241.
5. Anderson SJ, Wapnir I, Dignam JJ, et al. Prognosis after ipsilateral
breast tumor recurrence and locoregional recurrences in patients treated by
breast-conserving therapy in five National Surgical Adjuvant Breast and
Bowel Project protocols of node-negative breast cancer. J Clin Oncol. 2009;
27(15):2466-2473.
6. Fisher B, Dignam J, Bryant J, et al. Five versus more than five years of
tamoxifen therapy for breast cancer patients with negative lymph nodes and
estrogen receptor-positive tumors. J Natl Cancer Inst. 1996;88(21):1529-1542.
7. Fisher B, Dignam J, Mamounas EP, et al. Sequential methotrexate and
fluorouracil for the treatment of node-negative breast cancer patients with
estrogen receptor-negative tumors: Eight-year results from National Surgical
Adjuvant Breast and Bowel Project (NSABP) B-13 and first report of findings
from NSABP B-19 comparing methotrexate and fluorouracil with conventional cyclophosphamide, methotrexate, and fluorouracil. J Clin Oncol. 1996;
14(7):1982-1992.
2005;353(16):1673-1684.
9. Solin LJ, Orel SG, Hwang WT, et al. Relationship of breast magnetic
resonance imaging to outcome after breast-conservation treatment with
radiation for women with early-stage invasive breast carcinoma or ductal
carcinoma in situ. J Clin Oncol. 2008;26(3):386-391.
10. Azu M, Abrahamse P, Katz SJ, et al. What is an adequate margin for
breast-conserving surgery? Surgeon attitudes and correlates. Ann Surg Oncol. 2010;17(2):558-563.
11. Taghian A, Mohiuddin M, Jagsi R, et al. Current perceptions regarding
surgical margin status after breast-conserving therapy: Results of a survey.
Ann Surg. 2005;241(4):629-639.
12. Morrow M, Jagsi R, Alderman AK, et al. Surgeon recommendations and
receipt of mastectomy for treatment of breast cancer. JAMA. 2009;302(14):
1551-1556.
13. McCahill LE, Single RM, Aiello Bowles EJ, et al. Variability in
Reexcision Following Breast Conservation Surgery. JAMA. 2012;307(5):467475.
14. Hwang N, Schiller DE, Crystal P, et al. Magnetic resonance imaging in
the planning of initial lumpectomy for invasive breast carcinoma: Its effect on
ipsilateral breast tumor recurrence after breast-conservation therapy. Ann
Surg Oncol. 2009;16(11):3000-3009.
15. Kyndi M, Sorensen FB, Knudsen H, et al. Estrogen receptor, progesterone receptor, HER-2, and response to postmastectomy radiotherapy in
high-risk breast cancer: The Danish Breast Cancer Cooperative Group. J Clin
Oncol. 2008;26(9):1419-1426.
16. Mamounas EP, Tang G, Fisher B, et al. Association between the
21-gene recurrence score assay and risk of locoregional recurrence in nodenegative, estrogen receptor-positive breast cancer: Results from NSABP B-14
and NSABP B-20. J Clin Oncol. 2010;28(10):1677-1683.
17. Abdulkarim BS, Cuartero J, Hanson J, et al. Increased risk of locoregional recurrence for women with T1-2N0 triple-negative breast cancer
treated with modified radical mastectomy without adjuvant radiation therapy
compared with breast-conserving therapy. J Clin Oncol. 2011;29(21):28522858.
18. Ho AY, Gupta G, King TA, et al. Favorable Prognosis in Patients with
T1a/bN0 Triple Negative Breast Cancers Treated With Multimodality Therapy. Cancer. 2012 (In Press).
19. Fisher B, Jeong JH, Anderson S, et al. Twenty-five-year follow-up of a
randomized trial comparing radical mastectomy, total mastectomy, and total
mastectomy followed by irradiation. N Engl J Med. 2002;347(8):567-575.
20. Krag DN, Anderson SJ, Julian TB, et al. Technical outcomes of
sentinel-lymph-node resection and conventional axillary-lymph-node dissection in patients with clinically node-negative breast cancer: Results from the
NSABP B-32 randomised phase III trial. Lancet Oncol. 2007;8(10):881-888.
21. Reznik J, Cicchetti MG, Degaspe B, Fitzgerald TJ. Analysis of axillary
coverage during tangential radiation therapy to the breast. Int J Radiat
Oncol Biol Phys. 2005;61(1):163-168.
22. Giuliano AE, Hunt KK, Ballman KV, et al. Axillary dissection vs no
axillary dissection in women with invasive breast cancer and sentinel node
metastasis: A randomized clinical trial. JAMA. 2011;305(6):569-575.
23. Giuliano AE, McCall L, Beitsch P, et al. Locoregional recurrence after
sentinel lymph node dissection with or without axillary dissection in patients
with sentinel lymph node metastases: The American College of Surgeons
Oncology Group Z0011 randomized trial. Ann Surg. 2010;252(3):426-432;
discussion 32-33.
24. Grills IS, Kestin LL, Goldstein N, et al. Risk factors for regional nodal
failure after breast-conserving therapy: Regional nodal irradiation reduces
rate of axillary failure in patients with four or more positive lymph nodes. Int
J Radiat Oncol Biol Phys. 2003;56(3):658-670.
25. Yates L, Kirby A, Crichton S, et al. Risk Factors for Regional Nodal
Relapse in Breast Cancer Patients with One to Three Positive Axillary Nodes.
Int J Radiat Oncol Biol Phys (In Press). 2011.
26. Tuttle TM, Habermann EB, Grund EH, et al. Increasing use of
contralateral prophylactic mastectomy for breast cancer patients: A trend
toward more aggressive surgical treatment. J Clin Oncol. 2007;25(33):52035209.
27. Yao K, Stewart AK, Winchester DJ, Winchester DP. Trends in contralateral prophylactic mastectomy for unilateral cancer: A report from the National
Cancer Data Base, 1998-2007. Ann Surg Oncol. 2010;17(10):2554-2562.
28. Nichols HB, Berrington de Gonzalez A, Lacey JV, Jr., et al. Declining
incidence of contralateral breast cancer in the United States from 1975 to
2006. J Clin Oncol. 2011;29(12):1564-1569.
29. Lostumbo L, Carbine NE, Wallace J. Prophylactic mastectomy for the
prevention of breast cancer. Cochrane Database Syst Rev. (Online). 2010(11):
CD002748.
30. Hawley ST, Jagsi R, Morrow M, Katz SJ. Correlates of contralateral
prophylactic mastectomy in a population-based sample. J Clin Oncol. 2011;29
(Abstract 6010).
55
How Does Biology Affect Local

Therapy Decisions?
By Thomas A. Buchholz, MD
Overview: Breast cancer represents a biologically diverse

set of diseases. Previous data suggest that the estrogen/
progesterone receptor (ER/PR) status and HER2/neu status
are important determinants of prognosis and response to
various systemic treatments. Recent data also suggest that
these receptors correlate with outcomes of local-regional
therapies. Specifically, patients with ER-positive HER2/neunegative disease have an excellent outcome with radiation
ENOMIC AND molecular analyses of breast cancer

have elucidated that breast cancer represents a biologically diverse group of diseases. Studies indicate that molecular subtypes of breast cancer differ with respect to tumor
biology, prognosis and response to hormone therapy, targeted therapy, and chemotherapy. In part, these subtypes
are driven by differences in ER signaling and HER2/neu
status. While breast cancer biology has had a major role in
determining systemic treatment approaches, it is less clear
how biologic markers of disease should affect surgery and
radiation decisions. Fewer data exist correlating breast
cancer biology and local-regional treatment outcome compared with the available data correlating molecular markers
with response to systemic treatments, overall recurrence,
distant metastases, and death. This article will highlight
some of the emerging data that suggest biology plays an
important role in local treatment outcome of breast cancer.
Breast Conservation
The local-regional treatment outcomes after breastconservation surgery and whole-breast radiation are excellent. Improvements in imaging techniques, increased
attention to surgical margins, and increased use of systemic
therapy have helped reduce recurrence rates to very low
levels. For example, our institutional experience of BCT
(lumpectomy plus whole-breast radiation) for 974 patients
treated between 1973 and 1993 noted a 5-year risk of
in-breast recurrence of 5.7%. This compared with a 5-year
recurrence risk of only 1.3% noted in the 381 patients
treated between 1994 and 1996.1
Despite these overall excellent outcomes, recent data have
demonstrated that some subtypes of breast cancer have
higher local recurrence rates. For example, investigators
from the Harvard Radiation Oncology program analyzed a
cohort of 1,223 patients treated with BCT and noted 5-year
breast recurrence rates of 4.4% for patients with triplenegative disease, 9% for those with HER2-positive disease
(predated trastuzumab), and only 1% for patients with
ER-positive disease.2 These data are similar to a report
from University of Texas M. D. Anderson Cancer Center
(MDACC) that analyzed 911 patients treated with BCT for
primaries under 1 cm and negative lymph nodes and found
8-year breast recurrence rates of 18% in HER2/neu-positive
disease (predated trastuzumab), 11% for those with ERnegative disease, and only 4% for those with ER-positive
disease.3 Finally, an elegant study from British Columbia
that investigated more than 1,400 patients with BCT reported a higher 10-year local-regional recurrence rate in
56
treatments, either given as a component of breast-conservative

therapy (BCT) or for those with more advanced disease, when
given after mastectomy. For patients with triple-negative
disease, data suggest that the proportional benefits offered
from radiation in reducing local-regional recurrences may be
less. This article will highlight some of these data and discuss
strategies for new local-regional research avenues that are
based on breast cancer biologic subtype.
patients with HER2/neu-enriched (21%) and basal-like (16%)

cancers compared with the much more common luminal A
subtype (8%).4
Some groups have analyzed local recurrence after BCT
using more sophisticated genomic signatures. For example,
a cDNA microarray-based wound signature successfully
classified patients into two groups with either high or low
risk of local-regional recurrence.5 In a different study, investigators from the National Surgical Adjuvant Breast and
Bowel Project (NSABP) analyzed the 21-gene profile (Oncotype DX) in more than 1,500 patients with node-negative
ER-positive disease and reported that patients with a higher
gene profile score had a higher risk of local recurrence.6
It remains a question how data such as these should affect
local-regional treatment decisions. The majority of patients
treated with BCT have ER-positive or luminal A type
tumors and these data collectively indicate very low rates of
breast recurrence. Accordingly, the focus of future studies
in this population should be aimed at minimizing the cost,
inconvenience, and toxicity of such treatments. For patients
with higher-risk biologic features, improvements in systemic
treatments may help mitigate risks. For example, the
NSABP study of oncotype noted only an 8% 10-year risk for
patients with high recurrence scores who received chemotherapy and tamoxifen compared with a 16% risk for those
who received tamoxifen alone.6 Systemic treatments also
likely overcome some of the adverse effects of HER2/neu
overexpression. Data from the B-31/N9831 trials indicated
the use of adjuvant trastuzumab reduced the risk of localregional recurrence in HER2/neu-positive cancers by 42%
(crude event rates, 47 of 1,679 for no trastuzumab compared
with 27 of 1,672 with trastuzumab).7 In addition, a study
from Memorial Sloan-Kettering Cancer Center noted 3-year
local-regional recurrence rates of 10% for patients with
HER2/neu-positive tumors treated before trastuzumab and
only a 1% rate in those receiving adjuvant trastuzumab.8
The one cohort of patients who remain at higher risk of
recurrence after BCT appears to be those with triplenegative disease. For such patients, it is reasonable to help
From the Division of Radiation Oncology, Department of Radiation Oncology, University

of Texas M. D. Anderson Cancer Center, Houston, TX.
Address reprint requests to Thomas A. Buchholz, MD, 1515 Holcombe Blvd., Division of
Radiation Oncology, Unit 97, University of Texas M. D. Anderson Cancer Center, Houston,
TX 77030; email: tbuchhol@mdanderson.org.
1092-9118/10/1-10
BIOLOGY AND LOCAL THERAPY DECISIONS
mitigate this risk through routine use of chemotherapy,

achievement of negative surgical margin status, and use of
a tumor bed boost. Patients with triple-negative disease
should not be routinely recommended to undergo mastectomy. A recent study from Edmonton, Canada, evaluated
768 patients with T12N0 triple-negative disease and found
better 5-year local-regional recurrence-free survival after
BCT with radiation compared with mastectomy without
radiation (96% compared with 90%, respectively, p 0.027;
hazard ratio, 2.53; p 0.0264).9
Mastectomy with or without Radiation
Biologic subtypes also appear to affect the risk of localregional recurrence after mastectomy. For example, a British Columbia study that combined data from patients
treated with mastectomy alone and patients treated with
mastectomy plus radiation demonstrated that luminal A
subtype independently correlated with a lower risk of localregional recurrence after mastectomy compared with the
other subtypes.4 Interestingly, in an analysis of patients
KEY POINTS
Estrogen receptor (ER)-positive, HER2/neu-negative

breast cancers have very low local-regional recurrence rates after adjuvant radiation treatments.
Triple-negative breast cancer has higher rates of
local-regional recurrence after adjuvant radiation
compared with similar stages of ER-positive disease.
Breast-conservative therapy remains an appropriate
treatment option for patients with early-stage, triplenegative disease.
treated on the Danish postmastectomy radiation trials and

those who were randomly selected to not receive radiation,
the risks of local-regional recurrence for patients with ERpositive disease and those with ER-negative disease was
similar (34% and 31%, respectively).10 However, in the
patients randomly selected to receive postmastectomy radiation, the rates of local-regional recurrence were lower in
the patients with ER-positive disease compared with those
with ER-negative disease (4% and 14%, respectively).10
Similarly, a report of local-regional treatment outcome of
patients treated at MDACC with mastectomy, no radiation,
and adjuvant chemotherapy did not find ER status to independently correlate with local-regional recurrence (15% ER
positive and 12% ER negative).11 However, in a recursive
partition analysis of patients treated with postmastectomy
radiation, ER-negative disease proved to be the most powerful discriminator of local-regional recurrence risk.12
Radiation Resistance and Future Research
Collectively, the data from breast conservation and postmastectomy radiation raise the hypothesis that ER negativity or triple-negative disease may be a more radiationresistant histology than ER-positive disease. These data
were also supported by the most recent update of the Early
Breast Cancer Trialists Group meta-analysis of trials
comparing use with omission of radiation after breastconservation surgery. In these trials, the proportional reduction in the risk of recurrence with radiation was nearly
60% for ER-positive disease treated with tamoxifen, approximately 50% for ER-positive disease without tamoxifen
treatment, and only 35% for ER-negative disease. These
data indicate that additional research is needed to identify
potential molecular targets in triple-negative disease and to
design new therapeutic strategies to enhance radiation
effects in this subtype.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Thomas A. Buchholz*
REFERENCES
1. Cabioglu N, Hunt KK, Buchholz TA, et al. Improving local control with
breast-conserving therapy: A 27-year single-institution experience. Cancer.
2005;104:20-29.
2. Nguyen PL, Taghian AG, Katz MS, et al. Breast cancer subtype approximated by estrogen receptor, progesterone receptor, and HER-2 is associated
with local and distant recurrence after breast-conserving therapy. J Clin
Oncol. 2008;26:2373-2378.
3. Albert JM, Gonzalez-Angulo AM, Guray M, et al. Estrogen/progesterone
receptor negativity and HER2 positivity predict locoregional recurrence in
patients with T1a,bN0 breast cancer. Int J Radiat Oncol Biol Phys. 2010;77:
1296-1302.
4. Voduc KD, Cheang MC, Tyldesley S, et al. Breast cancer subtypes and
the risk of local and regional relapse. J Clin Oncol. 2010;28:1684-1691.
5. Nuyten DS, Kreike B, Hart AA, et al. Predicting a local recurrence after
breast-conserving therapy by gene expression profiling. Breast Cancer Res.
2006;8:R62.
6. Mamounas EP, Tang G, Fisher B, et al. Association between the 21-gene
recurrence score assay and risk of locoregional recurrence in node-negative,
estrogen receptor-positive breast cancer: Results from NSABP B-14 and
NSABP B-20. J Clin Oncol. 2010;28:1677-1683.

2005;353:1673-1684.
8. Kiess AP, McArthur HL, Mahoney K, et al. Adjuvant trastuzumab
reduces locoregional recurrence in women who receive breast-conservation
therapy for lymph node-negative, human epidermal growth factor receptor
2-positive breast cancer. Cancer. Epub 2011 Sep 1.
9. Abdulkarim BS, Cuartero J, Hanson J, et al. Increased risk of locoregional recurrence for women with T1-2N0 triple-negative breast cancer
treated with modified radical mastectomy without adjuvant radiation therapy
compared with breast-conserving therapy. J Clin Oncol. 2011;29:2852-2858.
10. Kyndi M, Sorensen FB, Knudsen H, et al. Estrogen receptor, progesterone receptor, HER-2, and response to postmastectomy radiotherapy in
high-risk breast cancer: The Danish Breast Cancer Cooperative Group. J Clin
Oncol. 2008;26:1419-1426.
11. Katz A, Strom EA, Buchholz TA, et al. Loco-regional recurrence
patterns following mastectomy and doxorubicin-based chemotherapy: Implications for postoperative irradiation. J Clin Oncol. 2000;18:2817-2827.
12. Woodward WA, Strom EA, Tucker SL, et al. Locoregional recurrence
after doxorubicin-based chemotherapy and postmastectomy: Implications for
breast cancer patients with early-stage disease and predictors for recurrence
after postmastectomy radiation. Int J Radiat Oncol Biol Phys. 2003;57:336-344.
57
THE MANY OPTIONS FOR BREAST IMAGING:

WHAT TO USE WHEN
CHAIR
Chris I. Flowers, MD
H. Lee Moffitt Cancer Center and Research Institute
Tampa, FL
SPEAKERS
Maxine Jochelson, MD
New York, NY
Karla Kerlikowske, MD
University of California, San Francisco
San Francisco, CA
Clinical and Imaging Surveillance Following

Breast Cancer Diagnosis
By Chris I. Flowers, MBBS, Blaise P. Mooney, MD, and Jennifer S. Drukteinis, MD
Overview: Breast cancer is the most common malignancy

affecting women worldwide. Women have a 1 in 8 lifetime risk
of breast cancer. Breast conservation therapy (BCT) is the
most common method of definitive treatment. Patients who
previously have had to undergo mastectomy may be now
eligible for BCT or a multitude of options for reconstruction,
either immediate or delayed. Surveillance imaging after a
breast cancer diagnosis is important because there is an
increased risk of recurrence developing in patients, and early
detection has been shown to improve survival. There is
currently no consensus on a protocol for imaging the postoperative breast. In patients who have undergone mastectomy,
OCAL RECURRENCE, second ipsilateral breast cancer, and contralateral breast cancer may develop in
women with a personal history of breast cancer.1-3 Lumpectomy followed by whole breast irradiation or BCT is the most
common treatment following a diagnosis of breast cancer.
BCT of clinical stage I and stage II breast cancer has become
a standard of care, with long-term studies showing no
significant difference (p 0.001) in survival rates between
those treated with BCT as opposed to mastectomy.4-10 Early
detection of local recurrence of breast cancer has been shown
to improve long-term survival (HR 2.44 [95% CI] if without
symptoms); therefore, it is important to optimize the use of
clinical and imaging tools for the patient with a history of
breast cancer to diagnose recurrence in its most early stages.
Locoregional recurrence occurs in approximately 5% of
patients at 5 years with a local failure rate of approximately
1% to 2.5% per year. In the immediate postoperative period,
suspicious findings likely represent residual disease, whereas
local recurrence typically occurs 37 years after BCT.11,12
Many factors affect local, regional recurrence and distant
metastasis. These include age at diagnosis, primary tumor
subtype and size, presence of local or regional node disease,
surgical treatment, use and type of radiation therapy, and
use of adjuvant hormonal therapy and chemotherapy.
With the increasing number and variability of oncoplastic
procedures, there are many potential ways in which local
recurrence can appear and be detected on imaging. Mammographic surveillance has shown utility in the detection
of recurrent disease, and in screening the contralateral
breast for early breast cancer. As a result, mammographic
follow-up has become the norm for women who have been
treated for breast cancer. There are, however, significant
variations in the way this has been implemented across the
United States and the world.
Most published data compare mammographic surveillance and physical examination. This combination has been
the standard of care and is most frequently used in discussion of follow-up protocols. In contrast, mammography of the
treated breast for recurrence is seldom performed routinely
for patients who have had a mastectomy and breast reconstruction, with physical examination being the primary tool
to detect recurrence. MRI has been shown to be a valuable
tool in evaluating the reconstructed breast following mastectomy. MRI may also have an increasing role in the
detection of recurrence has mostly been via clinical symptoms and physical exam, often at a later stage. New imaging
modalities, such as magnetic resonance imaging (MRI), ultrasound (US), and positron emission mammography (PEM) are
changing the way we image the postsurgical breast. MRI,
coupled with physical exam and mammography, approaches
100% sensitivity and high specificity for the identification of
recurrent disease. We present a review of major academic
institutions imaging protocols and discuss the advantages
of including MRI in traditional mammographic and clinical
exams.
treatment of patients who have undergone BCT or women

with an originally mammographically occult primary breast
cancer.
When discussing the follow-up examination of a patient
previously treated for breast cancer, we must remember that
the ultimate goal of imaging following a breast cancer
diagnosis is to detect recurrence or new primary cancer
before it is clinically detectable.
Surveillance of the contralateral breast is essentially a
type of screening exam, except that it is targeted to a
high-risk group. As a planned medical intervention, this
is often referred to as surveillance. However, in a patient
with breast conservation, the contralateral breast is being
screened.
There is very little data available to make specific guidelines about surveillance of the patient previously treated for
breast cancer. Therefore, a review of the limited literature
and surveillance protocols at a few selected academic centers
will be presented.
Effectiveness of Mammography
Mammography has been proven to be efficacious as a

screening modality in detecting breast cancer recurrence,
with various papers showing a 50% detection rate of recurrent tumors by mammography and the rest by physical
examination.13-15 Depending on the type of prior surgery,
there may be confounding and distracting features both on
physical examination and on imaging, and the number of
false-positive examinations may be higher than those for the
general population.
Recurrent tumors detected by surveillance mammography
are in general smaller and less invasive than those found
during clinical examination, as demonstrated by Lu and
colleagues.16 This systematic review compared isolated locoregional recurrence or contralateral cancers on survival.
The authors demonstrated better overall survival for mam-
From the H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL, and
University of South Florida, Tampa, FL.
Address reprint requests to Chris Flowers, MBBS, H. Lee Moffitt Cancer Center and
Research Institute, 12902 Magnolia Dr., Tampa, FL 33612; email: chris.flowers@
moffitt.org.
1092-9118/10/1-10
59
FLOWERS, MOONEY, AND DRUKTEINIS
mographically detected recurrence, or in asymptomatic patients. They also demonstrated an absolute reduction in
mortality of 17% to 28% in all patients with breast cancers
found early.
A recent review in the United Kingdom for the Health
Technology Assessment by Robertson and colleagues17 also
demonstrated that for screening for local recurrence, surveillance mammography sensitivity ranged from 64% to 67%
and specificity ranged from 85% to 97%. They also found
that for MRI, sensitivity ranged from 86% to 100% and
specificity was 93%. They concluded that mammography is
associated with a high sensitivity and specificity but that
MRI is the most accurate test for detecting local recurrence.
Developing Role of MRI Post-BCT Surveillance
MRI is emerging as a valuable tool in the treatment of

patients with a diagnosis of breast cancer. Conventional
imaging techniques of mammography and US can be difficult to interpret as a result of post-therapeutic changes, and
some centers are including breast MRI in the surveillance of
women who are post-BCT, as surveillance MRI in this
population has been shown to detect malignancy in 12%.18
In the Robertson review,17 MRI was found to be the most
accurate test for detecting ipsilateral tumor recurrence and
contralateral breast cancer in women previously treated for
primary breast cancer. MRI in combination with mammography, US, and clinical examination has been reported to
have a sensitivity as high as 100% and specificity of 89% for
detecting contralateral breast cancer in patients who have
undergone BCT.19 MRI has been shown to be extremely useful
in differentiating scar tissue from tumor recurrence, particularly in showing that nonenhancing areas have a high
negative predictive value for malignancy (88% to 96%).20,21
The use of MRI in the monitoring of the post-BCT breast
is variable and often at the discretion of the ordering
physician or surgeon. The American College of Radiology
practice guidelines state that breast MRI may be useful for
women with a prior history of breast cancer and suspicion of
recurrence when clinical, mammographic, and/or sonographic findings are inconclusive.22 Although women with
a previous diagnosis of breast cancer are at increased risk of
a second diagnosis, an American Cancer Society panel concluded that the increased risk due to a personal history of
breast cancer alone does not justify a recommendation for
overall screening in women post-BCT at the present time.23
Neither National Comprehensive Cancer Network nor
ASCO guidelines consider MRI for surveillance. Changes

occurring after BCT on MRI are similar to those described
on mammography and include architectural distortion,
edema, skin thickening, and occasionally a seroma. These
findings are typical and may stabilize or continue to decrease over time. Enhancement at the lumpectomy site can
be an expected finding depending on the time interval from
BCT. A minimal or small focal area of thin linear enhancement can be seen for up to 18 months, and in some cases
even longer.24,25 On follow-up imaging, enhancement should
demonstrate stability or decrease over time. Fat necrosis is
commonly present in the post-BCT breast and can be a
challenging pitfall, often leading to biopsy. Enhancement is
variable in appearance and kinetic patterns. Biopsy may be
avoided if the MRI signal is similar to that of adjacent fat on
unenhanced images and the lesion shows no internal enhancement. For this reason, nonfat-saturated T1-weighted
sequences should be included on all breast MRI patients.
Mass-like enhancement at the lumpectomy site is always
suspicious and requires biopsy. Similarly, clumped, new, or
increasing areas of nonmass-like enhancement (not associated with fat necrosis) should be considered suspicious.26
Much of the utilization of MRI is physician- or surgeondependent. Many centers stagger MRI and mammograms at
6-month intervals. The reasoning behind this is that the
breast is surveyed in a different modality every 6 months. At
our institution, we prefer that MRI be performed 1 year
following surgery, so that it coincides with the bilateral
mammographic exam. This clarifies indeterminate findings
in the contralateral breast and allows normal postsurgical/
radiation therapy enhancement to decrease. Either approach is acceptable, as there are no specific data-driven
guidelines or a consensus on recommendations for MRI
surveillance intervals in this population. Despite breast
cancer survivors having a high risk of second cancer,
Punglia and Hassett27 recommend against routine MRI,
suggesting selective use by lifetime risk estimates.
Breast Conservation Surveillance
All agree that surveillance of patients who are post-BCT is

necessary. However, there is a lack of consensus on how
frequently and for how long we should provide routine
surveillance after treatment. Subsequently, there are several ways that surveillance can take place and may vary
from institution to institution (Table 1). The majority of
centers surveyed for this article have a common initial
2-year pathway, changing to annual follow-up at year 3.
KEY POINTS
Benign Findings Mimicking Malignancy
There are many benign findings that can mimic malignancy in the postsurgical breast, but essentially they may be
broken down into two categories: a) developing microcalcification, and b) variants of fat necrosis. The aggressive variant of fat necrosis can produce a very hard, spiculate scar.
The matrix may also calcify without the characteristic
dystrophic calcification that is normally associated with it.
This variant often produces a fine, lace-like, rapidly developing calcification, prompting biopsy.
Calcifications may develop in a patient who has been
treated for ductal carcinoma in situ (DCIS), especially if
there was noncalcified disease followed by radiation therapy. There may be an increase in necrosis due to apoptosis of
60
Mammographic surveillance has a high sensitivity

and specificity for recurrent malignancy.
Magnetic resonance imaging is the most sensitive
and accurate tool for evaluation of the postoperative
breast.
Physical examination has an important role in surveillance.
Robust conclusions cannot be made because of the
limited evidence base.
Further research comparing surveillance mammography and newer diagnostic tests is required.
SURVEILLANCE FOLLOWING BREAST CANCER DIAGNOSIS

Table 1. Examples of Breast Conservation Surveillance Protocol by Institution
Institution
Mammography
Physical Examination
M. D. Anderson Cancer Center

University of California, Los Angelas
Moffitt Cancer Center
Brigham and Womens Hospital and
Dana-Farber Cancer Center
At 6 mo, then annual diagnostic for 5 yr; screening from year 6

Every 6 mo for 5 yr, then annual screening
Every 6 mo for 2 yr, then annual diagnostic
Every 6 mo for 2 yr, then annual diagnostic until yr 10
At 6 mo after completion of radiotherapy, then annual diagnostic
Every 6 mo for 5 yr, then annual

Left to primary care
Every 6 mo for 5 yr
Abbreviations: mo, months; yr, years.
cells in the ducts, with subsequent developing ductal microcalcifications. Some centers advocate a postsurgical, preradiation therapy mammogram of the treated side as a
baseline for just this purpose.
In general, calcifications developing in the first 2 years are
dystrophic and thereafter are more likely as a result of
recurrent disease.
For palpable findings, targeted breast US is the initial
imaging modality of choice. Findings are usually related to
scarring or to a lump caused by development of fat necrosis.
Imaging Appearances of Recurrent Disease
In general, a recurrent cancer shows similar appearances

to the original malignancy, such that a patient with DCIS
will have developing microcalcifications in the treated breast,
and high-grade tumors with DCIS may present as a developing focal asymmetry containing calcifications (Fig. 1).
We recommend that spot magnification views be part of

the initial post-lumpectomy mammogram for patients in
whom calcified DCIS is present, whether invasive or in situ.
This will act as the new baseline exam for comparison.
Developing microcalcifications usually ring alarm bells but
can be commonly seen in patients prone to developing fat
necrosis.
Special types of tumors, such as angiosarcomas, require
more specialized follow-up, which can be a combination of
Doppler US and MRI. The permeative vascular nature of the
recurrent tumor is best seen on MRI (Fig. 2).
Mastectomy without Reconstruction
A patient who has been treated by mastectomy has a

contralateral breast cancer risk of approximately 7%, and so
surveillance of the treated breast is almost entirely covered
by physical examination, with targeted US scan of any
palpable area. The contralateral breast will continue to be
monitored by a regular annual mammographic examination,
with additional imaging being reserved for patients who are
symptomatic or who have a palpable finding.
Male Survivors of Breast Cancer
These are the forgotten survivors of cancer. The surveillance protocol for a male patient should be the same as that
for a woman who has had a mastectomy without reconstruction. Most recurrences in men should be palpable, hence the
importance of a regular physical exam, and possibly breast
awareness on the part of the man. If the patient has had a
Fig. 1. Example of recurrent DCIS with a sharp margin corresponding to the edge of the radiotherapy boost site.
Abbreviation: DCIS, ductal carcinoma in situ.
Fig. 2. MRI image showing enhancing nodule adjacent to seroma/

lumpectomy site. Patient is 5 years post-surgery and 2 months
post-completion of 5 years of tamoxifen.
Abbreviation: MRI, magnetic resonance imaging.
61
lumpectomy rather than mastectomy, then there is still no

reason to vary the protocol from that of a female patient.
The Reconstructed Breast
The reconstructed breast has not been the subject of

imaging surveillance until more recently. Some advise the
use of mammographic surveillance of transverse rectus
abdominus myocutaneous (TRAM) flap reconstructions,
sometimes called tramograms, which can be performed
just like any other surveillance mammogram.28,29 The findings of recurrent cancer are typical and have been reported
before they become palpable, advancing the stage of diagnosis. Others advocate for only selecting women with a high
risk-of-recurrence score to have tramograms (Fig. 3).
The most sensitive test for the reconstructed breast is
MRI, although it is a relatively expensive test that is not
always covered by health insurance in this situation, even
Fig. 4.
gram.
Developing microcalcifications in scar on routine mammo-
though it is the most effective test. According to Devon,30

MRI is useful in the detection of locally recurrent tumors in
patients who have undergone breast reconstruction with a
TRAM flap. It detects cancer recurrences at an earlier stage
than noncompliant annual surveillance.
Improved access to dedicated breast MRI is making evaluation of the reconstructed breast more affordable and
accessible. Early detection may potentially save health care
dollars by catching recurrence before loco-regional spread,
but there are no data to show that this improves survival.
False-Positive Examinations on Follow-up
Fig. 3. Tramogram image showing TRAM recurrence in the right

XCCL mammogram. Lesion was not initially not detected until the
mammogram finding showed the lesion.
Abbreviations: TRAM, transverse rectus abdominus myocutaneous;
XCCL, exaggerated craniocaudal view.
62
There are frequently false-positive clinical findings after

reconstruction, mainly with palpable lumps. Palpable lesions should be first evaluated with US. The palpable lumps
are generally multiple, appear in the upper outer quadrant
or 6 oclock positions, and are caused by fat necrosis. This is
more common in women with implant reconstruction than
SURVEILLANCE FOLLOWING BREAST CANCER DIAGNOSIS
Fig. 5. Palpable mass near scar; US shows typical features of fat

necrosis.
Abbreviation: US, ultrasound.
TRAM flaps in our practice but may vary according to the

distributions of types of reconstruction in your group. Mammography often demonstrates early dystrophic microcalcifications, which can simulate an aggressive recurrent DCIS
(Fig. 4 and Fig. 5).
Bilateral Mastectomy with Reconstruction
Little is known about the utility of MRI as a diagnostic or

surveillance tool after bilateral mastectomy with reconstruction. The current role of MRI in treating patients who have
had a mastectomy is primarily as a diagnostic tool when
clinical findings need clarification. Most recurrences adjacent to an implant are clinically palpable, and there is little
evidence to support the use of MRI in asymptomatic postmastectomy with reconstruction patients. Patients who
have undergone reconstruction with TRAM or deep inferior
epigastric perforator flaps can be evaluated with MRI;
however, routine surveillance is not common practice, as the
risk of nonpalpable recurrence is extremely low.30 MRI has
been found useful in differentiating benign from malignant
findings in patients with palpable abnormalities or pain
after TRAM reconstruction (Fig. 6).30
Possible Developing Methods for
Recurrence Detection
To detect additional tumors or metastatic disease at an

early stage, advanced imaging techniques, including diagnostic mammographic images, focused breast US, breast
Fig. 6. Vague nodularity on physical exam. US of nodularity

showed no correlation. MRI demonstrates obvious local recurrence
and metastases.
Abbreviations: US, ultrasound; MRI, magnetic resonance imaging.
MRI, and positron emission tomography imaging is often

required. Breast-specific gamma imaging, PEM, and molecular breast imaging may have a problem-solving role, especially in the dense breast or where there are complications
from ruptured silicone implants. Radiologists who have
completed advanced training in breast cancer detection
should interpret these imaging techniques. Imaging should
be coupled with clinical exams by specialty-trained clinicians familiar with signs and symptoms of recurrence and
should occur at set time intervals. The downside of these
newer techniques is the radiation dose from the radioisotopes currently used, although new techniques are focused
on reducing the radiation dose.
Advanced reconstruction and radiation techniques can be
difficult to interpret unless communication between members of the health care team is timely and accurate, and
the radiologist is familiar with expected appearances. To
achieve the goal of recurrence detection in the early stages,
before it is clinically evident, radiologists and clinicians
must be familiar with the behavior of subtypes of breast
cancer and different surgical/medical/oncoplastic approaches to their treatment.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Chris I. Flowers*
Blaise P. Mooney*
Jennifer S. Drukteinis*
REFERENCES
1. Voogd AC, van Tienhoven G, Peterse HL, et al. Local recurrence after
breast conservation therapy for early stage breast carcinoma: detection,
treatment, and outcome in 266 patients. Cancer. 1999;85:437-446.
2. Hietanen P, Miettinen M, Makinen J. Survival after first recurrence in
breast cancer. Eur J Cancer Clin Oncol. 1986;22:913-919.
3. Nomura Y, Tsutsui S, Murakami S, et al. Prognostic impact of second

cancer on the survival of early breast cancer patients. Int J Oncol. 1999;14:
1103-1109.
4. Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a
randomized trial comparing total mastectomy, lumpectomy, and lumpectomy
63

plus irradiation for the treatment of invasive breast cancer. N Engl J Med.
2002;347:1233-1241.
5. Fisher B, Jeong JH, Anderson S, et al. Twenty-five-year follow-up of a
randomized trial comparing radical mastectomy, total mastectomy, and total
mastectomy followed by irradiation. N Engl J Med. 2002;347:567-575.
6. Fisher B, Redmond C, Poisson R, et al. Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or
without irradiation in the treatment of breast cancer. N Engl J Med.
1989;320:822-828.
7. Jacobson JA, Danforth DN, Cowan KH, et al. Ten-year results of a
comparison of conservation with mastectomy in the treatment of stage I and
II breast cancer. N Engl J Med. 1995;332:907-911.
8. van Dongen JA, Bartelink H, Fentiman IS, et al. Factors influencing
local relapse and survival and results of salvage treatment after breastconserving therapy in operable breast cancer: EORTC trial 10801, breast
conservation compared with mastectomy in TNM stage I and II breast cancer.
Eur J Cancer. 1992;28A:801-805.
9. Veronesi U, Cascinelli N, Mariani L, et al. Twenty-year follow-up of a
randomized study comparing breast-conserving surgery with radical mastectomy for early breast cancer. N Engl J Med. 2002;347:1227-1232.
10. Veronesi U, Luini A, Del Vecchio M, et al. Radiotherapy after breastpreserving surgery in women with localized cancer of the breast. N Engl
J Med. 1993;328:1587-1591.
11. Kurtz JM, Amalric R, Brandone H, et al. Local recurrence after
breast-conserving surgery and radiotherapy. Frequency, time course, and
prognosis. Cancer. 1989;63:1912-1917.
12. Recht A, Silen W, Schnitt SJ, et al. Time-course of local recurrence
following conservative surgery and radiotherapy for early stage breast cancer.
Int J Radiat Oncol Biol Phys. 1988;15:255-261.
13. Fowble B, Solin LJ, Schultz DJ, et al. Breast recurrence following
conservative surgery and radiation: patterns of failure, prognosis, and pathologic findings from mastectomy specimens with implications for treatment.
14. Joseph E, Hyacinthe M, Lyman GH, et al. Evaluation of an intensive
strategy for follow-up and surveillance of primary breast cancer. Ann Surg
Oncol. 1998;5:522-528.
15. Orel SG, Troupin RH, Patterson EA, et al. Breast cancer recurrence
after lumpectomy and irradiation: role of mammography in detection. Radiology. 1992;183:201-206.
16. Lu WL, Jansen L, Post WJ, et al. Impact on survival of early detection
of isolated breast recurrences after the primary treatment for breast cancer:
a meta-analysis. Breast Cancer Res Treat. 2009;114:403-412.
17. Robertson C, Ragupathy SK, Boachie C, et al. Surveillance mammography for detecting ipsilateral breast tumour recurrence and metachronous
64
contralateral breast cancer: a systematic review. Eur Radiol. 2011;21:24842491.

18. Brennan S, Liberman L, Dershaw DD, et al. Breast MRI screening of
women with a personal history of breast cancer. AJR Am J Roentgenol.
2010;195:510-516.
19. Viehweg P, Rotter K, Laniado M, et al. MR imaging of the contralateral
breast in patients after breast-conserving therapy. Eur Radiol. 2004;14:402408.
20. Nunes LW, Schnall MD, Orel SG. Update of breast MR imaging
architectural interpretation model. Radiology. 2001;219:484-494.
21. Schnall MD, Blume J, Bluemke DA, et al. Diagnostic architectural and
dynamic features at breast MR imaging: multicenter study. Radiology.
2006;238:42-53.
22. American College of Radiology. ACR Practice Guideline for the Performance of Contrast Enhanced Magnetic Resonance Imaging (MRI) of the
Breast (revision). 2008.
23. Saslow D, Boetes C, Burke W, et al. American Cancer Society guidelines for breast screening with MRI as an adjunct to mammography. CA
Cancer J Clin. 2007;57:75-89.
24. Heywang-Kobrunner SH, Schlegel A, Beck R, et al. Contrast-enhanced
MRI of the breast after limited surgery and radiation therapy. J Comput
Assist Tomogr. 1993;17:891-900.
25. Li J, Dershaw DD, Lee CH, Joo S, Morris EA. Breast MRI after
conservation therapy: usual findings in routine follow-up examinations. AJR
Am J Roentgenol. 2010;195:799-807.
26. Drukteinis JS, Gombos EC, Raza S, et al. MR imaging assessment of
the breast after breast conservation therapy: distinguishing benign from
malignant lesions. Radiographics. 2012;32:219-234.
27. Punglia RS, Hassett MJ. Using lifetime risk estimates to recommend
magnetic resonance imaging screening for breast cancer survivors. J Clin
Oncol. 2010;28:4108-4110.
28. Eidelman Y, Liebling RW, Buchbinder S, et al. Mammography in the
evaluation of masses in breasts reconstructed with TRAM flaps. Ann Plast
Surg. 1998;41:229-233.
29. Helvie MA, Bailey JE, Roubidoux MA, et al. Mammographic screening
of TRAM flap breast reconstructions for detection of nonpalpable recurrent
cancer. Radiology. 2002;224:211-216.
30. Devon RK, Rosen MA, Mies C, et al. Breast reconstruction with a
transverse rectus abdominis myocutaneous flap: spectrum of normal and
abnormal MR imaging findings. Radiographics. 2004;24:1287-1299.
31. Lee JM, Georgian-Smith D, Gazelle GS, et al. Detecting nonpalpable
recurrent breast cancer: the role of routine mammographic screening of
transverse rectus abdominis myocutaneous flap reconstructions. Radiology.
2008;248:398-405.
Advanced Imaging Techniques for the

Detection of Breast Cancer
By Maxine Jochelson, MD
Overview: Mammography is the only breast imaging examination that has been shown to reduce breast cancer mortality.
Population-based sensitivity is 75% to 80%, but sensitivity in
high-risk women with dense breasts is only in the range of
50%. Breast ultrasound and contrast-enhanced breast magnetic resonance imaging (MRI) have become additional standard modalities used in the diagnosis of breast cancer. In
high-risk women, ultrasound is known to detect approximately
four additional cancers per 1,000 women. MRI is exquisitely
sensitive for the detection of breast cancer. In high-risk
women, it finds an additional four to five cancers per 100
women. However, both ultrasound and MRI are also known to
lead to a large number of additional benign biopsies and
AMMOGRAPHY IS the only breast imaging examination shown to reduce breast cancer mortality, with
a population-based sensitivity of 75% to 80%. Current
screening guidelines for normal-risk women (15% lifetime
risk) recommend starting screening at 40 and continuing
until life expectancy is less than 5 years. Women at higher
risks (intermediate 15% to 20% and high 20%) begin
screening at an earlier age, generally 10 years earlier than
the youngest family member presented with her breast
cancer. Sensitivity of mammography in high-risk women
with dense breasts is only in the range of 50%. Both
ultrasound and MRI are standard examinations used to
improve on mammographic sensitivity; each has its own
flaws. New technologies have been and continue to be
developed to improve the breast cancer detection rate. These
include improvements on the standard techniques (mammography, ultrasound, and MRI) as well as new platforms
for breast imaging developed from body imaging tools, which
include CT and radionuclide breast imaging. This article
will review the most salient newer breast imaging technologies, as well as their potential roles.
Mammography and Mammography-Based Techniques
Digital mammography (FFDM) has almost entirely replaced analog (film-screen) mammography. It has been
shown to be more sensitive in women younger than 50 with
dense breast tissue.1 However, overall sensitivity remains
the same as with analog. The advent of digital mammography has enabled more advanced techniques, which are
added to the digital platform. These include digital breast
tomosynthesis (DBT) and contrast-enhanced spectral mammography (CESM).
Tomosynthesis
In 2011, DBT was called the most exciting new technology

in breast imaging. It is U.S. Food and Drug Administration
(FDA)-approved for performance with a screening digital
mammogram. In 1997, it was stated that DBT allowed the
radiologist to see through the structured noise of normal
breast tissue to improve the detection and characterization of early-stage breast cancer.2 By seeing through this
noise, additional lesions may be detected, and better margin
analysis can be performed. Additionally, the radiologist can
short-term follow-up examinations. Many new breast imaging

tools have improved and are being developed to improve on
our current ability to diagnose early-stage breast cancer.
These can be divided into two groups. The first group is those
that are advances in current techniques, which include digital
breast tomosynthesis and contrast-enhanced mammography
and ultrasound with elastography or microbubbles. The other
group includes new breast imaging platforms such as breast
computed tomography (CT) scanning and radionuclide breast
imaging. These are exciting advances. However, in this era of
cost and radiation containment, it is imperative to look at all of
them objectively to see which will provide clinically relevant
additional information.
determine that a mass was merely a confluence of shadows. Specificity is, therefore, theoretically enhanced. To date
no large prospective trials have been reported regarding the
ability of DBT to screen for cancer as a solo imaging tool.
What has been well documented is its ability to reduce the
need for callbacks.3,4 More recently, DBT images were
compared to routine spot films and found to show equivalent
mass characterization. DBT found seven additional cancers
along with five additional false-positive findings.5 The detection of clustered microcalcifications has been more problematic. Initially, detection was inferior with DBT because
of the very thin (1 mm) slices and the blurring inherent to
tomosynthesis. Spangler and colleagues showed superior
sensitivity (84% vs. 75%) and specificity (71% vs. 64%) with
FFDM, which also detected more cancers with calcifications
but the differences were not significant.6 More recently,
Destounis and colleagues compared the two techniques in
103 patients and found them equivalent.7 Gur and colleagues compared FFDM alone to FFDM combined with
DBT in 125 patients in a retrospective evaluation. There
were both more true-positives and true-negatives with DBT
with an overall 16% performance improvement (p 0.01).8
There are several limitations: 1) Radiation dose is twice
that of FFDM. The thought is that the reduction in callbacks
will reduce the effect of that additional radiation, as spot
films generate a higher dose than DBT. However, most
women who are screened do not get called back for additional views. That being said, the dose does fall within the
MQSA guidelines; 2) DBT only evaluates anatomy, while
many of the other emerging technologies also evaluate
physiology; 3) the interpretation of DBT takes much longer
than that of FFDM, limiting throughput; 4) reimbursement
is limited or absent.
This is a promising technology without a defined role
(Fig. 1). Possibilities include routine screening for everyone,
high-risk screening, and as a diagnostic tool after abnormal
From the Memorial Sloan-Kettering Cancer Center, New York, NY.

Address reprint requests to Maxine Jochelson, MD, Memorial Sloan-Kettering Cancer
Center, 300 E. 66th St., New York, NY 10065; email:jochelsm@mskcc.org.
1092-9118/10/1-10
65
MAXINE JOCHELSON
Fig. 1. A clip from the newspaper regarding 3-day as a new

technology for screening.
screening. There are multiple ongoing studies to better

define its strengths and weaknesses and to determine its
ideal role.
Contrast-Enhanced Mammography
Contrast-enhanced breast MRI functions by both defining

anatomic abnormalities and evaluating physiologic changes
related to the development of neovascularity, which are
visualized by their rapid contrast enhancement in cancers.
MRI has proven to be exquisitely sensitive for the detection
of breast cancer, with a sensitivity range from 79% to
98%.9,10 A meta-analysis has shown that MRI also detects
mammographically occult multicentric or multifocal disease
in approximately 16% of patients with known breast cancer.9
Although breast MRI is extremely sensitive, specificity is
more limited, leading to additional workups and benign
biopsies. Additionally, it is expensive, and good quality
KEY POINTS
66
Mammography is the only breast imaging exam

known to reduce breast cancer mortality, but sensitivity is limited particularly in high-risk women with
dense breast tissue.
The digital mammography platform has allowed for
developments such as tomosynthesis and contrastenhanced mammography, which have the potential to
improve sensitivity.
Ultrasound detects additional cancers but at the cost
of many benign biopsies. Potential advances in ultrasound technology include automated whole-breast
scanning, elastography, and microbubbles.
Breast MRI remains the most sensitive examination
for the detection of early-stage breast cancer in a
high-risk screening population.
Radionuclide breast imaging is able to detect breast
cancers independent of breast density, but its high
extramammary radiation dose precludes use in
screening.
breast MRI is not universally available. Patients with pacemakers, certain aneurysm clips or other metallic hardware,
and severe claustrophobia are unable to undergo MRI.
Contrast-enhanced mammography has been developed to
potentially duplicate the capacity of MRI to detect and stage
breast cancer with the use of both anatomy and physiology.
It has been investigated both as an adjunct to mammography and an alternative to MRI.11,12 It uses the same iodinated contrast used for CT in the same doses.
Hardware and software adaptations to digital mammography units that automate a dual energy technique have
been developed. This technology, known as ContrastEnhanced Spectral Mammography (CESM), received FDA
approval in 2011. It provides two images in each view: a
low-energy image, which is below the K-edge of iodine
(33.2keV), and a high-energy image, just above. The two
images are combined and processed so that the background
breast tissue is essentially subtracted out, maximizing
the ability to see areas of enhancement (Fig. 2). There is a
20% additional radiation dose, which is the equivalent of
one extra mammographic view. The examination takes approximately 10 minutes to perform and is only slightly
longer than a mammogram to read. It is well tolerated by
patients.
Dromain and colleagues recently reported a comparison
of contrast-enhanced digital mammography (CEDM) plus
mammography with mammography alone and mammography plus ultrasound in 142 lesions in 120 patients. Sensitivity for CEDM plus mammography was 93% compared with
78% (p 0.001) for mammography alone. Specificity was
unchanged. There was improvement in sensitivity and specificity between CEDM plus mammography and ultrasound
plus mammography, but it was not significantly better.
Additionally, all 23 multifocal lesions were detected.13
Jochelson reported early results of a comparison of CEDM
with breast MRI in 26 patients with known breast cancer.
The two examinations each detected 25 of 26 index tumors
compared with the 22 detected by mammography alone.
MRI was more sensitive than CEDM for the detection of
additional lesions (7 vs. 5), but there were no false-positive
CEDM examinations and seven false-positive MRIs.14 Results on additional patients will be reported later this year.
The early data for CESM are promising but require
confirmation. Screening studies need to be performed. If
larger studies confirm early results, CESM could become
available as a possible poor womans MRI. Larger multiinstitutional trials are ongoing.
Ultrasound
Targeted breast ultrasound is of value in further defining

mammographic and MRI abnormalities as well as evaluating palpable lesions. Ultrasound-guided core biopsies are
an effective means of obtaining histologic diagnosis. Ultrasound is inexpensive, utilizes no radiation, and is widely
available.
The use of ultrasound for screening is more problematic.
Multiple studies, including a large ACRIN trial evaluating
the utility of ultrasound screening in high-risk patients,
have shown that ultrasound will detect additional cancers
in three to four of 1,000 patients (as compared with MRI,
which detects additional cancers in four to five of 100
patients).15,16 In doing so, a large number of additional
lesions are detected leading to short-interval follow-ups or
IMAGING TECHNIQUES FOR BREAST CANCER
Fig. 2. Three images of the left breast

from left to right: normal left mammogram; contrast-enhanced mammography showing extensive enhancement
caused by multifocal breast cancer in
the lower half of the breast; breast
MRI also showing multifocal cancer.
Contrast-enhanced mammogram corresponds well to the MRI.
biopsies with a positive predictive value (PPV) of only 9%

to 10%. Scans are operator-dependent and, therefore, less
reproducible than other modalities. A lesion not detected is
not documented.
To eradicate operator dependence, automated whole
breast ultrasound (AWBU) machines have been developed.
Images are reproducible, and there is 3D capability. Additionally, studies need not be interpreted in real time. Kelly
prospectively compared mammography alone to mammography plus AWBU in 4,419 women and showed an improved
diagnostic yield of 7.2 of 1,000 compared with 3.6 of 1,000
and equivalent PPV of 38%.17 Shin found a 92% detection
rate for lesions larger than 1.2 cm.18
Other advances have been geared toward improving the
sensitivity, specificity, and PPV. Color Doppler detects tumor vascularity. This improves specificity of B-mode ultrasound. Microbubbles have been injected as contrast agents
to improve the performance of Doppler. Although this can
improve sensitivity, specificity has been more problematic
because of its ability to detect smaller and not necessarily
malignant vessels. Liu has shown that there is good histologic correlation when using this contrast agent. Unfortunately, this technique is labor intensive and a technically
difficult procedure to perform. It is unlikely to be useful in
every day practice.
Elastography is a tool that has been developed to improve
the specificity of breast ultrasound. Because most breast
cancers are stiff, there is theoretically less displacement
on compression. There are two types of elastography. There
is more experience with static elastography, in which the
operator compresses the lesion with the special probe and
measures displacement. Unfortunately, this is also operator
dependent, and the results have been disappointing.19 Shear
strain elastography is being studied. The waves are emitted
perpendicular to the transducer, and the speed of their
traversing the tissue is measured with faster speeds occurring in hard tissue. This is less operator dependent, and
early results indicate that this may be the better way to

measure elasticity.
Although these are interesting attempts at improving
screening ultrasound, a great deal more work is necessary.
Radionuclide Breast Imaging
Radionuclide breast imaging evaluates physiology. Sestamibi and 18-FDG detect cancers by different mechanisms of
action. Their ability to detect lesions in the breast is independent of breast density, and with 18-FDG, it is independent of hormone status.
MIBI or Gamma Imaging
In the early 1990s, incidental breast cancers were seen in

patients having sestamibi cardiac scans. At that time, attempts at dedicated breast imaging had limited sensitivity
for small lesions because of the large collimators that were
distant from the breasts. Since then, new technology has
become available, wherein high resolution detectors are
used to mildly compress the breasts, which are positioned as
they are with mammography. There are two different systems: molecular breast imaging (MBI) based on a semiconductor base and breast-specific gamma imaging (BSGI),
which uses a scintillating crystal detector. The results are
similar and will be discussed together. Patients are ideally
done between days 2 and 14 of the menstrual cycle. They
receive 740 1,110 MBq. There is a 1-minute wait. Images
are then acquired for approximately 10 minutes per view.
Two views of each breast are provided. Studies show sensitivities of 91% to 96% overall but somewhat lower for
smaller lesions: 69% for lesions less than 5 mm in one
study20 and 89% for sub-centimeter lesions.21 That study
also had a specificity of only 60%no improvement over
MRI. Detection of additional foci of cancer on staging exams
is 9% to 10%, but in doing so there are a large number of
false positives. For example, Brem reported on 159 patients
67
MAXINE JOCHELSON
in whom there were 46 additional lesions, only 14 of which

were malignant.22
Positron Emission Mammography (PEM)
ing, and assessment of treatment response in patients receiving neoadjuvant chemotherapy.

MRI and Conclusions
The sensitivity of whole-body 18-FDG PET for detection of

a primary breast cancer is lowapproximately 40%. PEM
was developed to better evaluate the breasts. The scanner
again resembles a mammography machine. Twelve images
are generated for each view of the breast, providing a sort of
3D image. PEM is performed like whole-body PET: 4 6
hours of fasting, injection of about 10 mCi of FDG, followed
by a 1-hour rest period. Each view takes 10 minutes. PEM
can also be performed after a whole-body exam without
additional tracer. Multiple studies have shown sensitivities
of more than 90% (including DCIS) even for smaller lesions.23 Specificity is also reported to be more than 90%. In
large trials comparing PEM to MRI in patients with known
breast cancer, MRI was more sensitive on a lesion level in
the ipsilateral breast, more sensitive in the contralateral
breast, but less specific.24 PEM can be performed with
different tracers.
Radionuclide imaging is less expensive than MRI, and the
scanners themselves need less space. The limiting factor for
their use in routine screening is the high dose of radiation,
particularly to extramammary tissues: 50mGy to lower
large intestine for sestamibi and 59mGy to bladder was for
PEM. Vendors of both technologies are reporting promising
preliminary results using half the dose, but even those doses
may be too high to be used in yearly screening. Other
potential uses include preoperative staging, problem solv-
Breast MRI is the gold standard breast imaging test

against which all other new technologies are compared,
particularly in terms of sensitivity, which is more than 95%.
The absence of radiation is a major asset. Initial inability to
diagnose DCIS and low specificity has been limiting. Both
have improved over the years, but the perception of poor
specificity remains. MRI is clearly the ideal way to screen
women at high risk for breast cancer, but is too expensive to
be used in larger populations, and good quality MRI is not
universally available. As newer technologies become available, however, there are many lessons from the initial
experience with breast MRI that should be applied: 1) not all
MRI scanners/scans are equal, and although most of the
technologies discussed here are easier to perform, equipment performance standards should be defined. 2) There is
always a learning curve in the interpretation of new technologies. Minimal training requirements should be set. 3) A
vocabulary with standard terminology to describe abnormalities and what they mean (such as the BIRADS system)
should be developed. 4) In this era of needed cost and
radiation containment, we cannot do every test on every
woman. It is imperative that in addition to determining
sensitivity, specificity, and other factors, we also learn in
which situations these new technologies can do the most
good.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Maxine Jochelson*
REFERENCES
1. Pisano ED, Gatsonis C, Hendrick E, et al. Diagnostic performance of
digital versus film mammography for breast-cancer screening. N Engl J Med.
2005;353:1773-1783.
2. Niklason LT, Christian BT, Niklason LE, et al. Digital tomosynthesis in
breast imaging. Radiology. 1997;205:399-406.
3. Kopans D, Moore R. Digital breast tomosynthesis (DBT) NCI 3000women trial. RSNA. 2009.
4. Poplack SP, Tosteson TD, Kogel CA, Nagy HM. Digital breast tomosynthesis: Initial experience in 98 women with abnormal digital screening
mammography. AJR Am J Roentgenol. 2007;189:616-623.
5. Noroozian M, Hadjiiski L, Rahnama-Moghadam S, et al. Digital breast
tomosynthesis is comparable to mammographic spot views for mass characterization. Radiology. 2012;262:61-68.
6. Spangler ML, Zuley ML, Sumkin JH, et al. Detection and classification
of calcifications on digital breast tomosynthesis and 2D digital mammography: A comparison. AJR Am J Roentgenol. 2011;196:320-324.
7. Destounis S, Murphy P, Seifert P, et al. Clinical experience with digital
breast tomosynthesis in the characterization and visualization of breast
microcalcifications. Amer J Radiol. 2011;196:A1-A3.
8. Gur D, Bandos AI, Rockette HE, et al. Localized detection and classification of abnormalities on FFDM and tomosynthesis examinations rated
under an FROC paradigm. AJR Am J Roentgenol. 2011;196:737-741.
9. Morris EA, Liberman L, Ballon DJ, et al. MRI of occult breast carcinoma
in a high-risk population. AJR Am J Roentgenol. 2003;181:619-626.
10. Berg WA. Rationale for a trial of screening breast ultrasound: American College of Radiology Imaging Network (ACRIN) 6666. AJR Am J
Roentgenol. 2003;180:1225-1228.
11. Dromain C, Thibault F, Adler G. Dual energy contrast-enhanced digital
68
mammography: Preliminary Clinical Results 94th Scientific Assembly of the

RSNA 2007 Annual Meeting Program. 2007;326.
12. Diekmann F, Marx C, Jong R, et al. Diagnostic accuracy of contrastenhanced digital mammography as an adjunct to mammography. ECR. 2007.
13. Dromain C, Thibault F, Muller S, et al. Dual-energy contrast-enhanced
digital mammography: Initial clinical results. Eur Radiol. 2011;21:565-574.
14. Jochelson M, Sung J, Dershaw DD, Heerdt A. Bilateral-Dual Energy
Contrast-enhanced Digital Mammography: Initial Experience. Presented at
96th Annual RSNA Scientific Assembly and Annual Meeting; November 2010;
Chicago, IL.
15. Berg WA, Blume JD, Cormack JB, et al. Combined screening with
ultrasound and mammography vs mammography alone in women at elevated
risk of breast cancer. JAMA. 2008;299:2151-2163.
16. Kolb TM, Lichy J, Newhouse JH. Occult cancer in women with dense
breasts: Detection with screening US-diagnostic yield and tumor characteristics. Radiology. 1998;207:191-199.
17. Kelly KM, Dean J, Comulada WS, Lee SJ. Breast cancer detection
using automated whole breast ultrasound and mammography in radiographically dense breasts. Eur Radiol. 2010;20:734-742.
18. Shin HJ, Kim HH, Cha JH, et al. Automated ultrasound of the breast
for diagnosis: Interobserver agreement on lesion detection and characterization. AJR Am J Roentgenol. 2011;197:747-754.
19. Kohr J, Sung, J, Malak, S, et al. Impact of elastography on assessing
the likelihood of malignancy of ultrasound lesions. Poster presented at RSNA
97th Annual Meeting; 2011; Chicago, IL.
20. Hruska CB, Boughey JC, Phillips SW, et al. Scientific Impact Recognition Award: Molecular breast imaging: A review of the Mayo Clinic experience. Am J Surg. 2008;196:470-476.
IMAGING TECHNIQUES FOR BREAST CANCER

21. Brem RF, Floerke AC, Rapelyea JA, et al. Breast-specific gamma
imaging as an adjunct imaging modality for the diagnosis of breast cancer.
Radiology. 2008;247:651-657.
22. Brem RF, Shahan C, Rapleyea JA, et al. Detection of occult foci of
breast cancer using breast-specific gamma imaging in women with one
mammographic or clinically suspicious breast lesion. Acad Radiol. 2010;17:
735-743.
23. Berg WA, Weinberg IN, Narayanan D, et al. High-resolution fluorodeoxyglucose positron emission tomography with compression (positron emission mammography) is highly accurate in depicting primary breast cancer.
Breast J. 2006;12:309-323.
24. Berg WA, Madsen KS, Schilling K, et al. Breast cancer: Comparative
effectiveness of positron emission mammography and MR imaging in presurgical planning for the ipsilateral breast. Radiology. 2011;258:59-72.
69
UPDATE OF THE OXFORD OVERVIEW: NEW

INSIGHTS AND PERSPECTIVES IN THE ERA OF
PERSONALIZED MEDICINE
CHAIR
Harold J. Burstein, MD, PhD
Boston, MA
SPEAKERS
Jonas Bergh, MD, PhD
Karolinska Institutet, University Hospital
Stockholm, Sweden
Kathleen I. Pritchard, MD
Sunnybrook Health Sciences Center
Toronto, ON, Canada
Update of the Oxford Overview: New

Insight and Perspectives in the Era of
Personalized Medicine
By Kathleen I. Pritchard, MD, Jonas Bergh, MD, PhD, and Harold J. Burstein MD, PhD
Overview: There is great appreciation for the heterogeneity

of breast cancers, particularly of hormone-receptor positive
breast cancers. A goal of modern oncology managing such
heterogeneity is to determine how to individualize therapy
based on the specific pathological and biological features of a
given tumor. Two distinctive clinical literatures exist to guide
treatment of hormone-receptor-positive breast cancer. The
Oxford Overview, a seminal meta-analysis effort, has recently
been updated, and suggests that nearly all patients with
ER-positive tumors benefit from adjuvant endocrine therapy.
HE OXFORD Overview of the Early Breast Cancer

Trialists Collaborative Group (EBCTCG) dates back to
1984 when investigators responsible for trials of systemic
adjuvant systemic therapy from all around the globe met
initially, not in Oxford, but at Heathrow Airport to examine
the first meta-analysis of adjuvant systemic therapy trials.
The Overview has always involved collaboration between
the Oxford Secretariat led by Sir Richard Peto and a
consortium of investigators. A series of distinguished past
chairs have included I. Craig Henderson, MD, a prominent
breast cancer medical oncologist initially from Harvard
University and later the University of California San Francisco Helen Diller Family Comprehensive Center, and
William C. Wood, MD, then Chief of Surgery at Emory
University. The EBCTCG is currently chaired by Kathy
Pritchard, MD, of Toronto and Martine Piccart, MD, PhD,
of Brussels. The Steering Committee and Executive of the
EBCTCG both of which include members of the Oxford
Secretariat and clinical investigators (the Trialists) extensively meet, teleconference, and e-mail to bring analyses
and publications to fruition. Between 2005 and 2011,
data collection and analyses have resulted in five major
publications.1-5
The Overview concept dating back to 1984 has not
changed. Collaboration between physicians and biostatisticians based in Oxford and around the world was sought,
built, and sustained. Data were initially collected from all
randomized trials of systemic adjuvant and, later, localregional therapy. The Overview methodology involves the
collection of individual patient data, which includes a wide
variety of items such as dates of randomization and treatment allocation. Patients can be stratified by age, node
status, and other criteria and the log-rank statistics from
each trial are combined to give an overall estimate of the
effect of different treatments either in the whole patient
population or in various stratified subsets.
Outcomes include recurrence, which can be adjusted to
include or exclude contralateral breast cancers and deaths.
Deaths from unknown causes are usually included with
deaths from breast cancer unless specifically stated otherwise. Recurrences can be divided into local and/or distant
with and without contralateral breast cancers. The
EBCTCG has also been interested in collecting information
on deaths from cardiac events, stroke, and other cancers.
In addition, the overview finds that nearly all subsets of

patients with ER-positive tumors also benefit from modern
adjuvant chemotherapy regimens. Meanwhile, retrospective
subset analyses of specific trials or populations suggests that
the benefits of chemotherapy are not so uniform, and in
particular that molecular diagnostics assays can identify patients who do not warrant chemotherapy. This article will
highlight recent data and controversies in personalizing adjuvant breast cancer therapy.
In 1984, the Oxford Overview showed unequivocally for

the first time that tamoxifen improved survival and that
cyclophosphamide, methotrexate, and 5-fluorouracil (CMF)
chemotherapy improved survival. It was also shown that
ovarian ablation, which had been mainly tested in small
underpowered trials with nonsignificant results, did, in and
of itself, improve overall survival, particularly in women
whose tumors had positive estrogen receptors (ER). By 1990,
it became clear that 5 years of tamoxifen was better than 1
or 2 years and that tamoxifen effects were greater in women
with ER-positive cancers. It was first shown in 1990 that
tamoxifen reduced the rate of contralateral breast cancer
and that chemotherapy was effective in both older and
younger women.
By 1995, the huge effect of 5 years of tamoxifen was
clearly demonstrated, and it was obvious from both direct
and indirect comparisons within the Overview that 5 years
of tamoxifen was superior to shorter durations of treatment.
It was seen for the first time that tamoxifen prevented
contralateral breast cancers only in women whose initial
tumors were ER positive. That year, the Overview also
demonstrated that anthracycline-containing regimens, at
least when given in higher dosages, were better than CMFtype chemotherapy.
By 2000, the Overview was able to report on long-term
results, such as 15-year outcomes with chemotherapy, demonstrating sustained benefit in older and in younger women.
There was great controversy at this time suggesting that,
particularly in the CMF trials, the effects of chemotherapy
might be greater in women with ER-negative tumors than in
those with ER-positive tumors, a controversy which continues to this day. In 2000, it was also clearly seen that the
15-year effects of 5 years of tamoxifen were sustained and
of great magnitude. The ATLAS, ATtOM, and a few other
small trials were combined and opened the door to the
suggestion that 5 years of tamoxifen might not be optimal
and that longer tamoxifen might further reduce disease-free
From the Odette Cancer Center, McMaster University, Hamilton, ON, Karolinska
Institutet, Stockholm, Sweden, and the Dana-Farber Cancer Institute, Boston, MA.
Address reprint requests to Harold J. Burstein, Dana-Farber Cancer Institute, 450
Brookline Avenue, Boston, MA 02215; email: hburstein@partners.org.
1092-9118/10/1-10
71
PRITCHARD, BERGH, AND BURSTEIN
Fig. 1. Effects of approximately 5 years

of tamoxifen on the 15-year probabilities
of recurrence and of breast cancer mortality for ER-positive disease.
Abbreviations; ER, estrogen receptor;
RR, recurrence rate; SE, standard error;
(O-E)/V, (observed-expected)/variance.
Reprinted from The Lancet, 378, Early
Breast Cancer Trialists Collaborative
Group, Davies C, Godwin J, et al. Relevance
of breast cancer hormone receptors and
other factors to the efficacy of adjuvant
tamoxifen: Patient-level meta-analysis of
randomised trials, 771784, 2011, with
permission from Elsevier.
recurrence. This remains an important and unresolved

question. It was also clear that ovarian ablation and/or
suppression was effective but not significantly so when
added to chemotherapy, perhaps because of chemotherapyinduced amenorrhea.
In 2005, structural changes were required in the Overview
process. The Trialists formed a new steering committee and
organized subcommittees, many of which continue today to
work very effectively. Many new trials were added and there
were many additional years of follow-up for all major questions. However, data from many major trials, particularly
those of taxanes, were missing in 2005.
In 2006, the Trialists and their subcommittees met and
a series of priorities were set. These included investigations
of the type of anthracycline-based regimen; all taxane trials;
aromatase inhibitors; trastuzumab; and chemoendocrine
therapy, particularly in relation to the ER-positive compared with ER-negative issue in pre- and postmenopausal
women. These meetings led five important publications on
tamoxifen, chemotherapy, and loco-regional therapy.1-5
Nearly 30 years since its inception, the Overview remains
highly relevant and informative. In comparison to individual
trials even huge trials tallying thousands of patients as
has become common in early-stage breast cancerthe Overview has several methodologic virtues that make it a unique
data resource. In particular, the Overview is important for
(1) having all the worldwide data; (2) avoiding publication
KEY POINTS
72
The Oxford Overview suggests benefits for adjuvant

endocrine therapy for all patients with ER positive
breast cancer.
The Oxford Overview suggests benefits for anthracycline- and taxane-based adjuvant chemotherapy
regardless of nodal, ER, or PR status.
Molecular diagnostic assays may identify ER-positive
tumors that may not warrant chemotherapy.
Reconciling the historic overview and newer personalized approaches to early breast cancer is a
compelling clinical challenge.
bias; (3) giving average effect sizes; and (4) clarifying the
timeframes of effects through large sample size and longterm follow-up.
Recent Notable Findings from the Overview
Endocrine Therapy
In the main tamoxifen Overview, more than 54,500

women were studied in trials of tamoxifen compared with
no tamoxifen and more than 45,000 in trials of longer
compared with shorter tamoxifen. The effects of 5 years of
tamoxifen on breast cancer recurrence and mortality are
shown in Fig. 1.
Fundamentally, tamoxifen has long and strong sustained
effects on local recurrence, contralateral breast cancer, and
distant and multiple recurrences. Tamoxifen benefits as
measured by proportional risk reduction are similar regardless of age, stage, grade, or tumor size and with and without
background chemotherapy.
Figure 2 examines the effect of ER and progesterone
receptor (PgR) expression on the benefits with tamoxifen.
Tamoxifen is similarly effective in patients with ER-positive
disease regardless of PgR expression. Patients with ERnegative but PgR-positive tumors do not get benefit from
tamoxifen, nor do patients with ER- and PgR-negative
tumors.
Figure 3 explores the relationship of quantitative levels of
ER with the benefits of tamoxifen. It demonstrates that
higher ER levels are associated with stronger effects of
tamoxifen.
Figure 4 shows the effects of tamoxifen over time. They
are large in years 0 to 1, 2 to 4 and 5 to 10 in terms of
recurrence, but by year 10 and beyond, the effects no longer
increase although the previous gains are not lost. In mortality however, the benefits come out a little later in years 2 to
4 and then years 5 to 9 and they persist into the 10 to
15year follow-up. Thus, there is a carryover effect on
recurrence of 5 years of tamoxifen that goes on to at least 10
years and in mortality that goes on to at least 15 years.
The Overview has been pivotal in demonstrating the
benefits of endocrine therapy for ER-positive breast cancer.
Five years of tamoxifen in ER-positive disease reduces
recurrences by a relative risk of 38%, breast cancer deaths
by approximately 30%, and all deaths by approximately
UPDATE OF THE OXFORD OVERVIEW
Fig. 2. Relevance of measured ER and

PR status to the effects of tamoxifen on the
10-year probably of recurrence.
Abbreviations: ER, estrogen receptor; PR,
progesterone receptor; RR, recurrence rate;
SE, standard error; (O-E)/V, (observedexpected)/variance.
Group, Davies C, Godwin J, et al. Relevance
of breast cancer hormone receptors and
other factors to the efficacy of adjuvant
tamoxifen: Patient-level meta-analysis of
randomised trials, 771784, 2011, with
permission from Elsevier.
22%. Contralateral breast cancer is reduced by approximately 40%. Tamoxifen for 5 years benefits all women with
ER-positive disease and benefits those with very high levels
of ER even more. To date, ER expression is the sole deter-
minant of likely benefit from tamoxifen. Endometrial cancer,

however, is increased by 2.3-fold following 5 years of tamoxifen. The rate of endometrial cancer increases as much as
4-fold with 10 years of tamoxifen.
Fig. 3. Relationship of quantitative

levels of ER with benefits of tamoxifen.
Abbreviations: O-E, observed-expected; ER, estrogen receptor; SE, standard
error.
Group, Davies C, Godwin J, et al. Relevance of breast cancer hormone receptors and other factors to the efficacy of
adjuvant tamoxifen: Patient-level metaanalysis of randomised trials, 771784,
2011, with permission from Elsevier.
73
Fig. 4. Abbreviations: ER, estrogen receptor; y, year; SE, standard error; (O-E)/V, (observed-expected)/variance.
Reprinted from The Lancet, 378, Early Breast Cancer Trialists Collaborative Group, Davies C, Godwin J, et al. Relevance of breast cancer
hormone receptors and other factors to the efficacy of adjuvant tamoxifen: Patient-level meta-analysis of randomised trials, 771784, 2011,
with permission from Elsevier.
Radiation Therapy
Chemotherapy
The Overview has also enabled powerful analyses of the

effects of radiotherapy for early-stage breast cancer. The
proportional effects of radiotherapy after breast-conserving
surgery are substantial for both node-negative and nodepositive disease as shown in Fig. 5. There are substantial
reductions in any recurrence and improvements in breast
cancer survival with postsurgical radiotherapy after breastconserving surgery.
An important question regarding radiotherapy for breast
cancer has been whether all-cause mortality would be enhanced by radiation treatments or whether late (rare) side
effects of radiation might negate treatment benefits in
breast cancerspecific mortality. The Overview, with its
large number of trials and patients, is uniquely positioned to
address this question. Recent data (Fig. 6) indicate that
radiotherapy after breast-conserving surgery reduces breast
cancer recurrence and all-cause mortality among both nodepositive and node-negative disease. A one in four rule
applies for patients with pN0 and pN1 disease in that one
death is prevented for every four recurrences that are
prevented. These benefits are not substantially reduced by
side effects and are durable through 15 years of follow-up.
The recent overview analysis was published in The Lancet

in December 2011 and was based on the analysis of 100,000
randomly selected patients treated in different randomized
studies.5
In short, the overview contained the quinquennial update
of the following:
Nontaxane chemotherapy compared with taxanes in
44,000 patients
Different anthracycline based regimens in 6,000 patients
Anthracyclines in the comparisons compared with CMF
in 18,000 patients
Randomized studies in 32,000 patients comparing no
chemotherapy with polychemotherapy
For this EBCTCG round, the chemotherapy regimens
were, for the first time, grouped based on scheduling and
dose intensity, with particular focus on the CMF- and
anthracycline-based regimens. All outcome analyses were
done based on individual patient tumor data with as long of
follow-up as possible for each study. Individual patient data
were also included from some unpublished randomized stud-
74
Fig. 5. Effect of radiotherapy after

breast-conserving surgery on 10-year risk
of any (loco-regional or distant) first recurrence on 15-year risk of breast cancer
death in women with pathologically verified node status.
Abbreviations: SE, standard error; RR, recurrence rate; BCS, breast-conserving surgery; RT, radiotherapy.
Group, Darby S, McGale P, et al. Effect of
radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year
breast cancer death: Meta-analysis of individual patient data for 10,801 women in
17 randomized trials, 17071716, 2011,
with permission from Elsevier.
ies, thereby reducing the potential risks of publication bias

of studies with a positive outcome, the latter being published earlier. Data were analyzed for recurrence, breast
cancerspecific survival, and overall mortality.
A summary of major findings includes the following results:

The use of taxanes added a statistically significant
reduced risk of breast cancer death by 14% (2.7% absolute
Fig. 6. Effect of radiotherapy after breast-conserving surgery on 10-year risk of any (loco-regional or distant) first recurrence on 15-year
risks of breast cancer death and death from any cause in 10,801 women (67% with pathologically node-negative disease) in 17 trials.3
Abbreviations: SE, standard error; RR, recurrence rate; BCS, breast-conserving surgery; RT, radiotherapy.
Reprinted from The Lancet, 378, Early Breast Cancer Trialists Collaborative Group, Darby S, McGale P, et al. Effect of radiotherapy after
breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: Meta-analysis of individual patient data for 10,801 women
in 17 randomized trials, 17071716, 2011, with permission from Elsevier.
75
Fig. 7. Studies comparing the add-on effect by taxanes, most AC x 4. Demonstration

of the outcome (breast cancer survival and
overall survival) when the comparator arm
contained higher doses of the nontaxane
drugs, mostly anthracyclines.
Abbreviations: AC, doxorubicin/cyclophosphamide; RR, recurrence rate; SE, standard error; (O-E)/V, (observed-expected)/
variance.
Breast Cancer Trialists Collaborative Group,
Peto R, Davies C, et al. Comparisons between different polychemotherapy regimens for early breast cancer: Meta-analyses
of long-term outcome among 100,000
women in 123 randomized trials, 432 444,
2012, with permission from Elsevier.
gain, p 0.0005, 2.2% absolute overall survival gain)

compared with anthracycline-based regimens at 8-years
follow-up. This effect was not statistically significant with
only a 6% relative effect if the anthracycline control arm was
given with a high cumulative anthracycline dose (Fig. 6).
There was no difference in schedule or taxane drug
(paclitaxel or docetaxel).
There was no difference whether endocrine therapy was
given concurrently or in sequence with chemotherapy.
There were no age-related effects by the delivered
chemotherapies, taxane, or anthracycline-based regimens.
Indeed, there was a trend suggesting greater chemotherapy
effect in the age 55 to 69 group than in younger women.
The anthracycline and taxane regimens had similar
antitumoral effects irrespective of patient age, tumor ER
status, whether tamoxifen was given, the combination of ER
by HER2, or the interaction of ER and grade. (Fig. 7).
76
Four cycles of doxorubicin and cyclophosphamide produced a similar result as six courses of standard CMF at 10
years follow-up. Anthracycline regimens with higher doses,
however, reduced the breast cancer mortality by about 4%
at 10 years (a relative improvement of 22%, p 0.004)
compared with CMF. The overall mortality was improved by
the same extent.
In comparison with no chemotherapy, CMF-like regimens and anthracycline-based regimens reduced overall
mortality by 6% at 10 years. No subgroup could be readily
identified that did not benefit from chemotherapy. The
CMF-based studies revealed a less distinct message; the
effect seemed less obvious for the elderly (irrespective of
ER status). This could be because of a combination of factors:
lower compliance in those studies run decades ago; no access
to modern antiemetic regimens; a lack of supportive care
measurement; and less patient motivation to accept toxicity
Fig. 8. Breast cancer mortality for anthracyclines compared with no adjuvant chemotherapy. Outcomes for different cumulative anthracyclines doses, age groups, node status, ER status. and the combination of ER status and histopathologic grade. All forest plots reveal a very similar
pattern of the same relative magnitude of chemotherapy.
Abbreviations: ER, estrogen receptor; CAF, cyclophosphamide/doxorubicin/fluorouracil; SE, standard error; 4AC/EC, four cycles of doxorubicin/cyclophosphamide or epirubicin/cyclophosphamide; NS, nodal status.
Reprinted from The Lancet, 379, Early Breast Cancer Trialists Collaborative Group, Peto R, Davies C, et al. Comparisons between different
polychemotherapy regimens for early breast cancer: Meta-analyses of long-term outcome among 100,000 women in 123 randomized trials,
432 444, 2012, with permission from Elsevier.
77
since an eventual benefit by adjuvant therapies was not

known at the time these studies were done.
The Era of Personalized Breast Cancer Treatment
During the last decade, there have been major attempts

to tailor adjuvant chemotherapy and, to a lesser degree,
the choices for adjuvant endocrine therapy, based on consideration of the unique biologic features of breast cancers. This
effort has grown from recognition of major breast cancer
subsets defined by widely used histopathologic markers
(ER, PgR, and human epidermal growth factor receptor 2
[HER2]) as well as traditional pathology (grade, proliferation)
and molecular diagnostics (intrinsic subsets/OncotypeDX
characteristics).6 These schemes of prognostication and
therapy prediction separate breast cancer into distinct subgroups, with different management, based principally on
tumor histopathology, abetted by detailed insights from
gene expression profiling of tumors.
A number of retrospective analyses from prospective,
randomized trials have suggested that the proportional
benefits of chemotherapy are not, in fact, the same for all
kinds of breast cancer.6-9 In addition, previous studies have
also revealed that chemotherapyin particular CMF-based
therapiesmay have no or little value to patients with high
ERs levels and that the major antitumoral effect for the
premenopausal patients may be because of induction of
amenorrhea.10 These subset analyses have suggested that
the clinical benefit of chemotherapy is most dramatic in
tumors with low or no ER expression (including triplenegative cancers), overexpression of HER2, higher grade,
and higher proliferative scores. Conversely, endocrine therapies are most effective in tumors with higher levels of ER
expression, lower levels of HER2, lower grade, and lower
proliferative scores. Retrospective subset analyses of specific
phase III studies have frequently found that the benefit of
adding taxane-based therapy is least pronounced among
cancers that are ER positive and HER2 negative (the majority of most breast cancers), although these data must now
be seen in the global context of the Overview findings to the
contrary. However, consistent with the findings in single
adjuvant trials, neoadjuvant studies have suggested that
complete pathologic response to chemotherapy is far less
common in tumors that are ER positive. Molecular assays
have emerged that appear to predict the likelihood of chemotherapy benefit, particularly in addition to adjuvant
endocrine therapy. Tumors classified as having a luminal A
intrinsic subtype or a low recurrence score on the OncotypeDX assay, tend to derive minimal benefit from the
addition of chemotherapy to endocrine therapy. It is unclear
whether this is because the relative gain by chemotherapy is
the same, but with an absolute gain that is too small to affect
outcomes among patients with low-risk disease, or whether
there truly is no benefit with chemotherapy. These assays
correlate well if imperfectly with other tumor features such
as grade, proliferation, and quantitative levels of ER and
HER2, fitting with subset hypotheses about the importance
of those pathologic features.
Finally, the advent of anti-HER2 therapy in the adjuvant
setting with trastuzumab has radically altered the natural
history of HER2-positive breast cancers. These tumors,
which traditionally had a less favorable prognosis, now
appear to have as good a prognosis as HER2-negative
tumors. The use of trastuzumab for HER2-positive cancers
78
and the additional use of adjuvant endocrine therapy for

ER-positive cancers has meant that issues of endocrine
therapy or not or chemotherapy or not are less compelling
now than decades ago. The persistent question has become:
are there sufficient data to tailor treatments based on
specific biomarker subsets, or do all patients need the same
treatments?
Can We Reconcile the Overview and the Individual
Treatment Paradigms?
We now confront the paradox of the Overview. In large

measure, the Overview argues for adjuvant chemotherapy
for all women, regardless of tumor biology or stage. This
blanket observation seems at odds with the growing literature that suggests that individual tumorsand thus individual patientsrespond differently to chemotherapy and
endocrine therapy. Can these observations be reconciled?
The reason why the EBCTCG analysis demonstrates relatively similar antitumoral effects in all the different subgroups is so far not understood. There are two particular
strengths of the Overview that should not be forgotten.
First, the Overview is an enormous data repository that
dwarfs in the number of events and the duration of follow-up
of the available data from any given study. This gives the
Overview power to see effects that might be missed in
smaller, retrospective efforts. Of note, the previously published reports on selectivity of the chemotherapy effects are
the result of subgroup analysis on materials with much less
statistical power and in some studies lacking inclusion of all
randomly selected patients and overlapping 95% confidence
intervals for some of the interesting findings.
Secondly, by including data from almost all adjuvant
trials, the Overview minimizes the temptation to focus on
positive, published studies, a bias particularly notable in the
literature on biomarker subsets where negative studies
are all but unpublishable. These are important considerations that should give clinicians pause before dismissing
the Overview findings.
At the same time, there are features of the Overview
design that may overstate the benefits of chemotherapy. The
Overview reports on proportional risk reduction and related
absolute differences in outcome. However, for most patients,
the absolute gains are the driving force for decisions on
chemotherapy. Differences in absolute benefit from chemotherapy clearly vary based on tumor stage and on the
residual degree of risk that remains despite adjuvant endocrine therapy. In addition, the Overview has not been able to
capture adequate information on chemotherapy-induced
amenorrhea so as to tease out the effects of chemotherapy on
ovarian function in women with ER-positive tumors, though
chemotherapy benefit is notably similar regardless of age
and ER status.5
More critically, perhaps, is that differences in treatment
effect may hinge on consideration of multiple variables that
the Overview has not as yet grappled with in detail. It could
simply be because the EBCTCG data just describes mean
effects for all the different subgroups without enabling the
identification of distinct subgroups with claimed different
biology and outcome. The characterization of subsets of
ER-positive cancers, in particular, by molecular assays suggests that simultaneous consideration of ER levels, HER2
expression, and grade/proliferation is important for classifying cancers and determining treatment benefit. As yet, the
Overview analyses looking at quantitative hormone receptor

levels or considering 2 2 analyses of ER by HER2 or ER
by grade may not be sufficiently multiplexed to figure out
which patients can avoid chemotherapy. A major limitation
with the present Overview is the lack of data on proliferation, gene expression, modern immunohistochemical measurements of receptors, quality controlled pathology
classification, and central marker review. Information on
these variables would, of course, have added value to the
EBCTCG processes and analyses, and some investigators
may claim that these shortcomings may explain some of the
results. Despite these shortcomings, the example of ER
values in the overall EBCTCG suggest very robust results in
relation to adjuvant tamoxifen, further supporting the validity of the EBCTCG findings.4
Additional methodologic details may be relevant. Several
studies that have identified selective benefit for chemotherapy have centrally reanalyzed ER or HER2 data on some
patients, a detailed pathologic step not preformed in the
EBCTCG database.8,10 Similarly, the Overview used a surrogate strategy for high proliferation/higher OncotypeDx
scores by using grade as reported by local laboratories (in
particular, poorly differentiated cancers) that have not been
confirmed. Furthermore, the EBCTCG demonstration of
similar effects of chemotherapies, independent of studied
subgroups, could merely be a reflection of an inherent
common biology of breast cancer with a similar type of
sensitivity to the commonly used cytostatics. However, this

seems counterintuitive based on present extensive knowledge of breast cancer biology in relation to outcome and
therapy strategies, but it has happened before in the era of
science that conclusions made with the best intentions have
required modification.
A possible explanation to the general findings of a similar
relative magnitude in anti breast cancer effects in the
adjuvant setting by taxanes and anthracyclines could be
that all epithelial breast cancers seem to share so far
unidentified gene cassettes associated with a similar sensitivity to chemotherapy. This could, to some extent, be
potentially substantiated by ongoing studies on human
breast cancer stem cells from primary cultures, which indicate a common phenotype for most breast cancer stem cells
in relation to a specific receptor status, despite the fact that
cancers per se have different tumor and receptor characteristics (Johan Hartman, Irma Fredriksson, Jonas Bergh,
verbal communications, March 1, 2012).
At its heart, the task of reconciling the invaluable data
from the ongoing Oxford Overview with the emerging data
from subset studies using novel markers remains the fundamental challenge for adjuvant therapy for breast cancer.
Ongoing collaboration and research will be critical for helping clinicians and patients understand these different but
important clinical datasets and solving the riddle of the
Overview paradox.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Kathleen I. Pritchard
GlaxoSmithKline;
Novartis; Ortho
Biotech; Pfizer;
Roche; Sanofi;
YM BioSciences
AstraZeneca;
GlaxoSmithKline;
Novartis; Pfizer;
Roche; Sanofi
Jonas Bergh
Affibody; Amgen;
AstraZeneca;
Bayer;
GlaxoSmithKline;
i3innovus; Onyx;
Pfizer; Sanofi;
Tapestry
Pharmaceuticals
AstraZeneca;
Pfizer; Roche;
Sanofi
Research
Funding
Expert
Testimony
Other
Remuneration
AstraZeneca;
Novartis; Pfizer;
Sanofi
Merck; Pfizer;
Roche
AstraZeneca;
Roche
Harold J. Burstein*
REFERENCES
1. Early Breast Cancer Trialists Collaborative Group. Adjuvant polychemotherapy in estrogen-receptor-poor breast cancer: Meta-analysis of individual patient data from randomized trials. Lancet. 2008;371:29-40.
2. Early Breast Cancer Trialists Collaborative Group. Overview of the
randomized trials of radiotherapy in ductal carcinoma in situ (DCIS) of the
breast. J Natl Cancer Inst Monogr. 2010;41:162-177.
3. Early Breast Cancer Trialists Collaborative Group. Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast
cancer death: Meta-analysis of individual patient data for 10,801 women in 17
randomized trials. Lancet. 2011;378:1707-1716.
4. Early Breast Cancer Trialists Collaborative Group. Relevance of breast
cancer hormone receptors and other factors to the efficacy of adjuvant
tamoxifen: Patient-level meta-analysis of randomised trials. Lancet. 2011;
378:771-784.
5. Early Breast Cancer Trialists Collaborative Group. Comparisons between different polychemotherapy regimens for early breast cancer: Metaanalyses of long-term outcome among 100,000 women in 123 randomized
trials. Lancet. 2012;379:432-444.
6. Goldhirsch A, Wood WC, Coates AS, et al. Strategies for subtypesdealing with the diversity of breast cancer: Highlights of the St. Gallen
International Expert Consensus on the Primary Therapy of Early Breast
Cancer 2011. Ann Oncol. 2011;22:1736-1747.
7. Paik S, Tang G, Shak S, et al. Gene expression and benefit of
chemotherapy in women with node-negative, estrogen receptor-positive
breast cancer. J Clin Oncol. 2006;24:3726-3734.
8. Hayes DF, Thor AD, Dressler LG, et al. HER2 and response to paclitaxel
in node-positive breast cancer. N Engl J Med. 2007;357:1496-1506.
9. Berry DA, Cirrincione C, Henderson IC, et al. Estrogen-receptor status
and outcomes of modern chemotherapy for patients with node-positive breast
cancer. JAMA. 2006;295:1658-1667.
10. Karlsson P, Sun Z, Braun D, et al. Long-term results of International
Breast Cancer Study Group Trial VIII: Adjuvant chemotherapy plus goserelin
compared with either therapy alone for premenopausal patients with nodenegative breast cancer. Ann Oncol. 2011;22:2216-2226.
79
GENE PATENTING: EFFECTS ON BIOTECHNOLOGY

AND ONCOLOGY
CHAIR
Kenneth Offitt, MD, MPH
New York, NY
SPEAKERS
Roger D. Klein, MD, JD
Blood Center of Wisconsin
Milwaukee, WI
Hans Sauer, PhD, JD
Biotechnology Industry Organization
Houston, TX

By Roger D. Klein, MD, JD
Overview: The use of genetic information to design and guide

therapies and to develop novel diagnostic procedures creates
important patent issues. Patents on human gene sequences
have likely helped stimulate the introduction of new biologics;
however, their role and that of patents on genotype-phenotype
HE PRACTICE of personalized medicine has been

described as giving the right drug, to the right patient,
at the right dose, at the right time. Central to the concept of
personalized patient management is knowledge and application of somatic and heritable molecular pathologic (molecular genetic) changes associated with disease susceptibility,
prognosis, therapeutic responsiveness, and susceptibility to
side effects. Although we are far from achieving truly individually designed care for most patients, in oncology the era
of targeted therapies linked to measurable biomarkers has
arrived. Diagnostic tests have long been important in cancer
management, but the high sensitivity, specificity, and direct
relationship to therapy of many nucleic acid-based analyses
adds enormously to their utility.
Because they have been rationally designed to directly
influence biochemical pathways implicated in malignant
progression, targeted cancer therapies potentially offer patients greater benefits with less morbidity than most currently used chemotherapeutic agents. Moreover, by allowing
for prospective selection of individuals most likely to benefit
from a given therapy, molecular pathology-guided approaches benefit patients by rescuing antineoplastic agents,
the efficacy of which is confined to select populations. Without the ability to specifically identify potential responders,
such drugs may mistakenly be abandoned. Finally, personalized medicine encourages cost-effective use of expensive
therapies by restricting administration to patients most
likely to benefit and least likely to suffer harms from them.
The use of molecular pathologic information to develop

and prescribe therapies raises important new patent issues.
In the United States, patents confer on the patent holder
the right to exclude others from making, using, or selling an
invention for 20 years from the filing date. Under the Patent
Act, patentable inventions must be novel, nonobvious, and
useful.1 Patent claims define an inventions features much
as a deed delineates the boundaries of a plot of land. Patent
applications are submitted to the U.S. Patent and Trademark Office (USPTO) where they are rejected or allowed and
issued. Importantly, [p]rocesses, machines, manufactures,
and compositions of matter can be patented,2 but patents
may not be obtained on products of nature or, under the
natural phenomenon doctrine, laws of nature, natural
phenomena, and abstract ideas.3
Patents have been integral to the discovery and commercialization of drugs and biologics. The product exclusivity
patents confer has been instrumental in ensuring that
manufacturers can raise sufficient capital to identify, test,
obtain regulatory approval for, and bring to market new
therapeutic agents. Absent enforceable patent rights for
new drugs, investment dollars would likely be redeployed to
correlations in diagnostic testing is highly controversial.

Genotype-phenotype associations are at the heart of personalized medicine. The intellectual property rules by which these
biologic relationships are governed have profound implications for the growth of individualized medicine.
areas with a more favorable risk-reward balance. So clear

are the positive incentives patents provide the pharmaceutical industry, even patent system critics have acknowledged them.4 Similarly, the stimulatory effects of patent
protection on pharmaceutical development may extend to
patents granted on human genes, when patented genes are
cloned to produce biologic therapies and vaccines.5
By contrast, patents on relationships between genetic
variants and clinical phenotypes are controversial. Such
patents have been widely criticized and their legitimacy
challenged, whether the patents directly claim genotypephenotype associations or gene sequences themselves.5,6
Because genotype-phenotype associations are the center of
personalized medicine, the intellectual property rules by
which these biologic relationships are governed have profound implications for its growth.
Clinical implementation of molecular pathology procedures is substantially less costly than the discovery, development, and commercialization of in vivo pharmaceuticals.
Manufacturers typically invest hundreds of millions of dollars over a prolonged period of time to bring a drug or
biologic to market. Drug candidates must be extensively
tested in preclinical and clinical studies before U.S. Food
and Drug Administration (FDA) approval. Unlike therapeutics, most molecular pathology procedures are designed,
developed, and validated by individual diagnostic laboratories at limited expense. The wide availability of automated
nucleic acid extractors, thermal cyclers, and DNA sequencing instruments combined with broad licensing of standard
molecular methods and the free accessibility of online gene
sequence databases has rendered development straightforward for many molecular pathology procedures.
Laboratory quality is ensured by the Clinical Laboratory
Improvement Amendments, accreditation by the College of
American Pathologists and other professional societies, and
individual state requirements. Most molecular pathology
services do not undergo FDA clearance or approval, although laboratories serving New York patients submit tests
to the states Clinical Laboratory Evaluation Program for
pre-implementation review. The small minority of molecular
pathology procedures manufactured and sold as kits to
clinical laboratories must be cleared or approved by the
FDA. However, this process is much less burdensome and
orders of magnitude less expensive for in vitro diagnostic
kits than for drugs.
From the Medical College of Wisconsin, Milwaukee, WI.

Address reprint requests to Roger D. Klein, MD, JD, 27500 Cedar Road #808, Beachwood,
OH 44122; email: roger.klein@aya.yale.edu.
1092-9118/10/1-10
81
ROGER D. KLEIN
The majority of gene-related discoveries have been made

by federally funded academic researchers, for whom patents
are unlikely to have provided primary or even significant
motivation.7 Further, given the relative ease by which
molecular pathology procedures are introduced into practice, patents are unnecessary for the clinical implementation
of molecular pathologic discoveries. Rather, enforcement of
gene patents in the diagnostic realm results in the elimination of already existing services and dissuades laboratories
from adding new ones.8 The resultant elimination of competition is likely to result in higher prices, decreased quality,
reduced innovation, and diminished patient access to these
important medical services. Moreover, loss of academic
providers will adversely affect training and related clinical
research.
Greater interest in companion diagnostic development by
the pharmaceutical industry may alter the preceding dynamics. Control of companion diagnostics could allow drug
manufacturers to ensure assay standardization, enforce
specific quality requirements, and provide additional
sources of revenue. Companion diagnostics could serve as
vehicles by which companies preserve market share for
linked drugs, assuming they avoid patent misuse or antitrust violations. Under this model, gene-related patents may
create incentives for marker discovery and in vitro diagnostic commercialization. However, such patents could prove a
double-edged sword by creating holdout problems for drug
vendors who do not own the underlying molecular pathologic
relationships.
Thousands of U.S. patents have been issued on human
gene sequences and genotype-phenotype associations, but
their legal status remains uncertain. Several cases presently
making their way through the courts may clarify and add
reason to the law in this area.
Recent Legal Developments
In Mayo Collaborative Servs. v. Prometheus Labs., Inc.,9

Prometheus sued Mayo Clinic in the District Court for the
KEY POINTS
82
Personalized medicine in the form of targeted drug

therapies is already making important contributions
to oncology care.
Patents on human gene sequences have likely helped
stimulate the introduction of novel biologics and
pharmaceutical agents.
Conversely, patents on human gene sequences and
genotype-phenotype correlations appear to reduce
the availability of and patient access to already
existing diagnostic services, while increasing costs
and decreasing innovation in the development of
diagnostic methods.
The intellectual property rules by which patents on
human genes and genotype-phenotype associations
are governed will have a profound impact on the
advancement of personalized cancer care.
Several key legal cases now before the courts may add
clarity and guidance to the law governing this area.
Southern District of California for infringement of a process

patent covering the postadministration correlation between
thiopurine drug activity and side effects, and blood levels of
the metabolites 6-methyl mercaptopurine (6-MMP) and
6-thioguanine (6-TG). The patent claims at issue include
administering a drug, determining the level of a metabolite of the drug, and correlating this level with therapeutic
efficacy or side effects.
The district court found as a matter of law that the patent,
which essentially claims the reference range for the drugs,
covered an unpatentable natural phenomenon. The Court of
Appeals for the Federal Circuit (CAFC), the national patent
appeals court, reversed the lower court, finding that the
patent covers a treatment method. Further, the CAFC held
that the in vivo metabolism of thiopurine agents constituted
a transformation of matter, consistent with a patentable
process on an application of a natural phenomenon as
opposed to a patent on the phenomenon itself.
The Prometheus Labs case was appealed to and accepted
by the U.S. Supreme Court. Oral arguments were held
before the Court on December 7, 2011. Of note, during oral
arguments, the attorney for Prometheus Labs acknowledged
that it had patented a fact they identified, and that
physicians can infringe the patent merely by thinking about
the biologic relationship between metabolite levels and patient response. However, because the case involved an exogenously administered drug rather than a genetic variant
intrinsic to a patient, and the CAFC considered the patent
a therapeutic method patent, and Prometheus defended
its validity on this basis, the implications for genotypephenotype association patents are unclear.
In a case addressing legal standards for patent obviousness, KSR Intl Co. v. Teleflex Inc., 550 U.S. 398 (2007),
Teleflex sued KSR for infringement of a patent that claimed
the combination of an adjustable brake pedal and an electronic sensor. The district court found the patent obvious,
but the CAFC reversed the lower court decision. The Supreme Court in turn reversed the CAFC, ruling that the
patent was obvious. The CAFC, the high court stated, had
applied overly restrictive criteria in finding the patent
nonobvious. Importantly, the Supreme Court held that
obviousness to try a problem-solving approach can render
a patent obvious if there is a demonstrated need for a
discovery and a finite number of identified, predictable
solutions.
Many if not most patented genes were initially mapped
to a chromosomal region before discovery. Moreover, many
genes are involved in sequential biochemical pathways in
which disease-related changes were known before specific
genetic variations were identified. Therefore, for many of
these discoveries, it was arguably obvious to look for variants in the potentially responsible genes among the finite
number of available solutions. As a result, under KSR Intl
Co. v. Teleflex Inc. (KSR) many gene-related patents may be
subject to challenge on obviousness grounds. In an early
application of the KSR ruling, the USPTO refused to award
a patent on the gene sequence of the natural killer cell
activationinducing ligand. Although the sequence had not
been previously described, the USPTO found it obvious in
light of the prior art. The CAFC upheld the USPTOs
decision.10
Finally, in Association for Molecular Pathology v. United
States Patent and Trademark Office, a lawsuit sponsored
GENE PATENTS AND EFFECTS ON PERSONALIZED CANCER CARE
by the American Civil Liberties Union, a number of medical

and professional societies, health care providers, and patients with breast cancer sued Myriad Genetics and the
USPTO. The plaintiffs seek to invalidate key claims of
patents covering wild-type and mutated sequences of the
BRCA1 and BRCA2 genes and associations between those
sequences and the predisposition to breast and/or ovarian
cancer. In addition to arguing that the patents claim unpatentable natural products and natural phenomena, the plaintiffs asserted that they violate Article I, section 8, clause 8
and the First Amendment to the U.S. Constitution.
On March 29, 2010, in a landmark decision, the court held
that both composition of matter claims on the human
BRCA1 and BRCA2 gene sequences, and process claims
covering the correlations between mutations in these genes
and a predisposition to breast cancer are invalid as a matter
of law. These patent claims, the court ruled, were directed
toward unpatentable subject matter.10
On July 29, 2011, although upholding the district courts

finding of invalidity of the genotype-phenotype correlation
claims, the CAFC held that human gene sequences are
patent-eligible subject matter, reversing ruling of the lower
court. The plaintiffs have appealed the case to the Supreme
Court.
Conclusion
Genes as chemicals have an important role to play in

therapeutic development. However, ready access to the
information contained within patients genes is also critical
to the advancement of personalized medicine. Therefore, a
prudent course for the courts when addressing gene-related
patents is to strike a balance, upholding their validity when
the patents protect drug development, while denying their
enforceability when the patents create the potential for
genetic testing monopolies.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Roger D. Klein*
REFERENCES
1. 35 U.S.C. 101-103 (2006).
2. 35 U.S.C. 101 (2006).
3. Diamond v. Diehr. 450 U.S. 175, 185 (1981).
4. Bessen J, Meurer MJ. Patent Failure: How Judges, Bureaucrats, and
Lawyers Put Innovators at Risk. Princeton, NJ: Princeton University Press; 2008.
5. Klein RD. Gene patents and genetic testing in the United States. Nat
Biotechnol. 2007;25:989-990.
6. Secretarys Advisory Committee on Genetics, Health, and Society. Revised Draft Report on Gene Patents and Licensing Practices and Their Impact
on Patient Access to Genetic Tests (2010). http://oba.od.nih.gov/SACGHS/
sacghs_documents.html. Accessed February 10, 2010.
7. Klein RD, Mahoney MJ. LabCorp v Metabolite Laboratories: the Supreme Court listens, but declines to speak. J Law Med Ethics. 2008;36:141149.
8. Cho MK, Illangasekare S, Weaver MA, et al. Effects of patents and
licenses on the provision of clinical genetic testing services. J Mol. Diagn.
2003;5:3-8.
9. In v. Mayo Collaborative Servs. Prometheus Labs., Inc. 581 F. 3d 1136
(Fed. Circ. 2009), cert. granted (No. 10-1150).
10. In re Kubin, 561 F. 3d 1351 (Fed. Cir. 2009).
11. Assn for Molecular Pathology v. U.S. Patent and Trademark Office. 653
F. 3d 1329 (Fed. Cir. 2011), petition for cert. filed.
83
BREAST CANCER CHEMOPREVENTION:

IF NOT NOW, WHEN?
CHAIR
Houston, TX
SPEAKERS
Nancy E. Davidson, MD
University of Pittsburgh Cancer Center
Pittsburgh, PA
Worta McCaskill-Stevens, MD, MS
Bethesda, MD
Chemoprevention for Breast Cancer:

Overcoming Barriers to Treatment
By Abenaa M. Brewster, MD, MHS, Nancy E. Davidson, MD, and
Worta McCaskill-Stevens, MD, MS
Overview: Evidence from placebo-controlled, randomized

clinical trials supports the use of chemoprevention in women
at high risk for developing breast cancer, and two agents
tamoxifen and raloxifeneare U.S. Food and Drug Administration (FDA)-approved for the indication. Despite clinical
guidelines that recommend physicians counsel high-risk
women about the use of chemoprevention and the estimated
2.4 million women in the United States who meet eligibility
criteria for net benefit, the uptake of breast cancer chemoprevention has been exceedingly low. Assessments of the risks
and benefits of chemoprevention are aided by the availability
of models that can be used to estimate of the risk benefit
ratio. However, many physicians remain unaware of these
EVERAL RANDOMIZED clinical trials have provided

evidence for the chemoprevention of invasive breast
cancer by selective estrogen receptor modulators (SERMs).
The Breast Cancer Prevention Trial (BCPT) randomly selected 13,388 women age 35 or older with a 5-year predicted
risk for breast cancer of 1.66% or higher to receive either
tamoxifen or placebo.1 At a mean follow-up of 4 years, there
was a 49% reduction in the risk of developing invasive breast
carcinoma and a 50% reduction in the risk of noninvasive
breast cancer among women who were assigned to the
tamoxifen arm. The absolute risk reduction was 21.4 cases
per 1,000 women over 5 years. The benefit of tamoxifen was
limited to estrogen receptorpositive tumors, and the greatest risk reduction (86%) was observed in women with a
history of atypical hyperplasia. However, tamoxifen was also
associated with a 2.5-fold increased risk of endometrial
carcinoma and an increased risk of stroke, deep venous
thrombosis, pulmonary embolus, and cataracts.1 The results
of the Study of Tamoxifen and Raloxifene (STAR) demonstrated that raloxifene was as effective as tamoxifen in
reducing the risk of breast cancer in postmenopausal women
at increased risk of breast cancer with less adverse events of
thromboembolic events, endometrial cancer, and cataracts.2
Similar risks for ischemic heart disease, fractures, and
stroke were found for both drugs. There was a nonstatistically significant higher risk of noninvasive breast cancer
from raloxifene, which diminished at longer follow-up.2
Additional evidence for the efficacy of SERMs is derived
from the European randomized trials of tamoxifen compared
with placebo, which have demonstrated a more modest
benefit of tamoxifen for reducing the risk of invasive breast
cancer.3
The significant reduction in contralateral breast cancer
seen in the adjuvant aromatase inhibitor treatment trials
and the potentially serious toxicities associated with SERMs
led to the investigation of aromatase inhibitors for the
primary prevention of breast cancer. Goss et al published
the first results of a randomized placebo-controlled trial
designed to investigate exemestane for the primary prevention of invasive breast cancer. The National Cancer Institute
of Canada Clinical Trials Group MAP.3 randomly selected
4,560 postmenopausal women to receive exemestane or
resources to determine patient eligibility for chemoprevention

and lack the time to provide informed counseling to their
patients. The barriers for patients acceptance of chemoprevention treatment include fear of side effects and the perception that chemoprevention will not substantially lower their
risk of developing breast cancer. Despite these challenges,
there are substantial opportunities to increase the utilization
of chemoprevention. These strategies include education, dissemination of user-friendly risk benefit models, and the support of research efforts focused on identifying biomarkers that
can more accurately select women most likely to develop
breast cancer and predict responsiveness of treatment.
placebo. The inclusion criteria were similar to the BCPT and

STAR trials and included a 5-year predicted risk for breast
cancer of 1.66% or higher, history of atypia, or lobular
carcinoma in situ (LCIS). Women with ductal cancer in situ
were also eligible but represented a small portion of the
study participants (2.5%). At a median follow-up time of 35
months, exemestane was found to significantly reduce the
annual risk of invasive breast cancer by 65% (p 0.002).4
Arthritis (p 0.01) and hot flashes (p 0.001) were more
common in the exemestane treatment group, but there was
no statistically significant difference between the groups
in rates of new diagnoses of osteoporosis or cardiovascular
disease. Another phase III trial, IBIS-II, has randomly
selected 6,640 high-risk women to receive either placebo or
anastrozole and the results are eagerly awaited.
Evaluating RiskBenefit Indices for Chemoprevention
The FDA-approved indication to reduce the incidence of

invasive breast cancer for tamoxifen includes both premenopausal and postmenopausal women. However, raloxifene
which was FDA approved in 2007is indicated for postmenopausal women with osteoporosis or at high risk of
breast cancer based on the Gail model 5-year projected risk
of 1.66% or higher. Adverse events that are associated with
these agents raise concerns among women and physicians
about the benefits and risks, especially in women at high
risk of developing breast cancer who are otherwise healthy.
In addition to chronic diseases that accompany increasing
age, the risks of endometrial cancer, venous thromboembolic
events, bone fractures, and cataracts complicate the counseling for tamoxifen and raloxifene.
In 1998, the National Cancer Institute convened a workshop to determine the best ways of communicating the
results of the BCPT. The primary goal of this workshop was
From the University of Texas M. D. Anderson Cancer Center, Houston, TX; University of
Pittsburgh Cancer Institute, Pittsburgh, PA; and National Cancer Institute, Bethesda, MD.
Address reprint requests to Abenaa Brewster, MD, MHS, University of Texas M. D.
Anderson Cancer Center, 1155 Herman P. Pressler, PO Box 301439, Houston, TX 77230;
email: abrewster@mdanderson.org.
1092-9118/10/1-10
85
BREWSTER, DAVIDSON, AND STEVENS
to develop a tool to identify those women whose benefits

from the use of tamoxifen outweighs the risks in reducing
the incidence of breast cancer. Of particular interest was the
need to explore the influence of tamoxifen on those health
outcomes that disproportionately affect different populations. Such a tool was developed by Gail et al, which weighed
the benefits and risk of other health outcomes in the absence
of tamoxifen and the effects of tamoxifen on other health
outcomes.5 The benefits and risks for the use of tamoxifen
were described in the format of tables by race, age, presence
or absence of a uterus, and the projected 5-year risk of
invasive breast cancer. Net benefits increased with increasing risk and decreased with age and in postmenopausal
women with an intact uterus. Younger women with elevated
risks of breast cancer had the best net benefit for tamoxifen
use. Importantly, this study showed that the adverse events
varied by age and race, preferences by the women, and
varied weights for specific events were critical complements
for weighing risk and benefits of tamoxifen.
Limitations of the 1999 Gail et al data included the lack
of incidence rates for stroke and venous thromboembolic
events from populations other than the non-Hispanic white
population and the uncertainty of applying the estimated
benefits and risks to all populations. A comparison of the
benefits and risks of tamoxifen and raloxifene for postmenopausal women to assist in communication was identified as
a research and clinical need. In addition, the availability of
KEY POINTS
86
Evidence from randomized, placebo-controlled clinical trials supports the use of chemoprevention for
reducing the risk of primary invasive breast cancer,
and there are three options: tamoxifen, raloxifene,
and exemestane.
Risk benefit models have great potential for decision
making about chemoprevention in high-risk clinical
practices and primary care practices if used in combination with a full assessment of the womans health
and preferences.
Guidelines for the prescription of tamoxifen and
raloxifene are provided by the United States Preventive Services Task Force, American Society of Clinical
Oncology, the Canadian Task Force on Preventive
Health Care, and the National Comprehensive Cancer Network.
Only a small portion of women who are eligible for
chemoprevention receive treatment, and the barriers
to uptake include lack of education about breast
cancer, inadequate physician training in risk assessment, lack of time to provide counseling, concerns
about side effects, cost, and misconceptions about
actual and perceived risks and benefits of treatment.
An active area of cancer prevention research is the
identification of biomarkers and clinical features to
improve the selection of women most likely to benefit
from breast cancer chemoprevention and to predict
and monitor treatment responsiveness.
improved incidence rates for invasive cancer for Hispanic

women from the Surveillance, Epidemiology, and End Results (SEER), modifications to the Gail model prediction tool
for the black population, and baseline incidence rates for
other health outcomes from the Womens Health Initiative
could enrich a risk benefit analysis of tamoxifen and raloxifene.
Freedman et al evaluated the probability of having a
health event in 5 years in the absence and presence of
tamoxifen and raloxifene.6 Risks and benefits were defined
as life-threatening (invasive breast cancer, hip fractures,
endometrial cancer, stroke, and pulmonary emboli) and
severe events (in situ breast cancer and deep vein thrombosis). An index of expected number of life-threatening events
without chemoprevention in 10,000 women minus the expected number of life-threatening events if tamoxifen and
raloxifene were used was presented in color-coded tables.
The authors used a complex statistical modeling approach to
simulate weighted outcomes to provide levels of evidence for
risk estimation. Strong evidence of a positive net benefit of
tamoxifen or raloxifene compared with no treatment was
deemed to exist if the net benefit was positive in 90% or
greater of replications in Bayesian boot-strap testing for
variability. Freedman et al showed that raloxifene results in
a better risk benefit profile for black, Hispanic, and nonHispanic white postmenopausal women with a uterus when
compared with tamoxifen. Similar benefits for postmenopausal black and non-Hispanic white women without a
uterus were seen for tamoxifen and raloxifene.
There are several limitations of the tables presented by
Freedman et al. Since the Gail model performs poorly at
predicting risk of breast cancer among women at higher risk
levels, generalization of the results to these women is
limited. In addition, the benefit of raloxifene over tamoxifen
may continue to diminish with longer-term follow-up and
the time line of the analysis was limited to 5 years.7 Despite
these limitations, risk benefit models have great potential
of assisting cancer care providers in high-risk clinical practices and primary care practices if used in combination with
a full assessment of the womans health and preferences.
Guidelines for Breast Cancer Chemoprevention
In 2002, the United States Preventive Services Task Force

(USPSTF) issued a recommendation statement that women
who are at both high risk of developing breast cancer and
low risk for adverse events should be counseled about
chemoprevention for breast carcinoma.8 Chemoprevention
was not recommended for women at low or average risk for
whom the adverse health events associated with tamoxifen
or raloxifene may outweigh the benefits. Although no set
definition of high risk was specified, the USPSTF referred
to the entry eligibility for the BCPT, which was a predicted
5-year risk of greater than 1.67% defined using the Gail
model.1 This guideline for the prescription of tamoxifen and
raloxifene has been supported by the American Society of
Clinical Oncology (ASCO), the Canadian Task Force on
Preventive Health Care, and the National Comprehensive
Cancer Network (NCCN).9-11 Based on the results of the
MAP.3 trial, exemestane was included as one of the choices
for breast cancer chemoprevention by the NCCN panel of
experts. Their recommendation for exemestane was limited
to postmenopausal women age 35 or older with a Gail model
5-year predicted risk higher than 1.66% or a history of LCIS.
CHEMOPREVENTION FOR BREAST CANCER
There are no data directly comparing the benefits and risks

of exemestane with those of tamoxifen and raloxifene and it
is not currently FDA approved for primary breast cancer
risk reduction.
Uptake of Breast Cancer Chemoprevention in the
United States
Despite clinical guidelines recommending tamoxifen or

raloxifene for the risk reduction of primary breast cancer,
the majority of eligible women elect not to use chemoprevention. It is estimated that approximately 10 million women
(16%) would be eligible for chemoprevention and approximately 2.4 million would have a positive risk benefit profile
for tamoxifen treatment.12 Unfortunately the uptake of
breast chemoprevention has been substantially less than
predicted. Waters et al used data from the National Health
Interview Survey between 2000 and 2005 to estimate the
number of women taking tamoxifen for chemoprevention.13
In 2000, two years after the FDA approved tamoxifen for
breast cancer chemoprevention, only 0.2% of women ages 40
to 79 had taken tamoxifen to prevent a diagnosis of breast
cancer. In 2005, 0.08% of women had taken tamoxifen,
which represented a 50% decrease in the number of women
receiving tamoxifen from 120,000 in 2000 to 60,000 in
2005.13 Armstrong et al surveyed 350 primary care physicians between 2002 and 2004 and found that only 27% had
prescribed tamoxifen for breast cancer prevention in the
previous year, which was higher than would have been
predicted based on the prevalence of women reporting taking the medication.14 There are no available data available
on the uptake of raloxifene, however, the general perception
is that public acceptance remains low for reasons that
remain unclear given its lower toxicity profile compared
with tamoxifen.15
Physician Barriers to Prescribing Breast Cancer
Chemoprevention
It is well recognized that women are more likely to accept

preventive services if recommended by their physicians.16-18
The reluctance of physicians to prescribe tamoxifen or raloxifene is therefore likely to be a major reason for the low
uptake of chemoprevention. Since primary care physicians
(PCPs) serve as the gatekeepers for implementing screening
recommendations and usually have the first interaction with
women at increased risk of developing breast cancer, studies
investigating physicians attitudes and acceptance of chemoprevention have focused on this group of health care providers.
The PCP role in the preventive care setting is to assess a
womans risk of developing breast cancer and provide a
recommendation for a risk-reduction option. How a PCP
views a womans risk of developing breast cancer has been
shown to be an important predictor of whether a prescription of tamoxifen will be offered.19,20 Tchou et al found that
women with a histologic diagnosis of atypical hyperplasia
or LCIS and women with a Gail model 5-year predicted risk
of 5.0% or greater were more likely to be offered tamoxifen
than women at lower risk.19 In a random sample of 822
Californian PCPs, Haas et al presented hypothetical patient scenarios to assess how a womans breast cancer risk
would influence her physicians recommendations for riskreduction options.20 Although 80% of PCPs would recom-
mend mammography screening and counseling for high-risk

women on lifestyle modifications, a minority endorsed the
use of tamoxifen. Some of the physician-perceived barriers
to risk-reduction counseling included not sufficiently
trained in counseling techniques and not sufficiently informed about risk-reduction options. Lack of time with
patients was the most frequently cited barrier, and 13% of
physicians reported insufficient reimbursement as an impediment to counseling.21 Interestingly, physicians concern
about the side effects of chemoprevention has not been
reported to be a significant deterrent to prescribing tamoxifen or raloxifene.14 However, more PCPs report prescribing
raloxifene (30%) compared with tamoxifen (10%) for primary
breast cancer risk reduction, which is not surprising since
Freedman et al showed that raloxifene results in a better
risk benefit profile for women with a uterus when compared
with tamoxifen.6,21 A good understanding of how to determine a womans eligibility for tamoxifen and ability to
determine whether the benefits of treatment outweigh the
risks is a significant predictor of whether a PCP prescribes
tamoxifen. Unfortunately in two surveys of PCPs conducted
by Armstrong et al and Sabatino et al, only 15% and 9% of
PCPs, respectively, felt capable of making this clinical assessment, which highlights the challenges faced by PCPs in
using and interpreting the risk benefit models that have
been developed.14,22 Contrary to the experience of PCPs, a
survey of 851 oncologists revealed that 73% discussed breast
cancer chemoprevention with their patients.23 The high
percentage of oncologists who discuss chemoprevention recommendations compared with PCPs may reflect differences
in the receipt of formal and informal training in cancer
prevention and encounters with women in the high-risk
clinical setting who are highly motivated to reduce their risk
of developing breast cancer.
Patient Barriers to Accepting Breast Cancer
Chemoprevention
Few women who meet eligibility criteria based on an

assessment of the risks and benefits of chemoprevention
treatment elect to initiate treatment. Ropka et al conducted
a meta-analysis of studies that evaluated real or hypothetical decisions made by women about breast cancer chemoprevention. The mean uptake of tamoxifen for the five
studies that reported real decision rates was 14.8% (range,
0.5% to 51.2%) and 24% (range, 5.7% to 60%) for the studies
reporting hypothetical decisions.24
The most commonly cited reason for not taking tamoxifen
is fear of the side effect of endometrial cancer, thromboembolic disease, and menopausal symptoms.16,25,26 In a study
of 43 women who were eligible to take tamoxifen for primary
prevention, the most feared side effects were endometrial
cancer and thromboembolic disease, and the majority of
women tended to overestimate their risk of developing these
side effects even after a presentation of the balance of risks
and benefits.25 When asked to estimate the risk of developing a side effect associated with tamoxifen, the majority
perceived that the risk would be 40% or greater. A survey of
29 women interviewed in a family practice setting revealed
additional concerns about nausea and vomiting, symptoms
mostly attributed to chemotherapy. In this study, the authors proposed that the name chemoprevention might
itself act as a barrier to patient acceptance because it is often
confused with chemotherapy.26
87
Previous studies have shown that women tend to overestimate their risk of developing breast cancer and women
with a higher perceived risk of breast cancer are more likely
to be accepting of chemoprevention.16,27 Many women, however, have the perception that chemoprevention will not
substantially lower their risk of developing breast cancer
even after receiving information about the 50% risk reduction associated with tamoxifen treatment.28,29 Acceptance
of chemoprevention treatment has been shown to be significantly higher among women with a history of atypical
hyperplasia or LCIS compared with patients at risk on the
basis of other factors (56% vs. 28%, p 0.0001).19
Additional barriers that contribute to the reluctance of
women to use chemoprevention are related to physician
difficulty in accurately selecting individuals most likely to
develop breast cancer and the lack of a biomarker that can
be used to monitor the effect of the drug on risk. Unlike the
monitoring of cardiovascular risk factors (e.g., cholesterol
and blood pressure), which provide a measurable target for
assessing the efficacy of cholesterol or hypertension lowering
agents, there is no available reliable biomarker to measure
the preventive effect of chemoprevention that may serve to
motivate patients to accept treatment.15 Salant et al interviewed women seen at a high-risk clinic, of which the
majority were black, and found that the women understood
risk not as a numerical probability or chronic disease state
but as an immediate physical sign or symptom warranting
medical intervention. Therefore, despite meeting criteria for
being high-risk using the Gail model, many women did not
feel at high risk and therefore were not interested in
taking breast cancer chemoprevention treatment.30 This
finding has implications for the support of research efforts
aimed at identifying biomarkers and clinical features that
identify high-risk states and individualize the risk prediction of invasive breast cancer as a means of improving the
uptake of chemoprevention treatment.31
Among low-income women, acceptance of chemoprevention is dependent also on how much it costs and whether the
cost is covered by health insurance.28,30 At least three
studies have found that education and lower income are
inversely associated with breast cancer chemoprevention.27,32,33 These studies did not report on the accuracy of
the womens understanding of the risks and benefits of the
chemoprevention treatments, which is an important consideration since a greater understanding of the risk benefit
profile of chemoprevention has been demonstrated to result
in decreased patient acceptance of chemoprevention treatment.24 To overcome the significant patient-related barriers
to the uptake of chemoprevention, a comprehensive strategy
is needed that educates women about breast cancer and
their competing health risks and incorporates easily accessible decision aids that accurately convey the risks and
benefits of chemoprevention treatment.
Strategies to Improve the Uptake of Breast Cancer
Chemoprevention
Several investigators have explored the use of educational

or decision-making aids to increase the uptake of breast
cancer chemoprevention.34,35 The goal of the decision aids is
to facilitate a discussion between the patient and provider
about breast cancer chemoprevention and help guide informed decisions.36 Key components of the decision aids
88
have included education about breast cancer, an assessment

of the patients risk of breast cancer using the Gail or Claus
risk assessment models, integration of the patients risk of
breast cancer within the context of population risk, and an
assessment of the change in risks and benefits that would be
expected with tamoxifen or raloxifene treatment.34-36 In two
studies that have evaluated patient willingness to take
chemoprevention after receiving a decision aid, the uptake
rates have been disappointingly low (range, 4% to 6%).34,35
In a randomized trial designed to test whether a decision aid
intervention increased a womans interest in chemoprevention, among the 6% of high-risk women who received the
decision aid and said that they were likely to take the drug,
less than 1% had initiated therapy after the intervention.35
The challenges for PCPs considering prescribing breast
cancer chemoprevention are identifying women eligible for
treatment based on an acceptable risk benefit profile and
communicating the net benefit estimate using a balanced
approach that allows the patient to consider her personal
preferences. Increased PCP awareness of the online Gail
risk model and access to the risk benefit tools for tamoxifen
that were published in 1999 and updated in 2011 to include
raloxifene are needed to facilitate this process. It has been
advocated that the risk benefit models for breast cancer
chemoprevention should be more interactive and available
online, accessible to both physicians and patients.37 In
addition, the risk benefit models should incorporate additional variables, such as 5-year and lifetime mortality,
comorbidities including menopausal symptoms, and competing health risks that are essential to an informed decisionmaking discussion about whether to use chemoprevention
treatment.37,38 To increase awareness and education about
cancer preventive and control strategies, the ASCO Cancer
Prevention Committee is in the process of developing curricular material to disseminate to the educational programs
of a spectrum of disciplines, including family practice and
internal medicine. A recent breast cancer consensus statement recommended that to advance efforts for breast cancer
prevention, the term chemoprevention should be replaced
with the term preventive therapy to remove its inappropriate association with chemotherapy.39
A major hurdle in promoting breast cancer chemoprevention to the general population is the recognition that many
women who receive chemoprevention treatment will never
have developed breast cancer anyway. Because of the poor
discriminatory accuracy of the Gail risk model in determining individual level risk of breast cancer, many women and
their physicians remain uncertain about the benefits of
treatment and concerned about the potential serious toxicities and side effects that may affect quality of life. In
addition, advocacy groups including the Breast Cancer National Coalition are skeptical of chemoprevention and express concern about the lack of data on the long term
side-effects of the chemoprevention drugs.40 Strategies to
refine risk and predict responsiveness to chemoprevention
treatment are being explored and include random periareolar fine-needle aspiration to evaluate for cellular atypia in
otherwise normal breast tissue, breast density as a surrogate marker for monitoring response to treatment, and
single nucleotide polymorphisms associated with increased
susceptibility.36,39
CHEMOPREVENTION FOR BREAST CANCER

Conclusion
Breast cancer is the most common malignancy among

women in the United States affecting approximately
225,000 women annually. Evidence from randomized,
placebo-controlled clinical trials supports the use of chemoprevention for reducing the risk of primary invasive breast
cancer, and there are three options: tamoxifen, raloxifene,
and exemestane. Risk benefit models indicate that raloxifene results in a better risk benefit profile for black, Hispanic, and non-Hispanic white postmenopausal women with
a uterus when compared with tamoxifen. Similar benefits
for postmenopausal black and non-Hispanic white women
without a uterus are seen for tamoxifen and raloxifene. Only
a small portion of women who are eligible for chemopreven-
tion receive treatment, and the barriers to uptake include

lack of education about breast cancer, inadequate physician
training in risk assessment, lack of time to provide counseling, concerns about side effects, cost, and misconceptions
about actual and perceived risks and benefits of treatment.
Strategies to close this gap have focused on educating
women about breast cancer and their competing health risks
as well as developing easily accessible decision aids and risk
assessment models that accurately convey the risks and
benefits of chemoprevention treatment. An active area of
research is the identification of biomarkers and clinical
features to improve the selection of women most likely to
benefit from chemoprevention and to predict and monitor
responsiveness of treatment.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Abenaa M. Brewster*
Nancy E. Davidson
GlaxoSmithKline
(U)
Breast Cancer
Research
Foundation; NIH
Worta McCaskill-Stevens*
REFERENCES
1. Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen for prevention
of breast cancer: report of the National Surgical Adjuvant Breast and Bowel
Project P-1 Study. J Natl Cancer Inst. 1998;90:1371-1388.
2. Vogel VG, Costantino JP, Wickerham DL, et al. Update of the National
Surgical Adjuvant Breast and Bowel Project Study of Tamoxifen and Raloxifene (STAR) P-2 Trial: Preventing breast cancer. Cancer Prev Res (Phila).
2010;3:696-706.
3. Nelson HD, Fu R, Griffin JC, et al. Systematic review: comparative
effectiveness of medications to reduce risk for primary breast cancer. Ann
Intern Med. 2009;151:703-715.
4. Goss PE, Ingle JN, Ales-Martinez JE, et al. Exemestane for breastcancer prevention in postmenopausal women. N Engl J Med. 2011;364:23812391.
5. Gail MH, Costantino JP, Bryant J, et al. Weighing the risks and benefits
of tamoxifen treatment for preventing breast cancer. J Natl Cancer Inst.
1999;91:1829-1846.
6. Freedman AN, Yu B, Gail MH, et al. Benefit/risk assessment for breast
cancer chemoprevention with raloxifene or tamoxifen for women age 50 years
or older. J Clin Oncol. 2011;29:2327-2333.
7. Amir E, Goodwin P. Breast cancer chemoprevention gets personal. J
Clin Oncol. 2011;29:2296-2298.
8. U.S. Preventive Services Task Force. Chemoprevention of breast cancer:
recommendations and rationale. Ann Intern Med. 2002;137:56-58.
9. Visvanathan K, Chlebowski R, Hurley P. American Society of Clinical
Oncology practice guideline update on the use of pharmacologic interventions
including tamoxifen, raloxifene, and aromatase inhibition for breast cancer
risk reduction. J Clin Oncol. 1999;27:3235-3258.
10. Levine M, Moutquin JM, Walton R, et al. Chemoprevention of breast
cancer. A joint guideline from the Canadian Task Force on Preventive Health
Care and the Canadian Breast Cancer Initiatives Steering Committee on
Clinical Practice Guidelines for the Care and Treatment of Breast Cancer.
CMAJ. 2001;164:1681-1690.
11. National Comprehensive Cancer Network. NCCN clinical practice
guidelines in oncology: Breast cancer risk reduction, v. 3.2011. http://www.
nccn.org/professionals/physician_gls/f_guidelines.asp#detection. Accessed February 13, 2012.
12. Freedman AN, Graubard BI, Rao SR, et al. Estimates of the number of
US women who could benefit from tamoxifen for breast cancer chemoprevention. J Natl Cancer Inst. 2003;95:526-532.
13. Waters EA, Cronin KA, Graubard BI, et al. Prevalence of tamoxifen use
for breast cancer chemoprevention among U.S. women. Cancer Epidemiol
Biomarkers Prev. 2010;19:443-446.
14. Armstrong K, Quistberg DA, Micco E, et al. Prescription of tamoxifen
for breast cancer prevention by primary care physicians. Arch Intern Med.
2006;166:2260-2265.
15. Meyskens FL Jr., Curt GA, Brenner DE, et al. Regulatory approval of
cancer risk-reducing (chemopreventive) drugs: Moving what we have learned
into the clinic. Cancer Prev Res (Phila). 2011;4:311-323.
16. Bober SL, Hoke LA, Duda RB, et al. Decision-making about tamoxifen
in women at high risk for breast cancer: clinical and psychological factors.
J Clin Oncol. 2004;22:4951-4957.
17. Lerman C, Rimer B, Trock B, et al. Factors associated with repeat
adherence to breast cancer screening. Prev Med. 1990;19:279-290.
18. Kinney AY, Richards C, Vernon SW, et al. The effect of physician
recommendation on enrollment in the Breast Cancer Chemoprevention Trial.
Prev Med. 1998;27:713-719.
19. Tchou J, Hou N, Rademaker A, et al. Acceptance of tamoxifen chemoprevention by physicians and women at risk. Cancer. 2004;100:1800-1806.
20. Haas JS, Kaplan CP, Gregorich SE, et al. Do physicians tailor their
recommendations for breast cancer risk reduction based on patients risk?
J Gen Intern Med. 2004;19:302-309.
21. Kaplan CP, Haas JS, Perez-Stable EJ, et al. Factors affecting breast
cancer risk reduction practices among California physicians. Prev Med.
2005;41:7-15.
22. Sabatino SA, McCarthy EP, Phillips RS, et al. Breast cancer risk
assessment and management in primary care: provider attitudes, practices,
and barriers. Cancer Detect Prev. 2007;31:375-383.
23. Ganz PA, Kwan L, Somerfield MR, et al. The role of prevention in
oncology practice: results from a 2004 survey of American Society of Clinical
Oncology members. J Clin Oncol. 2006;24:2948-2957.
24. Ropka ME, Keim J, Philbrick JT. Patient decisions about breast cancer
chemoprevention: a systematic review and meta-analysis. J Clin Oncol.
2010;28:3090-3095.
25. Port ER, Montgomery LL, Heerdt AS, et al. Patient reluctance toward
tamoxifen use for breast cancer primary prevention. Ann Surg Oncol. 2001;
8:580-585.
26. Heisey R, Pimlott N, Clemons M, et al. Womens views on chemoprevention of breast cancer: qualitative study. Can Fam Physician. 2006;52:624625.
27. Melnikow J, Paterniti D, Azari R, et al. Preferences of Women Evaluating Risks of Tamoxifen (POWER) study of preferences for tamoxifen for
breast cancer risk reduction. Cancer. 2005;103:1996-2005.
28. Cyrus-David MS, Strom SS. Chemoprevention of breast cancer with
selective estrogen receptor modulators: views from broadly diverse focus
groups of women with elevated risk for breast cancer. Psychooncology.
2001;10:521-533.
89

29. Fagerlin A, Zikmund-Fisher BJ, Nair V, et al. Womens decisions
regarding tamoxifen for breast cancer prevention: responses to a tailored
decision aid. Breast Cancer Res Treat. 2010;119:613-620.
30. Salant T, Ganschow PS, Olopade OI, et al. Why take it if you dont
have anything? breast cancer risk perceptions and prevention choices at a
public hospital. J Gen Intern Med. 2006;21:779-785.
31. Temple R. Cancer chemopreventionthe cardiovascular model. Cancer
Prev Res. 2011;4:307-310.
32. Fasching PA, von Minckwitz G, Fischer T, et al. The impact of breast
cancer awareness and socioeconomic status on willingness to receive breast
cancer prevention drugs. Breast Cancer Res Treat. 2007;101:95-104.
33. Tija J, Micco E, Armstrong K. Interest in breast cancer chemoprevention among older women. Breast Cancer Res Treat. 2008;108:435-453.
34. McKay A, Martin W, Latosinsky S. How should we inform women at
higher risk of breast cancer about tamoxifen? An approach with a decision
guide. Breast Cancer Res Treat. 2005;94:153-159.
90
35. Fagerlin A, Dillard AJ, Smith DM, et al. Womens interest in taking
tamoxifen and raloxifene for breast cancer prevention: response to a tailored
decision aid. Breast Cancer Res Treat. 2011;127:681-688.
36. Ozanne EM, Klemp JR, Esserman LJ. Breast cancer risk assessment
and prevention: a framework for shared decision-making consultations.
Breast J. 2006;12:103-113.
37. Ravdin PM. The lack, need, and opportunities for decision-making and
informational tools to educate primary-care physicians and women about
breast cancer chemoprevention. Cancer Prev Res (Phila). 2010;3:686-688.
38. Mulley AG, Sepucha K. Making good decisions about breast cancer
chemoprevention. Ann Intern Med. 2002;137:52-54.
39. Cuzik J, DeCensi A, Arun B. Preventive therapy for breast cancer: a
consensus statement. Lancet Oncol. 2011;12:496-503.
40. The National Breast Cancer Coalition position statement on chemoprevention. Updated June 2011. www.breastcancerdeadline2020.org/know/
assets/... /chemoprevention.pdf. Accessed March 7, 2012.
CONTROVERSIES IN PROSTATE CANCER:

PSA SCREENING, CHEMOPREVENTION, AND
TREATMENT OF EARLY DISEASE
CHAIR
Barnett S. Kramer, MD, MPH
Rockville, MD
SPEAKERS
Gerald L. Andriole, MD
Washington University School of Medicine
St. Louis, MO
Timothy J. Wilt, MD, MPH
U. S. Department of Veterans Affairs
Minneapolis, MN
The Case for Prostate Cancer Risk Reduction

by 5-Alpha Reductase Inhibitors
By Youssef S. Tanagho, MD, MPH, and Gerald L. Andriole, MD
Overview: Prostate cancer remains a significant public

health problem. The current approach with prostate-specific
antigen (PSA)-based screening has questionable effects on
prostate cancerspecific mortality but is clearly associated
with overdiagnosis of prostate cancer, especially relatively
low-risk and low-volume tumors. Methods to decrease overdiagnosis include alterations in screening practices and, potentially, the use of 5-alpha reductase inhibitors. This article
reviews the major trials that have evaluated 5-alpha reductase
inhibitors in this setting: the Prostate Cancer Prevention Trial
(PCPT) and the Reduction by Dutasteride Prostate Cancer
Events Trial (REDUCE). Although these trials enrolled different
ROSTATE CANCER risk reduction is a potentially

valuable strategy, as current approaches with PSAbased screening, which have a potentially small effect on
prostate cancer mortality, are associated with a significant
risk of overdiagnosis of prostate cancer.1-3 Most men diagnosed with PSA-detected tumors in the United States receive aggressive treatments.4,5 These treatments often have
significant side effects and are costly in both human and
economic terms. Therefore, one plausible strategy to improve the efficacy of PSA-based screening is to consider the
use of 5-alpha reductase inhibitors, as they may reduce
overdiagnosis of prostate cancer and improve the utility of
PSA as a marker for aggressive disease,6-9 as well as reduce
adverse sequelae of BPH.
Overdiagnosis of Prostate Cancer
Welch and Black10 identified three factors that result in

overdiagnosis of prostate cancer: 1) the existence of a silent
disease reservoir; 2) the use of activities (such as screening)
that lead to its detection; and 3) tumors with a long natural
history and, hence, limited cancer-specific mortality. Prostate cancer may represent a textbook disease for overdiagnosis. Multiple studies have shown a high prevalence of this
disease on autopsy, ranging from approximately 30% of men
in their 30s to up to 80% of men in their 80s. Thus, there is
a large silent reservoir of this disease that could be clinically
detected.11-12 Second, it has been estimated that close to
70% of American men have undergone one or more PSA
tests, and evidence suggests that older men, those in their
70s and 80s, who stand to benefit the least from PSA-based
screening represent the age stratum most likely to be
screened.13 Moreover, use of relatively low PSA thresholds
as a guide to select patients for biopsy could result in the
diagnosis of prostate cancer in a high proportion of men in
their 60s to 80s.14 It has been estimated that 6 to 10 times as
many men are considered to have an abnormal PSA than are
destined to die of prostate cancer. Finally, most cases of
prostate cancer have a long natural history and a limited
cancer-specific mortality. Albertsen and colleagues15 looked
at survival among men with localized prostate cancer. Men
with Gleason score 8 to 10 disease and no comorbidities had
more than twice the chance of dying of other causes than of
prostate cancer within 10 years. Men who had one or more
significant comorbidities and Gleason score 8 to 10 disease
92
patient populations, their findings are complementary and

suggest a potential role for these agents in prostate cancer
risk reduction. Use of 5-alpha reductase inhibitors results in
an approximate 25% reduction in the detection of prostate
cancer, reduces diagnosis of high-grade prostatic intraepithelial neoplasia (PIN), and improves benign prostatic hyperplasia
(BPH)-related outcomes and the performance of PSA as a
diagnostic test for aggressive prostate cancer. Side effects
occur in a small percentage of men and consist of decreased
sexual function and libido as well as gynecomastia. The risk of
high-grade tumor development while receiving these agents is
uncertain.
were about 5 times as likely to have a nonprostate cancer

related death. A contemporary clinical trial, the Goteborg
screening study, demonstrated that in a screened population
of about 20,000 Scandinavian mena population demographically predisposed to a relatively high overall death
rate from prostate cancer only 122 died of prostate cancer
and more than 3,800 died of other causes at 14 years of
follow-up.3 Similarly, of 367 men with screen-detected, localized prostate cancers who were considered candidates for
radical prostatectomy in the observation arm of the U.S.
PIVOT trial, 31 died of prostate cancer, whereas 152 died of
other causes at a median 10-year follow-up.16
Estimates of the magnitude of overdiagnosis of prostate
cancer can be made by comparing the screened to the usual
care arms of randomized screening trials. The PLCO trial,
which enrolled men across 10 participating U.S. centers,
showed that the screened arm had a 17% (95% CI: 1122)
increased chance of diagnosis of prostate cancer compared
with the usual care arm at 7 to 10 years of follow-up.4
However, it is known that the usual care arm for PLCO was
heavily contaminated with PSA testing. If one were to
compare the usual care arm of PLCO to a matched contemporary population in the United States, the usual care arm
would have a 22% excess prostate cancer detection rate.
Furthermore, men in the usual care arm of PLCO had more
than twice as many prostate cancers discovered as would
have been otherwise anticipated for a similar group of men
in the pre-PSA era.17 In ERSPC, a multicenter European
trial, there was a 71% increase in the detection of prostate
cancer in the screened arm compared with the usual care
arm, and a 64% increase was noted in the Goteborg trial. In
most European sites, screening was performed at a 4-year
interval, whereas in Scandinavia, it generally occurred at a
2-year interval. One illustration of the effect of overdiagnosis of prostate cancer was offered by Caroll and colleagues,18
who used the Goteborg data to demonstrate that if 1,000
From the Division of Urology, Washington University School of Medicine, St. Louis, MO.
Address reprint requests to Gerald L. Andriole, MD, Division of Urologic Surgery,
Washington University in St. Louis, 4960 Childrens Place, Campus Box 8242, St. Louis,
MO 63110; email: andrioleg@wustl.edu.
1092-9118/10/1-10
PROSTATE CANCER RISK REDUCTION
men were followed for 14 years, screening would avert the

death of five of nine men destined to die of prostate cancer
while leading to the diagnosis of prostate cancer in 120 men.
In addition to the considerable laboratory expenses associated with PSA screening, the anxiety associated with
abnormal PSA test results and the known significant side
effects of prostate biopsyincluding sepsis and hospitalizationrepresent significant costs of prostate cancer overdiagnosis. Indeed, the largest cost associated with prostate
cancer overdiagnosis is overtreatment. In a Scandinavian
study, Fall and colleagues showed that during the first week
following the diagnosis of prostate cancer, patients experienced a 2.8-fold (95% CI: 2.53.2) increased relative risk of a
major cardiovascular event and a 8.4-fold (95% CI: 1.9 22.7)
increased risk of suicide.19 In PLCO, more than 90%
of screen-detected cancers have received aggressive therapy.4 Certainly, earlier detection and treatment of prostate
cancer in the PSA era have contributed to the decreasing
mortality of this disease. Nonetheless, the significant human and economic burden of prostate cancer overtreatment
presents a considerable challenge to current PSA-based
screening efforts.
Agents to Reduce Overdiagnosis of Prostate Cancer
Plausible approaches to prostate cancer risk reduction

include a variety of agents. The 5-alpha reductase inhibitors
and vitamin E and selenium have been the best studied.
SELECT evaluated vitamin E and selenium and failed to
show any reduction in the diagnosis of prostate cancer.20 A
more recent analysis of the data suggested that vitamin E
may actually increase the detection of prostate cancer.21
Five-alpha reductase inhibitors are a group of drugs with
antiandrogenic activity, commonly used in the treatment of
BPH. These agents exert their antiandrogenic effect by
blocking the conversion of testosterone to the active form
dihydrotestosterone (DHT); the latter acts as a potent cellular androgen that promotes prostate growth. Two isoenzymes of 5-alpha reductase are found in the human genome,
Type 1 and Type 2. Type 2 appears to be the more important
isoform in the pathophysiology of BPH, whereas both Types
1 and 2 are expressed in cancer and cancer-related (PIN)
tissue. Finasteride is a selective inhibitor of 5-alpha reductase Type 2, while dutasteride is an inhibitor of both isoenzymes. Both drugs have been linked to decreased sexual
KEY POINTS
Prostate-specific antigen (PSA)-based screening is

associated with overdiagnosis of prostate cancer.
Most prostate cancers detected in the United States
are treated aggressively.
Five-alpha reductase inhibitors decrease the detection
of low-risk prostate cancers during PSA screening.
It is uncertain whether 5-alpha reductase inhibitors
used in this setting promote aggressive prostate cancer.
Five-alpha reductase inhibitors improve the performance of PSA as a diagnostic test for aggressive
prostate cancer.
Table 1. Comparison of PCPT and REDUCE

PCPT
Design
Duration (yr)
Number of patients
Location
Age
Entry PSA
5- reductase inhibition
Negative baseline biopsy
Follow-up biopsy
Cores per study biopsy
Baseline Characteristics
Median age
Median PSA
Prostate Cancer Results
Placebo
Treated
7
18,882
United States
55
3
Type 2 only
No
7 yr FC
6
REDUCE
4
8,251
International
5080
2.510
Types 1 and 2
Yes
2 yr and 4 yr FC
10
63
1.1
63
5.7
24.4%
18.4%
21.1%
16.3%
Abbreviations: FC, for cause; PSA, prostate-specific antigen; yr, years.
function and libido as well as gynecomastia in a small subset

of men. In general, men apt to experience these side effects
do so during the first year; thereafter, rates are similar
between placebo and treated men. Most side effects are
reversible if the drug is stopped.
Five-alpha reductase inhibitors have been studied in two
large prospective randomized trials, both of which have
demonstrated a reduced incidence of prostate cancer diagnosis (see Table 1). PCPT evaluated men age 55 or older with
serum PSA values of 3 who were randomly selected to
receive placebo or 5 mg per day of finasteride and underwent
for cause and end-of-study biopsies at 7 years.22 This trial
showed a 25% (95% CI: 18.6 30.6, p 0.001) statistically
significant risk reduction in the diagnosis of prostate cancer
in men treated with finasteride. However, this study has
been criticized for three main reasons. First, there was a
very high detection rate for prostate cancer over the 7-year
course of the trial. This low-risk group of men in the
placebo arm (who had a normal rectal exam and a PSA 3)
had a 24.4% chance of a prostate cancer diagnosis after 7
years of follow-up. In contrast, average U.S. men have a 17%
lifetime risk of prostate cancer. Second, there were fewer
for cause biopsies in the finasteride arm. If the biopsy rates
had been equal between the two arms, the difference in
overall prostate cancer detection rates would have been
narrowed. Finally, a major concern of PCPT was the finding
on biopsy of more prostate cancers with a Gleason score of 7
or higher in the finasteride arm. This excess of Gleason score
7 to 10 cancer has been attributed to biases induced by
finasteridethe major ones being the increased sensitivity
of PSA and the volume bias. The latter reflects the effect of
prostate size on the ability of digital rectal exam and biopsy
to detect prostate cancer. Four independent groups have
adjusted the original data of PCPT for these biases against
finasteride. The 27% excess Gleason score 7 to 10 disease in
the finasteride arm is eliminated.23-26 It was on the basis of
these findings that ASCO and the American Urological
Association together recommended that men with a PSA of
3 or below who are undergoing screening with PSA should
be encouraged to talk to their physicians about the risks
and benefits of taking a 5-alpha reductase inhibitor.27 More
recently, the U.S. Food and Drug Administration (FDA)
used other models to adjust for bias in PCPT. One of these
models showed a persistent 1.51 (95% CI: 1.01 to 2.26)
93
TANAGHO AND ANDRIOLE
excess risk of Gleason score 8 to 10 prostate cancer in the

finasteride arm.28
The REDUCE trial enrolled men between the age of 50
and 75 who are considered to be at high risk for prostate
cancer because of an elevated PSA (between 2.5 and 10).29
Patients were required to have undergone a negative
prestudy biopsy that was centrally reviewed. Patients in the
REDUCE trial received 0.5 mg of dutasteride per day or
placebo and underwent protocol-specified biopsies at 2 years
and 4 years as well as for cause biopsies whenever indicated. During the first 2 years of the REDUCE trial, there
was statistically significant 22.4% (95% CI: 13.0 30.8, p
0.001) relative risk reduction in the diagnosis of prostate
cancer.
The design of the REDUCE trial with a planned second
biopsy at 4-year follow-up for men without cancer on either
the initial prestudy biopsy or the 2-year study-mandated
biopsy is important to consider. One would anticipate that at
randomization there would be an equal number of small
volume cancers in each arm. After the discovery of 142 more
cancers in the placebo arm and their removal from the study
following the year-2 biopsies, there were more men and more
cancers remaining in the dutasteride arm, thus giving rise to
an imbalance between the two study groups for the remaining duration of the study. Despite these imbalances, a
statistically significant 23.7% (95% CI: 9.9 35.3, p 0.001)
relative risk reduction in the detection of prostate cancer
was again noted during the second round of biopsies in
REDUCE. The finding of fewer cancer cases in the dutasteride arm despite three significant biases against dutasteride during the second round of biopsies (i.e., more cancers
waiting to be discovered, more men undergoing biopsy, and
a 35% smaller prostate) suggests that dutasteride is capable
of continually suppressing tumor growth.
There was no excess of Gleason score 7 to 10 cancers in the
dutasteride arm of REDUCE. There was, however, a numerical increase in Gleason score 8 to 10 cancers in the dutasteride arm (29 vs. 19). This excess occurred exclusively
during the second round of biopsies. During the first round,
there were 17 Gleason score 8 to 10 cancers in the dutasteride arm and 18 in the placebo arm. During the second
round of biopsies, there were 12 in the dutasteride arm and
one in the placebo arm. An explanation for this excess biopsy
Gleason score 8 to 10 during the second round of the
REDUCE trial has been offered by considering that the
excess Gleason score 5 to 7 cancers diagnosed in the placebo
group during the first round of biopsies almost certainly
included the cancers of some men in whom biopsy-detectable
Gleason score 8 to 10 cancers were actually present. Using
data from the study of Choo and colleagues that considered
men with Gleason score 4 to 7 prostate cancer who were on
active surveillance and underwent follow-up biopsy, approximately 8% of such patients were found to have Gleason
score 8 to 10 disease on follow-up biopsy. Eight percent of the
141 excess cancers detected in years 1 and 2 in the placebo
arm of REDUCE would project 11 additional Gleason score
8 to 10 tumors in the placebo arm. Thus, there was a
virtually identical number of Gleason score 8 to 10 cancers
in both arms during the entire study.
Other studies have looked at the occurrence of Gleason
score 8 to 10 cancers in men receiving dutasteride. These
include the COMBAT and the REDEEM studies. Neither of
these trials showed any increase in Gleason score 8 to 10
94
cancer in the dutasteride-driven arm after 3 to 4 years,

respectively.30,31
When one considers the raw data from the REDUCE trial
and adjusts it for the bias induced by prostate shrinkage and
increased PSA sensitivity in the dutasteride arm, the odds
of Gleason 7 to 10 cancer in the dutasteride arm actually
decreased by 38%.
In preparation for the Oncology Drugs Advisory Committee meeting in December 2010, cancers in the REDUCE trial
were reread using a modified Gleason score. The REDUCE
protocol specified that tumors would be graded using the
classic Gleason scoring system. When cancers in REDUCE
were reread using a modified Gleason scoring system, similar to the one used in PCPT, the number of Gleason score 8
to 10 cancers in the dutasteride arm rose from 29 to 32,
whereas the number in the placebo arm fell from 19 to 16.
Thus, there were twice as many Gleason score 8 to 10
cancers in the dutasteride arm when the modified Gleason
score was used. This doubling of Gleason score 8 to 10
cancers suggested a safety signal and the potential induction of Gleason score 8 to 10 cancers by dutasteride, as might
have been seen for finasteride in PCPT. Based on these
considerations, the FDA concluded that if 5-alpha reductase
inhibitors were used for prostate cancer risk reduction, one
might potentially anticipate one additional modified-score
Gleason score 8 to 10 cancer in the treated arm in order to
avert three or four lower-grade cancers. The FDA has stated
that health care professionals should be aware that 5-alpha
reductase inhibitors may increase the risk of prostate cancer
and that 5-alpha reductase inhibitors are not approved for
the prevention of prostate cancer.28
Effect of 5-Alpha Reductase Inhibitors on PSA
Multiple studies have shown that 5-alpha reductase inhibitors result in a statistically significant improvement in
the performance of PSA as a diagnostic test for prostate
cancer. These include the PLESS BPH trial,6 PCPT,7 and
the REDUCE trial.8,9 Moreover, when one looks at the
ability of PSA rise to detect clinically significant cancers, use
of the National Comprehensive Cancer Networkrecommended PSA velocity guidelines in REDUCE would have
resulted in the detection of 64% of Gleason score 7 to 10
cancers in the placebo arm, whereas 75% of Gleason score 7
to 10 cancers in the dutasteride arm exhibited a PSA rise off
nadir. These data highlight the importance of monitoring
PSA in men on 5-alpha reductase inhibitors. Most patients
experience about a 50% reduction in PSA by 6 to 12 months;
any rise from nadir should warrant a biopsy.
Conclusion
In summary, PSA-based screening is associated with

considerable overdiagnosis of prostate cancer, which, in
turn, results in a significant burden of overtreatment. Fivealpha reductase inhibitors decrease the detection of low-risk
prostate cancers during PSA screening, while at the same
time improving the performance of PSA as a diagnostic test
for aggressive prostate cancer. Five-alpha reductase inhibitors may, therefore, play an important role in focusing
prostate cancer screening efforts on the more lethal forms of
prostate cancer. Nevertheless, there remains an element of
uncertainty regarding a potential role for 5-alpha reductase
inhibitors used in this setting in promoting aggressive
prostate cancer.
PROSTATE CANCER RISK REDUCTION
Author
Youssef S. Tanagho*
Gerald L. Andriole
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Amarex; Amgen;
Augmenix;
Bayer; BristolMyers Squibb;
Cambridge Endo;
Caris MPI;
GlaxoSmithKline;
Janssen
Biotech; Myriad
Genetics; OrthoClinical
Diagnostics;
STEBA Biotech;
Viking Medical
Envisioneering
Honoraria
Amgen; Bayer;
Bristol-Myers
Squibb; Caris;
GlaxoSmithKline;
Janssen
Biotech; Myriad
Genetics; OrthoClinical
Diagnostics;
STEBA Biotech
Research
Funding
Expert
Testimony
Other
Remuneration
Caris;
GlaxoSmithKline;
STEBA Biotech
REFERENCES
1. Andriole GL, Crawford ED, Grubb RL 3rd, et al. Prostate cancer
screening in the randomized Prostate, Lung, Colorectal, and Ovarian Cancer
Screening Trial: mortality results after 13 years of follow-up. J Natl Cancer
Inst. 2012;104:125-132.
2. Schroder FH, Hugosson J, Roobol MJ, et al. Screening and prostatecancer mortality in a randomized European study. N Engl J Med. 2009;360:
1320-1328.
3. Hugosson J, Carlsson S, Aus G, et al. Mortality results from the
Goteborg randomised population-based prostate-cancer screening trial. Lancet Oncol. 2010;11:725-732.
4. Andriole GL, Crawford ED, Grubb RL 3rd, et al. Mortality results from
a randomized prostate-cancer screening trial. N Engl J Med. 2009;360:13101319.
5. Cooperberg MR, Broering JM, and Carroll PR. Time trends and local
variation in primary treatment of localized prostate cancer. J Clin Oncol.
2010;28:1117-1123.
6. Andriole GL, Guess HA, Epstein JI, et al. Treatment with finasteride
preserves usefulness of prostate-specific antigen in the detection of prostate
cancer: results of a randomized, double-blind, placebo-controlled clinical trial.
PLESS Study Group. Proscar Long-term Efficacy and Safety Study. Urology.
1998;52:195-202.
7. Thompson IM, Chi C, Ankerst DP, et al. Effect of finasteride on the
sensitivity of PSA for detecting prostate cancer. J Natl Cancer Inst. 2006;98:
1128-1133.
8. Andriole GL, Bostwick D, Brawley OW, et al. The effect of dutasteride on
the usefulness of prostate specific antigen for the diagnosis of high grade and
clinically relevant prostate cancer in men with a previous negative biopsy:
results from the REDUCE study. J Urol. 2011;185:126-131.
9. Marberger M, Freedland SJ, Andriole GL, et al. Usefulness of prostatespecific antigen (PSA) rise as a marker of prostate cancer in men treated with
dutasteride: Lessons from the REDUCE study. BJU Int. Epub 2011 Jun 23.
10. Welch HG, Black WC. Overdiagnosis in cancer. J Natl Cancer Inst.
2010;102:605-613.
11. Sakr WA, Grignon DJ, Haas BP, et al. Age and racial distribution of
prostatic intraepithelial neoplasia. Eur Urol. 1996;30:138-144.
12. Powell IJ, Bock CH, Ruterbusch JJ, et al. Evidence supports a faster
growth rate and/or earlier transformation to clinically significant prostate
cancer in black than in white American men, and influences racial progression and mortality disparity. J Urol. 2010;183:1792-1796.
13. Drazer MW, Huo D, Schonberg MA, et al. Population-based patterns
and predictors of prostate-specific antigen screening among older men in the
United States. J Clin Oncol. 2011;29:1736-1743.
14. Welch HG, Schwartz LM, and Woloshin S. Prostate-specific antigen
levels in the United States: implications of various definitions for abnormal.
15. Albertsen PC, Moore DF, Shih W, et al. Impact of comorbidity on
survival among men with localized prostate cancer. J Clin Oncol. 2011;29:
1335-1341.
16. Wilt TJ. The VA/NCI/AHRQ CSP#407: Prostate cancer Intervention
Versus Observation Trial (PIVOT): main results from a randomized trial
comparing radical prostatectomy to watchful waiting in men with clinically

localized prostate cancer. In: American Urological Association 2011 Annual
Meeting; 2011 May 17.
17. Pinsky PF, Blacka A, Kramer BS, et al. Assessing contamination and
compliance in the prostate component of the Prostate, Lung, Colorectal and
Ovarian (PLCO) Cancer Screening. Clin Trials. 2010;4:303-311.
18. Carroll PR, Whitson JM, Cooperberg MR. Serum prostate-specific
antigen for the early detection of prostate cancer: always, never, or only
sometimes? J Clin Oncol. 2011;29:345-347.
19. Fall K, Fang F, Mucci LA, et al. Immediate risk for cardiovascular
events and suicide following a prostate cancer diagnosis: prospective cohort
study. PLoS Med. 2009;6:1-8.
20. Lippman SM, Klein EA, Goodman PJ, et al. Effect of selenium and
vitamin E on risk of prostate cancer and other cancers: the Selenium and
Vitamin E Cancer Prevention Trial (SELECT). JAMA. 2009;301:39-51. Epub
2008 Dec 9.
21. Klein EA, Thompson IM, Tangen CM, et al. Vitamin E and the risk of
prostate cancer: the Selenium and Vitamin E Cancer Prevention Trial
(SELECT). JAMA. 2011;306:1549-1556.
22. Thompson IM, Goodman PJ, Tangen CM, et al. The influence of
finasteride on the development of prostate cancer. N Engl J Med. 2003;349:
215-224.
23. Cohen YC, Liu KS, Heyden NL, et al. Detection bias due to the effect of
finasteride on prostate volume: a modeling approach for analysis of the
Prostate Cancer Prevention Trial. J Natl Cancer Inst. 2007;99:1366-1374.
24. Pinsky PF, Parnes H, and Ford L. Estimating rates of true high-grade
disease in the Prostate Cancer Prevention Trial. Cancer Prev Res (Phila).
2008;1:182-186.
25. Redman MW, Tangen CM, Goodman PJ, et al. Finasteride does not
increase the risk of high-grade prostate cancer: a bias-adjusted modeling
approach. Cancer Prev Res (Phila). 2008;1:174-181.
26. Kaplan SA, Roehrborn CG, Meehan AG, et al. PCPT: Evidence that
finasteride reduces risk of most frequently detected intermediate- and highgrade (Gleason score 6 and 7) cancer. Urology. 2009;73:935-939.
27. Kramer BS, Hagerty KL, Justman S, et al. Use of 5- reductase
inhibitors for prostate cancer chemoprevention: American Society of Clinical
Oncology/American Urological Association 2008 Clinical Practice Guideline.
J Clin Oncol. 2009;27:1502-1516.
28. Theoret MR, Ning YM, Zhang JJ, et al. The risks and benefits of 5-
reductase inhibitors for prostate cancer prevention. N Engl J Med. 2011;365:
97-99.
29. Andriole GL, Bostwick DG, Brawley OW, et al. Effect of dutasteride on
the risk of prostate cancer. N Engl J Med. 2010;352:1192-1202.
30. Roehrborn CG, Andriole GL, Wilson TH, et al. Effect of dutasteride on
prostate biopsy rates and the diagnosis of prostate cancer in men with lower
urinary tract symptoms and enlarged prostates in the combination of Avodart
and Tamsulosin trial. Eur Urol. 2011;59:244-249.
31. Fleshner NE, Lucia MS, Egerdie B, et al. Dutasteride in localised
prostate cancer management: the REDEEM randomised, double-blind,
placebo-controlled trial. Lancet. Epub 2012 Jan 23.
95
Screening for Prostate Cancer with ProstateS p e c i fi c A n t i g e n : W h a t s t h e E v i d e n c e ?

By Pamela M. Marcus, PhD, and Barnett S. Kramer, MD, MPH
Overview: In October 2011, the U.S. Preventive Services Task

Force (USPSTF, or Task Force) released draft recommendations on prostate cancer screening with prostate-specific
antigen (PSA), concluding that PSA-based screening results
in small or no reduction in prostate cancerspecific mortality
and is associated with harms related to subsequent evaluation
and treatments, some of which may be unnecessary. This
statement was accompanied by a grade D recommendation,
which indicates that in the Task Forces judgment there is
moderate or high certainty that the service has no net benefit
or that the harms outweigh the benefits. The Task Force, an
independent panel of nonfederal (U.S.) experts in prevention
and evidence-based medicine, conducts systematic evidence
reviews of preventive health care services and makes recommendations about preventive services in primary care. Task
Force recommendations do not set U.S. federal policy but can
HE USPSTF IS an independent panel of nonfederal

(U.S.) primary care providers and scientists with expertise in prevention and evidence-based medicine. The Task
Force conducts systematic evidence reviews of preventive
health care services and makes recommendations about
preventive services in primary care. Task Force recommendations do not set U.S. federal policy but can and do
influence clinical practice as well as health care coverage
and reimbursement. Therefore, they are important and
often controversial, particularly when the recommendation
questions a common medical intervention. Under the Task
Forces purview is cancer screening, a prevention strategy
that has, over the years, proven to be contentious and
emotionally charged.
In October 2011, the Task Force released a draft of
recommendations on prostate cancer screening with PSA.1
They concluded that PSA-based screening results in small
or no reduction in prostate cancerspecific mortality and is
associated with harms related to subsequent evaluation and
treatments, some of which may be unnecessary. This statement was accompanied by a grade D recommendation,
which indicates that in the Task Forces judgment there is
moderate or high certainty that the service has no net
benefit or that the harms outweigh the benefits. As of this
writing, the recommendations are still in draft form, and the
Task Force is assessing a large number of comments sent to
them during the public comment period. Whatever the
ultimate conclusions of the Task Force, prostate cancer
screening with PSA will remain a Medicare benefit for men
of all ages and risk factor profiles.
Screening for prostate cancer with PSA is an example of
mass screening. A mass cancer screening program strives to
screen asymptomatic persons with certain characteristics
(usually based on age) and reduce the rate of disease-specific
mortality in the population being screened. For a mass
cancer screening program to be of value, mortality from the
target cancer must be reduced, but harms associated with
the screening process and downstream events triggered by
screening cannot outweigh the benefit. Persons who are
screened have no symptoms of the disease in question and
therefore are healthy with regard to that illness. It is
96
and do influence reimbursement and clinical practice. In this

article, we will present evidence the Task Force considered
when making its decision, including two highly influential
randomized controlled trials (RCTs) of prostate cancer
screening, the European Randomized Study of Prostate Cancer (ERSPC) and the Prostate, Lung, Colorectal and Ovarian
Cancer Screening Trial (PLCO). The two trials arrived at
different conclusions about the efficacy of routine prostate
cancer screening, but similar conclusions about the accompaniment of clinically relevant harms with prostate cancer
screening, including overdiagnosis (screen detection of cancers that never would be diagnosed in the absence of screening). We also will present other available evidence on benefits
and harms of PSA-based screening and consider that evidence
and the findings of ERSPC and PLCO in conjunction with one
another.
difficult to make healthy individuals any healthier, especially with an intervention like cancer screening, which puts
large numbers of people in harms way. Therefore, there
should be strong evidence of benefit from cancer screening
before mass screening programs are implemented.
In this article, we discuss the evidence that led to the Task
Forces decision to recommend against routine prostate
cancer screening with PSA. We first present prostate cancer
statistics for the United States. Before presenting the evidence that influenced the Task Forces recommendation, we
provide an overview of cancer screening, as such knowledge
is necessary for proper evaluation of evidence. For a thorough treatment of the topic, please see Prorok and colleagues.2
Prostate Cancer in the United States
Prostate cancer is the most frequently diagnosed nonskin

cancer and the second-leading cause of cancer death in U.S.
men. More than 240,000 cases are expected to be diagnosed
in 2012 and more than 28,000 men are expected to die of the
disease.3 Early prostate cancer is treatable but usually has
no symptoms; advanced prostate cancer, on the other hand,
is symptomatic but not curable. More than 90% of prostate
cancers in the United States are detected at a treatable
stage, however, and the 5-year relative survival for those
cancers approaches 100%. Because treatments often lead to
urinary and erectile dysfunction, some clinicians have adopted a watchful waiting or active surveillance strategy
rather than an initial surgical one for older patients and
those who appear to have less aggressive tumors.
Historically, prostate cancer incidence rates began to rise
rapidly in the mid- to late 1980s, continued to do so until the
From the Division of Cancer Control and Population Sciences, National Cancer Institute,
Bethesda, MD.
Address reprint requests to Pamela M. Marcus, PhD, Division of Cancer Control and
Population Sciences, National Cancer Institute, 6130 Executive Blvd., Room 4106,
Bethesda, MD 20892-7344; email: marcusp@mail.nih.gov.
1092-9118/10/1-10
PSA SCREENING: HARMS WITHOUT CLEAR BENEFIT
mid-1990s, and then started to fall. This trend is thought by

many to be primarily attributable to changes in use of PSA
screening rather than radical changes in risk factors or
classification schemes for invasiveness of disease. In the
early 1990s, at the height of PSA screening, the age-adjusted
prostate cancer incidence rate was in excess of 200 per
100,000 man-years, but in 2008 (the most recent data
available), it was about 150 per 100,000 man-years.4 Prostate cancer mortality peaked in 1993 at about 39 per 100,000
man-years but in 2008 had fallen to about 23 per 100,000
man-years.4
Cancer Screening
Cancer screening refers to the routine testing of persons

who are asymptomatic for cancer of a particular organ. The
goal of screening is to decrease the risk of death from a given
cancer by detecting the cancer at a time when treatment can
lead to cure. A reduction in disease-specific mortality with
screening, relative to established care without that exam,
KEY POINTS
A 2010 meta-analysis of prostate-specific antigens

(PSAs) effect on prostate cancer mortality that included five randomized controlled trials (including
the two largest, the European Randomized Study of
Prostate Cancer and the Prostate, Lung, Colorectal
and Ovarian Cancer Screening Trial) generated a
summary risk ratio of 0.88 (95% CI: 0.711.09),
indicating no statistically significant benefit of PSA
screening.
If PSA screening does reduce prostate cancer mortality, the reduction is likely to be small.
The harms associated with prostate cancer screening
(false-positive exams), diagnostic evaluation of positive exams (e.g., hematospermia, hematuria, sepsis),
treatment of prostate cancer (e.g., urinary incontinence and impotence), and overdiagnosis (screen detection of cancers that never would be diagnosed in
the absence of screening) are well-established and clinically relevant, and should be considered in concert with
any benefit of PSA screening that might exist.
In October 2011, the U.S. Preventive Services Task
Force (USPSTF) released a draft statement advising
against prostate cancer screening with PSA and assigned a grade D recommendation, which states
that in the Task Forces judgment there is moderate
or high certainty that the service has no net benefit or
that the harms outweigh the benefits. Nevertheless,
prostate cancer screening with PSA will remain a
Medicare benefit for men of all ages and risk profiles.
The USPSTF recommendation reflects two core concepts in the field of cancer prevention and screening:
1) it is difficult to make healthy persons even healthier; and 2) strong evidence of benefit should exist
before routinely administering a clinical intervention
to healthy persons if that measure is likely to incur
harm.
Table 1. The Potential Positive and Negative Consequences

of Cancer Screening
Positive
Reduced risk of death from the target cancer
Provides some reassurance to those who are concerned that they may have
cancer
Negative
False reassurance if cancer is present, which may lead to disregard of symptoms
Harms of the screening test (e.g., perforation from endoscopy)
False-positive exams, including the need to undergo diagnostic evaluation and
experience any harms that accompany it
Earlier detection of cancer that does not lead to a change in outcome
Identification of overdiagnosed* cancers, with accompanying unnecessary,
potentially harmful diagnostic evaluation and treatment
* Cancers that never would have been diagnosed in the absence of screening.
indicates that screening works. RCTs are used to determine

whether screening reduces cancer mortality, as that study
design provides the most definitive evidence. Other study
designs, such as case-control studies or comparisons of
temporal trends in screening and mortality, are subject to
powerful confounding because of the inability to control for
factors that are related to both screening and death due to
the disease of interest. Metrics such as an increase in 5-year
survival, number of cases detected, and number of earlystage cases are insufficient on their own to demonstrate a
benefit of screening as a result of the presence of certain
methodologic biases.
There are both positive and negative consequences of
cancer screening (Table 1). Screening may benefit patients
by reducing the risk of death from the target cancer, relative
to what it would be in the absence of that screening exam. A
negative screening exam can provide some reassurance to
those who are concerned that they may have cancer, although occasionally the reassurance is false and may lead to
disregard of cancer symptoms. Screening, however, can lead
to substantial harm. Medical misadventures may occur as
part of the screening exam itself. Many positive screening
exams will be false positives, which lead to unnecessary
anxiety as well as diagnostic evaluation that may be accompanied by harm. Screening can result in earlier detection of
a cancer that does not lead to a change in date of death; in
this instance, the patient spends more of his or her life as a
patient with cancer, but life expectancy and cause of death
are unchanged. Screening also can result in the detection of
overdiagnosed cancers, cancers that never would have been
diagnosed in the absence of screening. Patients with overdiagnosed cancers receive unnecessary treatment, experience
unnecessary treatment-related morbidity, and can even die
prematurely as a result of unnecessary treatment (e.g.,
lethal cancers induced by radiotherapy, postoperative death,
and heart disease or leukemia caused by chemotherapy).
The U.K. National Screening Committee has compiled a
list of criteria for appraising the viability, effectiveness, and
appropriateness of a screening program5 (www.screening.
nhs.uk/criteria/fileid9287). We have identified and adapted
the most relevant criteria to prostate cancer screening with
PSA, and the 13 we consider most relevant are listed in
Table 2. Among those criteria are at least five that either are
known not to be true for PSA screening or whose veracity is
uncertain: the natural history of the detected lesions is not
fully understood, PSA screening appears not to preferentially detect lesions that are most likely to progress, there is
97
MARCUS AND KRAMER

Table 2. Criteria for Implementation of a Screening Program
Applied to Prostate Cancer Screening*
Criteria for Screening
An important health problem
Recognizable latent or early asymptomatic stage
Available suitable test with agreed-upon cutoff values
Test acceptable to the population
Natural history of the detected lesions understood
Preferential detection of lesions likely to progress
Available, accepted treatment
Available facilities for diagnosis and treatment
Agreed-upon policy on whom to treat
Strong evidence (RCTs) of mortality/morbidity reduction
Benefits proven to outweigh harms
Quality assurance standards in place (including monitoring)
Affordable (cost-effectiveness)
* Adapted from www.screening.nhs.uk/criteria/fileid9287.
no standard clinical policy on whom to treat, strong evidence

(RCTs) of mortality/morbidity reduction does not exist, and
benefits have not been proven to outweigh harms.
Evaluation of PSA Screening for Prostate Cancer:
Prostate Cancer Mortality
Five RCTs of PSA prostate cancer screening (either alone

or in conjunction with another modality) have reported
prostate cancer mortality results.6 A 2010 meta-analysis of
those studies generated a summary risk ratio of 0.88 (95%
CI: 0.711.09), indicating no statistically significant benefit
of PSA screening.6 The meta-analysis included ERSPC7 and
PLCO,8 the two largest and most influential RCTs of prostate cancer screening. ERSPC, which randomly selected in
excess of 180,000 men ages 50 to 74 from 1991 to 2003,
reported a 20% prostate cancer relative mortality reduction
(95% CI: 0.67 0.98) with screening using the core group of
men ages 55 to 69 (more than 160,000), although the
prostate cancer mortality reduction was not statistically
significant in the entire study population (rate ratio: 0.85,
95% CI: 0.731.00). PLCO, which randomly selected almost
77,000 men ages 55 to 74 from 1993 to 2001, reported a
nonsignificant 9% (95% CI: 0.87 to 1.36) increase in prostate
mortality with screening after 13 years of follow-up. Both
trials, however, demonstrated clinically important levels of
harm associated with diagnostic evaluation of positive
screens and treatment of screen-detected prostate cancers.
Neither ERSPC nor PLCO was perfect. As to be expected,
researchers with an a priori belief that prostate cancer
screening is warranted believe that ERSPCs shortcomings
are not fatal, but those who take a more negative view of
prostate cancer screening feel that PLCO is the stronger of
the two trials. Those who make their decisions based on
systematic evidence reviews, such as the Task Force, look at
the evidence base as a whole and conclude that although a
small benefit of PSA screening may be possible, the harms
that can follow may outweigh that potentially small benefit.
ERPSC was a multicenter trial: Finland, Switzerland,
Italy, the Netherlands, Belgium, Switzerland, Spain, Portugal, and France participated, although data from the latter
two countries were excluded from the initially published
analyses. Given the unique characteristics of health care in
each country, intervention arm participants had different
experiences and access to different diagnostic evaluation
procedures and treatments. The screening interval varied by
98
country (2 to 7 years), as did the use of additional modalities

and the PSA value that defined a positive screen. In the
Netherlands, Belgium, Switzerland, and Spain, randomization occurred after informed consent was received, but in
Finland, Switzerland, and Italy, potential subjects were
identified using population registries and randomly selected, and then those selected for screening were asked to
provide written informed consent, thus raising the question
of whether two equivalent arms were produced. A substantial number of men randomly selected for the control arm
were unaware of their participation in the trial, so they were
less likely to receive cutting-edge treatment, in particular,
radical prostatectomy, because they had no relationship
with the screening centers, which were academic centers. It
has been demonstrated that participants in the screened
arm were more likely to receive therapy with curative
intent. Therefore, it is unclear whether screening led to a
reduction in mortality or whether differences in the treatment were responsible.
PLCOs randomization was more likely than ERSPCs to
have produced equivalent screening arms, as all participants signed their informed consent forms before they were
randomly selected. Furthermore, all 10 PLCO centers used
the same criterion to define a positive PSA screen, and
control participants knew they were in the trial and had a
relationship with the specialty medical center where the
intervention arm participants were screened. PLCOs
Achilles heel was a high rate of PSA screening in the
control arm. The trial had been designed to have 90%
statistical power to detect a 20% reduction in prostate
cancer mortality in the presence of no more than 20%
noncompliance (i.e., screening) in the control arm, but actual
noncompliance, measured annually, was between 40% and
54% at particular screening years. Given that rate of noncompliance, PLCOs statistical power was 90% for a mortality reduction no smaller than 40%, which may be unrealistic
for PSA screening.
Evaluation of PSA Screening for Prostate Cancer:
Harms
Although ERSPC and PLCO came to different conclusions

about whether PSA screening could reduce prostate cancer
mortality, they both did come to the conclusion that screening could lead to nontrivial levels of adverse events due to
diagnostic evaluation and treatment. They also both indicate
that overdiagnosis occurs with prostate cancer screening.
PSA screening itself, which requires a blood draw, has
infrequent side effects, and the ones that do resultinfection at the draw site, vasovagal reaction, discomfortare
considered to be minor. However, adverse events associated
with prostatic biopsy are of more concern. In PLCO, almost
13% of men had at least one false-positive PSA exam and of
those, more than 5% had a biopsy dictated by that exam9; at
the Netherlands ERSPC site, 50% of men who had a biopsy
as a result of a positive exam had hematospermia, 23% had
hematuria, and 4% had fever.10 In addition, there are
reports of a recent and consistent increase in urosepsis with
resistant organisms for men undergoing PSA-motivated
prostatic biopsy. Surveillance, Epidemiology, and End Results (SEER)-Medicare data indicate that the rate of hospitalization for biopsy-related infection within 30 days of that
biopsy has quadrupled, from about 0.25% before 2001 to
more than 1% in 2007.11 Hospitalizations also occur for
PSA SCREENING: HARMS WITHOUT CLEAR BENEFIT
Fig. 1. Prostate cancer in the United States, 1975 to

2008: new cases and deaths.
noninfectious biopsy-related complications, but not to the

same degree. A similar pattern for biopsy-related hospitalizations was observed in Ontario, Canada, during the 2000s.13
The complications associated with prostate cancer treatment are unpleasant and life-altering. Urinary incontinence
and erectile dysfunction occur as a result of treatment,
although the percentage of men whose morbidities are
attributable solely to treatment likely varies by patient
characteristics, treatment received, the definition of incontinence and erectile dysfunction employed, whether the
diagnosis is self-assigned, and the prevalence of these morbidities in the study population before screening, and thus is
difficult to pinpoint. Nevertheless, those receiving prostatectomy have been shown in research settings to have a 2- to
11-fold increase in risk of urinary incontinence and a 1.2- to
2.1-fold increase in risk of erectile dysfunction, relative to
men who are treated with watchful waiting.1 In the Task
Forces 2011 review,1 the smallest and largest absolute
percentile differences for urinary incontinence following
prostatectomy compared with watchful waiting were nine
(19% vs. 10%) and 40 (44% vs. 4%). For erectile dysfunction,
the figures were 21 (89% vs. 68%) and 36 (81% vs. 45%). In
a community-based (U.S.) study of experiences of men diagnosed with prostate cancer in 1994 and 1995 who had
radical prostatectomy, 87% of men reported that they had
total urinary control prior to their surgery as compared with
less than 40% at 6, 12, 24, and 60 months postsurgery.14
Eighty-one percent reported that they had erections firm
enough for intercourse before their surgery, as compared
with less than 30% at the same postsurgery time points.14
Radiation therapy may confer a lower relative risk of incontinence and erectile dysfunction, but androgen deprivation
therapy may increase the risk of the latter.15 Postoperative
deaths or cardiovascular events occur about 0.5% of the
time, although rates are age-dependent.
Men with overdiagnosed prostate cancers are harmed the
most by PSA screening. In the case of overdiagnosis, any
expense, inconvenience, discomfort, sexual dysfunction,
morbidity, hospitalization, or death that results from diagnostic evaluation for a positive screen and treatment of a
diagnosed cancer is entirely unnecessary. Figure 1 provides
convincing evidence of overdiagnosis: the number of prostate
cancer cases in the United States was substantially higher
in 2008 than it was in the early 1970s, but the prostate
cancer mortality rates between the early 1970s and 2008

were not very different. This increase in risk correlates with
two changes in medical practice: an increase in transurethral resection of the prostate (which could serendipitously
detect prostate cancers when performed for other reasons)
beginning in the 1970s and the adoption of PSA in the late
1980s. A true epidemic of prostate cancer cannot explain
this rise in incidence because the rise was unaccompanied by
extreme shifts in risk factor prevalence or the identification
of a new, widespread exposure that strongly increased risk;
even if it had, not all cancers would have been curable.
Additional data supporting the existence of overdiagnosis
come from a comparison of prostate cancer incidence and
mortality in Connecticut and the Seattle-Puget Sound
area,16 two areas that differed with regard to the uptake of
PSA screening in the 1980s and 1990s. Connecticut, where
PSA screening was less common, had consistently lower
prostate cancer incidence rates from 1987 through 2001, but
prostate cancer mortality rates were very similar, with
Seattle data showing a possible, although not statistically
significant, increase in prostate cancer mortality for men
ages 75 to 79. A nearly identical reduction in the prostate
cancer mortality rate in both regions over time has been
observed; if PSA screening was reducing mortality in addition to reductions due to treatment, the reduction for Seattle
would have been even more pronounced.
Microsimulation methods have been used to estimate the
percentage of prostate cancers that are considered to be
overdiagnosed disease. Using ERSPC data, Draisma and
colleagues estimated that among men screened annually
from age 55 to 67, 50% of screen-detected cases would be
overdiagnosed cases.17 Using SEER data, overdiagnosis
ranged from 23% to 42% of all screen-detected cases.18
Modeling using PLCO data is not yet available, but an
excess of cases existed in the intervention arm in every
study, including after 13 years of follow-up.8 Two years after
the last screen (i.e., the end of study year 7), the excess of
479 cases indicates that overdiagnosis could have accounted
for about 17% of cases in the screened arm. That figure is an
underestimate, as contamination occurred in the control
arm. Although it is unclear whether the extent of overdiagnosis is moderate or extreme, it is clear that it exists with
PSA screening.
99
MARCUS AND KRAMER

Conclusion
One could argue that neither ERSPC nor PLCO findings

can be considered definitive, as the shortcomings of those
trials provide an unreliable measure of PSAs actual benefit.
In fact, the Physician Data Query Prevention and Screening
Editorial Board concluded that the evidence for PSA screening is insufficient to determine whether that practice results
in a mortality reduction. But board members concluded as
well that based on solid evidence, screening with PSA
and/or digital rectal exam detects some prostate cancers
that would never have caused important clinical problems
and leads to some degree of overtreatment, and that based
on solid evidence, current prostate cancer treatments, in-
cluding radical prostatectomy and radiation therapy, result

in permanent side effects in many men.19
The harms associated with PSA screening for prostate
cancer are well-established; the benefit of screening, however, is not and therefore remains theoretical. Given that a
central principle of medicine is first, do no harm, the
medical community must recognize that routine PSA screening may run counter to that principle. Rather than simply
advocating the practice both in the office and in the public
arena, health care providers must engage in informed decision making with men, as the U.S. Centers for Disease
Control and Prevention currently supports,20 to make sure
they are aware of the potential benefits, harms, and remaining uncertainties.
Author
Pamela M. Marcus*
Barnett S. Kramer*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Chou R, Croswell JM, Dana T, et al. Screening for prostate cancer: a
review of the evidence for the U.S. Preventive Services Task Force. Ann Intern
Med. 2011;155:762-771.
2. Prorok PC, Kramer BS, Gohagan JK. Screening theory and study design:
the basics. In Kramer BS, Gohagan JK, Prorok PC (eds). Cancer Screening:
Theory and Practice. New York: Marcel Dekker, 1999;29-53.
3. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer
J Clin. 2012;62:10-29. Epub 2012 Jan 4.
4. Howlader N, Noone AM, Krapcho M, et al. (eds). SEER Cancer Statistics
Review 1975-2008. http://seer.cancer.gov/csr/1975_2008/. Accessed January
30, 2012.
5. UK National Screening Committee. Programme Appraisal Criteria.
http://www.screening.nhs.uk/criteria/fileid9287. Accessed January 30, 2012.
6. Djulbegovic M, Beyth RJ, Neuberger MM, et al. Screening for prostate
cancer: systematic review and meta-analysis of randomised controlled trials.
BMJ. 2010 Sep 14;341:c4543.
7. Schroder FH, Hugosson J, Roobol MJ, et al. Screening and prostatecancer mortality in a randomized European study. N Engl J Med. 2009;360:
1320-1328.
8. Andriole GL, Crawford ED, Grubb RL 3rd, et al. Prostate cancer
screening in the randomized Prostate, Lung, Colorectal, and Ovarian Cancer
Screening Trial: Mortality results after 13 years of follow-up. J Natl Cancer
Inst. 2012;104:125-132.
9. Croswell JM, Kramer BS, Kreimer AR, et al. Cumulative incidence of
false-positive results in repeated, multimodal cancer screening. Am Fam Med.
2009;7:212-222.
10. Ilic D, OConnor D, Green S, et al. Screening for prostate cancer.
Cochrane Database Syst Rev. 2006 Jul 19;3:CD004720.
11. Loeb S, Carter HB, Berndt SI, et al. Complications after prostate biopsy:
data from SEER-Medicare. J Urol. 2011;186:1830-1834. Epub 2011 Sep 23.
100
12. Welch HG. Should I be tested for cancer? Maybe not and heres why.
Berkeley and Los Angeles, CA, University of California Press, 2004.
13. Nam RK, Saskin R, Lee Y, et al. Increasing hospital admission rates for
urological complications after transrectal ultrasound guided prostate biopsy.
J Urol. 2010;183:963-969.
14. Penson DF, McLerran D, Feng Z, et al. 5-year urinary and sexual
outcomes after radical prostatectomy: results from the Prostate Cancer
Outcomes Study. J Urol. 2005;173:1701-1705.
15. Potosky AL, Davis WW, Hoffman RM, et al. Five-year outcomes after
prostatectomy or radiotherapy for prostate cancer: the Prostate Cancer
Outcomes Study. J Natl Cancer Inst. 2004;96:1358-1367.
16. Lu-Yao G, Albertsen PC, Stanford JL, Stukel TA, Walter-Corkery E,
Barry MJ. Screening, treatment, and prostate cancer mortality in the Seattle
area and Connecticut: fifteen-year follow-up. J Gen Intern Med. 2008;23:18091814.
17. Draisma G, Boer R, Otto SJ, et al. Lead times and overdetection due to
prostate-specific antigen screening: estimates from the European Randomized Study of Screening for Prostate Cancer. J Natl Cancer Inst. 2003;95:868878.
18. Draisma G, Etzioni R, Tsodikov A, et al. Lead time and overdiagnosis in
Prostate-Specific Antigen screening: importance of methods and context.
19. Department of Health and Human Services, National Institutes of
Health, National Cancer Institute. Prostate cancer screening. http://www.
cancer.gov/cancertopics/pdq/screening/prostate/HealthProfessional. Accessed
March 5, 2012.
20. Department of Health and Human Services, Centers for Disease
Control and Prevention. Prostate cancer screening: a decision guide. http://
www.cdc.gov/cancer/prostate/pdf/prosguide.pdf. Accessed March 5, 2012.
GLIOBLASTOMA: TAKING THE STANDARD OF CARE

TO THE CUTTING EDGE
CHAIR
Mark R. Gilbert, MD
Houston, TX
SPEAKERS
Daniel J. Brat, MD, PhD
Emory University
Atlanta, GA
Howard Colman, MD, PhD
University of Utah
Salt Lake City, UT
Glioblastoma: Biology, Genetics, and Behavior

By Daniel J. Brat, MD, PhD
Overview: Glioblastoma (GBM) is a highly malignant, rapidly

progressive astrocytoma that is distinguished pathologically
from lower-grade tumors by necrosis and microvascular hyperplasia. The global pattern of growth changes dramatically
with the development of GBM histology and is characterized
by hypoxia-driven peripheral expansion from a growing necrotic core. Necrotic foci present centrally in GBM and are
typically surrounded by pseudopalisading cellsa configuration that is relatively unique and long recognized as an
ominous prognostic feature. Theses pseudopalisades are severely hypoxic, overexpress hypoxia inducible factor-1 (HIF-1),
and secrete proangiogenic factors, such as vascular endothelial growth factor (VEGF) and interleukin 8 (IL-8). The microvascular hyperplasia that emerges in response promotes
peripheral tumor expansion. Recent evidence suggests that
pseudopalisades represent a wave of tumor cells actively
migrating away from central hypoxia that arises following a
LIOBLASTOMA (GBM; World Health Organization

[WHO] grade 4) is the highest grade astrocytoma and
has a dismal prognosis.1 Mean survival following the most
advanced treatment, including neurosurgery, radiotherapy,
and chemotherapy is only 60 to 70 weeks.2 When patients
receive only surgical resection, but are not treated with
adjuvant therapy, mean survival is a mere 14 weeks, underscoring the tremendous natural growth properties of these
tumors. Lower-grade infiltrative astrocytomas (i.e., WHO
grade 2 and 3 astrocytomas) are also ultimately fatal, but
have substantially slower growth rates and longer survivals
(3 to 8 years). Only once lower-grade tumors have progressed to GBM do they demonstrate accelerated growth
and rapid progression to death. Those GBMs that progress
from lower-grade gliomas are referred to as secondary
GBMs, whereas those GBMs that present without a lowergrade precursor are termed primary or de novo GBMs.
De novo tumors account for approximately 90% of all GBMs
and are both clinically and molecularly distinct from secondary GBMs, as will be discussed below.
Growth Patterns in Grade 23 Astrocytomas
Importantly, the biologic properties of GBM (grade 4)

are quite distinct from those of lower-grade astrocytomas
(grade 2 and 3) and suggest that it represents more than an
incremental step in malignancy. The unique neuroimaging
and pathologic features that emerge during the transition to
GBM provide the best insight into potential mechanisms
responsible for the enhanced growth. By magnetic resonance
imaging (MRI), grade 2 and 3 astrocytomas show hyperintense T2-weighted (or fluid attenuated inversion recovery
[FLAIR]) signal abnormalities, reflecting vasogenic edema
generated in response to diffuse infiltration by individual
tumor cells. Lower-grade tumors expand the involved brain,
but show mild or no contrast enhancement, suggesting an
intact blood-brain barrier and a lack of tumor necrosis.
Radial growth rates are modest, with annual increases in
diameter of 2 to 4 mm/yr.3 Histologic sections of grade 23
tumors reflect the imaging properties: neoplastic cells are
seen diffusely infiltrating between neuronal and glial processes, leading to architectural distortion and edema.4-5 As
102
vascular insult. Vaso-occlusive and prothrombotic mechanisms in GBM could readily explain the presence of pseudopalisading necrosis in tissue sections, the rapid peripheral
expansion on neuroimaging, and the dramatic shift to an
accelerated rate of clinical progression as a result of hypoxiainduced angiogenesis. The genetic alterations that coincide
with progression to GBM include amplification of epidermal
growth factor receptor (EGFR), deletion of CDKN2A, and
mutation or deletion of PTEN. Other diagnostic and prognostic
tests used in neuro-oncology include assessment of 1p/19q,
MGMT promoter methylation, IDH1, and p53. More recently,
the Cancer Genome Atlas data have indicated that there are
four robust transcriptional classes of GBM, referred to as
proneural, neural, classical, and mesenchymal. These classes
have genetic associations and may pave the road for future
development of targeted therapies.
astrocytomas progress from the lower end of grade 2 to the

upper end of grade 3, the degree of nuclear anaplasia
increases and the proliferative capacity creeps upward,
resulting in a more densely cellular tumor with greater
malignant potential. Thus, grade 23 astrocytomas represent a continuum of gradually increasing tumor grade and
biologic potential, with clinical behavior generally correlating with the properties of individual tumor cells.
Changes in Tumor Growth from Grade 23
Astrocytomas to Glioblastoma
Tumor dynamics are fundamentally altered in GBM in

comparison to grade 23 astrocytomas. Radial growth rates
for GBM can accelerate to values nearly 10 times greater
than those in grade 2 astrocytomas.3 MRI of GBM typically
reveals a central, contrast-enhancing component (rimenhancing mass) emerging from within the infiltrative
astrocytoma and rapidly expanding outward, causing a
much larger T2-weighted signal abnormality in the tumors
periphery.
Studies of cellular proliferation strongly suggest a fundamental difference in driving biologic forces between the
grade 23 astrocytomas and glioblastoma.6 The most reliable marker of proliferation is the Ki-67/MIB-1 antibody,
which identifies nuclei of cells in the G1, S, G2, and M
phases of the cell cycle, but is not expressed in the resting
phase, G0. In one study of proliferation in diffuse astrocytomas of grades 2, 3, and 4, Giannini and colleagues found that
MIB-1 index was highly correlated with survival on multivariate analysis among grade 2 and 3 astrocytomas. However, when GBMs (grade 4) were included in the analysis,
proliferation was no longer an independent marker of prog-
From the Department of Pathology and Laboratory Medicine, Winship Cancer Institute,
Emory University School of Medicine, Atlanta, GA.
Address reprint requests to Daniel J. Brat, MD, PhD, Department of Pathology and
Laboratory Medicine, Emory University Hospital, H-176, 1364 Clifton Rd. NE, Atlanta, GA
30322; email: dbrat@emory.edu
1092-9118/10/1-10
GLIOBLASTOMA: BIOLOGY, GENETICS AND BEHAVIOR
nosis.7 Similarly, the significance of MIB-1 prolifieration in

prognosticating grade 2 and 3 astrocytomas was also supported by an investigation by Hsu and colleagues, who
concluded that the MIB-1 index was useful in grade 2 and 3
tumors, but not GBM.8 When proliferation is studied in the
setting of the GBM histology alone, the prognostic significance is not evident. For example, in one study of 116 newly
diagnosed GBMs, the mean MIB-1 index was 12.5% and
varied from 0 to 76.4%. In this series, MIB-1 proliferation was
not associated with survival on either univariate or multivariate analysis.9 Thus, taking the evidence from these and other
studies together, proliferation is a biologically important,
prognostically significant driving force in grade 2 and 3 astrocytomas. However, once the disease has progressed to GBM,
biologic forces more significant than proliferation emerge.
Forces That Drive Gliobastoma Progression
What are these driving forces that emerge in GBM? The

histopathologic features that distinguish GBM from lower
grade astrocytomas are found near the contrast-enhancing
rim on MRI and include the following: 1) foci of necrosis,
usually with evidence of surrounding cellular pseudopalisades (pseudopalisading necrosis), and 2) microvascular
hyperplasia, a form of angiogenesis morphologically recognized as endothelial proliferation within newly sprouted
vessels.1 In contrast to lower-grade astrocytomas, these two
diagnostic findings of GBM are largely independent of tumor
cell properties (i.e., degree of anaplasia, proliferation, etc.),
yet carry an inordinate degree of prognostic power. Rather
than mere markers, these structures are more likely to be
mechanistically linked to the accelerated growth properties
that characterize the grade 3 to 4 transition.
An emerging model of tumor progression may explain the
development of pseudopalisades, the relationship between
pseudopalisades and angiogenesis, and the strong association between pseudopalisades and aggressive clinical behav-
KEY POINTS
The growth pattern of glioblastoma is fundamentally

distinct from lower-grade gliomas and represents
more than a small incremental change in tumor
grade.
Whereas tumor cell proliferation is prognosic in
lower-grade gliomas, hypoxia-driven mechanisms become more relevant in glioblastoma.
Vasco-occlusion and thrombosis are typical in glioblastoma and likely initiate or propagate perfusionlimited hypoxia and necrosis associated with tumor
progression.
Genetic tests that are performed for diagnostic and
prognostic purposes for glioblastoma include assessment of EGFR, PTEN, 1p/19q, MGMT promoter
methylation, IDH1, and p53.
Analysis of the Cancer Genome Atlas data has demonstrated four distinct transcriptional classes of
glioblastoma that have genomic correlates and prognostic significance: proneual, neural, classical, and
mesenchymal.
ior.10,11 This model hypothesizes a sequence that begins

with an infiltrating astocytoma of moderate to high cellularity (i.e., grade 3 astrocytoma) and continues with 1) vascular
occlusion within the tumor that is often associated with
intravascular thrombosis; 2) hypoxia in regions surrounding
vascular pathology; 3) outward migration of tumor cells
away from hypoxia, creating a peripherally moving wave
(pseudopalisade) and central necrosis; 4) secretion of
hypoxia-inducible, proangiogenic factors (VEGF, IL-8) by
pseudopalisading cells; 5) an exuberant angiogenic response
creating microvascular proliferation in regions adjacent to
central hypoxia; and 6) accelerated outward expansion of
tumor cells toward a new vasculature. The global growth
properties of GBM within the brain reflect a coalescence of
these microscopic processes and result in a peripherally
expanding tumor with a large degree of central necrosis.
Pseudopalisades Are Actively Migrating Tumor Cells
Pseudopalisading of cells around central degereneration

has been recognized for nearly a century as both a defining
feature of GBM and a morphologic finding that predicts
aggressive behavior. A commonly held belief has been that
pseudopalisades represent a rim of residual tumor cells
around a centrally degenerating clone of highly proliferative
cells. However, recent studies have refuted this and concluded that pseudopalisades represent a wave of actively
migrating tumor cells that are moving away from an area of
central hypoxia. Pseudopalisading cells are known to be
hypoxic, as demonstrated by their dramatic upregulation of
HIF-1, a nuclear transcription factor that orchestrates the
cells adaptive response to low oxygen.12 Gene expression
studies performed on microdissected pseudopalisading cells
have demonstrated upregulated gene transcripts in this
population that suggest a response to a hypoxic microenvironment, including those related to glycolysis, angiogenesis,
and cell cycle control.13 Hypoxic GBM cells in culture that
have similar upregulation of HIF-1 are more highly migratory than normoxic cells, and HIF-1 itself mediates many of
the critical promigratory mechanisms in gliomas and other
neoplastic cells.12 Moreover, hypoxic pseudopalisades express increased levels of extracellular matrix proteases
associated with invasion, including MMP-2 and uPAR.4,12
Thus, the combined evidence suggests that the pseudopalisades in GBM are formed by a population of hypoxic,
actively migrating neoplastic cells that have imposed themselves on a less mobile population, thereby creating a hypercellular zone around an evolving area of central necrosis.
Pseudopalisades Are Hypoxic Tumor Cells Migrating
Away from Vascular Pathology
The hypoxia that leads to increased tumor cell migration

to form pseudopalisades and to widespread invasiveness
could result from limitations in vascular perfusion within
the tumor (i.e., disruption of the blood supply) or from
reduced oxygen diffusion within the neoplasm, in part due to
increased metabolic demands of a growing tumor. Attenuated perfusion would lead to cell migration away from
central blood vessels that no longer provides the necessary
oxygen supply. Limitations in oxygen diffusion, on the other
hand, would cause tumor cells at greatest distance from
arterial supplies to become hypoxic and migrate toward
viable vessels. These mechanisms are not mutually exclu-
103
DANIEL J. BRAT
sive. However, a growing body of experimental and observational evidence favors the hypothesis that pseudopalisades
represent tumor cells migrating away from a dysfunctional
vasculature.12,14,15 Perhaps less appreciated, abnormal vessels can often be noted within the lumina of at least a subset
of pseudopalisades. A comprehensive survey of human GBM
specimens found that over 50% of pseudopalisades had
evidence of a central vascular lumen that was either degenerating or thrombosed.12 Thus, pseudopalisades around necrosis appear to represent hypoxic tumor cells migrating
away from vaso-occlusion and thrombosis.
Intravascular Thrombosis Accentuates and Propagates
Tumor Hypoxia
While precise initiators of vascular pathology in GBM

continue to be studied, it is becoming clear that intravascular thrombosis within these neoplasms can accentuate and
propagate tumoral hypoxia and necrosis. Intravascular
thrombosis within the tumoral tissue of GBM is a frequent
intraoperative finding by the neurosurgeon. Even more
impressively, thrombosed vessels within resected GBM
specimens can almost always be identified under the microscope (noted histologically in over 90% of GBMs).12,15 In
contrast, the frequency of microscopic thrombosis is much
lower in anaplastic astrocytoma (grade 3), a tumor that
lacks necrosis and angiogenesis. In the uncommon instance
when intravascular thrombosis is identified microscopically
in anaplastic astrocytomas (AAs, WHO grade 3), it is predictive of aggressive growth, similar to that of GBM, suggesting it may precede the necrosis and angiogenesis that
arise in GBM and are associated with a more rapid progression.15 Since the frequency of intravascular thrombosis
within neoplastic tissue is much higher in GBMs (grade 4)
than AAs (grade 3), critical pro-thrombotic events must
occur in this transition. In many instances, the intravascular thrombosis can be seen within or adjacent to the regions
of pseudopalisading necrosis, leading to the proposition that
vaso-occlusion due to thrombosis could directly initiate or
propagate hypoxia and necrosis in GBM.
Multiple factors likely contribute to intravascular thrombosis in GBM, including abnormal blood flow within a
distorted vasculature, increased interstitial edema, dysregulation of pro- and anticoagulant factors, and access of
plasma-clotting factors to tumoral tissue. Normal central
nervous system (CNS) blood vessels allow only limited
diffusion though their walls because of a highly restrictive
blood-brain barrier, which is formed primarily by endothelial tight junctions, but also has contributions from astrocytic foot plates, extracellular matrix, and endothelialpericytic interactions. This barrier becomes breached in
GBM and can be visualized radiologically by the presence of
contrast enhancement due to increased vascular permeability to contrast agents (e.g., gadolinium) and to proteins that
bind them, such as albumin. Damaged vessels appear fenestrated, show detachment of pericytes, and exhibit extracellular matrix alterations. All the factors that contribute to
increased permeability have not been defined, but VEGF
secretion by neoplastic cells is known to cause vascular
leakage.16 One result is to bring plasma coagulation factors
such as factor VII into the tissue spaces where they are
activated by binding tissue factor and result in thrombosis.
104
Angiogenesis Supports Peripheral Tumor Growth
If emerging models of GBM progression are valid, then

vascular pathology may underlie the development of hypoxia and necrosis in GBM. Although necrosis has long been
recognized as a marker of aggressive behavior in diffuse
gliomas, by itself it does not explain rapid tumor progression. Indeed, tumor cell death is the goal of most adjuvant
therapies. Instead, pseudopalisades that surround necrosis
in GBM are intimately related to microvascular hyperplasia,
a defining morphologic feature of GBM that is most often
noted in regions directly adjacent to pseudopalisades.11 This
exuberant angiogenic response attempts to lay down a new
vasculature for rapid neoplastic expansion, yet the proper
function of these distorted vessels has not been established.
The formation of new blood vessels from pre-existing ones is
tightly regulated process and follows a complex sequence in
response to pro- and antiangiogenic factors. Initial phases
require increased vascular permeability of parent vessels,
extravasation of plasma, and deposition of proangiogenic
matrix proteins. In response to the mitogenic effects of
proangiogenic cytokines, endothelial cells proliferate and
migrate along a chemotactic gradient into the extracellular
matrix. Once established, endothelial cells form tubes with a
central lumen, elaborate a basement membrane, and eventually recruit pericytes and smooth muscle cells to surround
the mature vessels.
One of the most critical proangiogenic factors produced by
pseudopalisades that is responsible for directing nearby
angiogenesis in GBM is VEGF. As noted above, pseudopalisading cells are severely hypoxic and express high levels of
hypoxia-inducible transcription factors, including HIF-1.
The VEGF gene contains a hypoxia-responsive element
(HRE) within its promoter that binds HIF-1, thereby activating transcription.17 VEGF concentrations in the cystic
fluid of human GBMs can reach levels that are 200 300-fold
higher than in serum. Inhibition of this HIF/VEGF pathway
suppresses tumor growth experimentally. Once expressed
and secreted, extracellular VEGF binds to its high affinity
receptors, VEGFR-1 and VEGFR-2, which are upregulated
on endothelial cells of high-grade gliomas, but not present in
normal brain. Receptor activation then leads to angiogenesis
in regions adjacent to pseudopalisades, eventually leading to
a vascular density in GBMs that is among the highest of all
human neoplasms.
A second proangiogenic factor that is highly upregulated
in GBMs is interleukin-8 (IL-8, CXCL8).18 Much like VEGF,
hypoxia/anoxia strongly stimulates IL-8 expression and its
expression is also found at highest levels within the pseudopalisades of GBM. Unlike VEGF, IL-8 has a more punctate distribution within pseudopalisades, and it remains
unclear if tumor cells or scattered infiltrating macrophages
are most responsible for the majority of its expression.
Hypoxic upregulation of IL-8 is not directly due to HIF
activation, but is more likely due to activation of AP-1 by
hypoxia/anoxia. The IL-8 receptors that could potentially
contribute to IL-8 mediated tumorigenic and angiogenic
responses in GBM include CXCR1 and CXCR2, both of
which are G-protein coupled.
The precise type of angiogenesis that is most evident
in GBM, microvascular hyperplasia, is characterized by
numerous enlarged, rapidly dividing endothelial cells, pericytes, and smooth muscle cells that form tufted microaggre-
gates at the leading edge of sprouting vessels. In its most

florid form, angiogenesis takes the shape of glomeruloid
bodiesa feature that is most characteristic of GBM, but is
also an independent marker of poor prognosis in other forms
of cancer.19 Since necrosis and hypoxia are located in the
GBMs core and near the contrast-enhancing rim, hypoxiainduced angiogenesis occurs further peripherally, favoring neoplastic growth outward. The permissive nature of the CNS
parenchymal matrix to diffuse infiltration by individual
glioma cells allows for this burst of peripheral expansion.4
Genetic Testing
A large and growing number of genetic alterations have

been identified in the diffuse gliomas, some of which are
used for diagnostic and prognostic purposes in neurooncology. The genetic alterations that coincide with progression to GBM include amplification of EGFR, deletion of
CDKN2A, and mutation or deletion of PTEN. Other diagnostic and prognostic tests used in neuro-oncology include
assessment of 1p/19q, MGMT promoter methylation, IDH1,
and p53.1
p53 Pathway Alteration
Alterations in the p53 tumor suppressor pathway are

frequent in infiltrative astrocytomas. The p53 pathway can
be altered mechanisms including mutation of TP53 and
deletion of the opposite allele, MDM2 gene amplification,
and p14ARF gene deletion. Point mutations in the DNAbinding region of TP53 are the most frequent source of
inactivation. The majority (50% to 60%) of lower-grade
(WHO grade 2) infiltrating astrocytomas show such TP53
mutations, suggesting it occurs early. GBMs that arise from
grade 2 and 3 astrocytomas, so-called secondary GBMs, have
similar frequencies of TP53 mutations. TP53 mutations are
less frequent (30%) in primary, or de novo, GBMs, yet the
p53 pathway is altered by other mechanisms in these
tumors. The prognostic significance of TP53 mutations and
p53 protein expression in astrocytomas has been debated.
Young age is a strong predictor of prolonged survival among
patients with astrocytomas, especially GBM. Since TP53
mutations occur more frequently in GBMs from young
patients, their significance must be separated from the
survival advantage of youth.
For glioma classification, TP53 mutations currently have
limited utility. Since p53 derived from mutant genes has a
longer cellular half-life than wild type, the protein accumulates in the nucleus. Thus, positive nuclear immunohistochemical staining for the p53 protein correlates, albeit
imperfectly, with the presence of TP53 mutations. Positive
p53 immunostaining can be occasionally helpful for distinguishing astrocytic from oligodendroglial differentiation,
since the oligodendrogliomas rarely harbor TP53 mutations.
1p/19q
There is a strong association of allelic losses on chromosomes 1p and 19q and the oligodendroglioma phenotype, and
60% to 80% of oligodendroglial neoplasms demonstrate combined 1p and 19q losses.1,6 Enthusiasm for defining genetic
subsets of oligodendrogliomas increased substantially with
the demonstration of prognostically distinct groups.20
1p and 19q losses are less frequent in other forms of
gliomas. For example, Smith and colleagues investigated the
allelic losses of 1p and 19q in 115 diffuse gliomas.21 Combined loss of 1p and 19q were seen in 11% of astrocytomas,
31% of the mixed oligoastrocytomas, and 64% of oligodendrogliomas. Thus, while most oligodendrogliomas showed
1p/19q loss, not all did. Moreover, a small percentage of
mixed gliomas and astrocytomas also showed similar deletions. More recent studies have demonstrated less frequent
1p/19q losses in diffuse astrocytomas.22
Similar studies have interrogated the prognostic significance of combined loss of 1p and 19q in diverse types of
diffuse gliomas and found them to be predictive of prolonged
overall survival only for patients with oligodendrogliomas.
Combined 1p and 19q losses are not predictive of prolonged
survival in astrocytomas or oligoastrocytomas of any
grade.22-23 However, diagnostic testing for 1p/19q is often
used in cases of diffuse gliomas that have ambiguous morphology, since 1p/19q codeletion is highly associated with
oligodedroglioma.
EGFR
Amplifications of the EGFR gene occur in approximately

40% of GBMs and 10% of anaplastic astrocytomas.1 Amplifications are much less frequent in low-grade astrocytomas
and are considered a late genetic event in the progression of
tumors to GBM. Either wild-type or mutated forms of EGFR
can be amplified. The most common EGFR amplification is a
mutated form lacking exons 27, which results in a truncated cell surface protein with constitutive tyrosine kinase
activity (EGFRvIII).
The significance of EGFR gene amplification or EGFR
protein overexpression as a prognostic marker in GBM has
been debated. Most comprehensive studies have concluded
that EGFR status is not prognostically significant in patients with GBM. However, therapies have been developed
that are directed at the overexpressed EGFR in GBMs.
Therefore, it may become critical to establish the EGFR
status of GBM as a part of the pathologic diagnosis in order
to predict pharmacologic responses to EGFR inhibitors. For
example, Mellinghoff and colleagues demonstrated that
those GBMs that have the best therapeutic response to
EGFR inhibitors are characterized by the coexpression of
EGFRvIII and PTEN.24
Losses on Chromosome 10/PTEN Mutation
Some of the most frequent genetic deletions in GBMs

involve chromosome 10 and occur as losses of the entire
chromosome or as losses of only the long or short arms. Most
interest focuses on 10q, since it is commonly implicated in
high-grade progression and is the location of PTEN. Located
at 10q23.3, PTEN is mutated in 20% to 40% of GBMs and
generally occurs in the setting of chromosome 10 allelic loss.
Both chromosome 10 losses and PTEN mutations are much
more frequent in diffuse forms of astrocytoma than oligodendrogliomas. They are also relatively late events in the
progression to GBM, since both are much more frequent in
GBMs than in AAs or grade 2 astrocytomas.
The high frequency of chromosome 10 losses in GBMs
( 80%) and the tight correlation of its loss with the GBM
phenotype might suggest that it would not be a useful
prognostic marker across tumor grades. Indeed, in studies of
high-grade astrocytomas (AAs and GBMs), loss of heterozygosity (LOH) of chromosome 10 (either 10p or 10q) has been
105
DANIEL J. BRAT
an independent predictor of poor prognosis in large part

because LOH was highly associated with AAs that had short
survivals. Some investigators have therefore suggested that
AAs with LOH of 10q may behave clinically as GBMs.
Overall, it appears that chromosome 10 losses could serve as
a marker of astrocytic differentiation in diffuse high-grade
gliomas and of poor prognosis among high-grade astrocytomas, with special relevance for predicting aggressive AAs.
MGMT
Current standard of care chemotherapy (temozolomide)

used to treat GBM acts by crosslinking DNA by alkylating at
the O6 of guanine. DNA crosslinking is reversed by the DNA
repair enzyme MGMT (O6-methylguanine-DNA methyltransferase). Thus, low levels of MGMT expression would be
expected to be associated with an enhanced response to
therapy. The expression level of MGMT is determined in
large part by the methylation status of the genes promoter.
This epigenetic silencing of MGMT occurs in 40% to 50% of
GBMs and can be assessed by its promoter methylation
status on polymerase chain reaction (PCR)-based tests of
genomic DNA. Epigenetic silencing of MGMT in tumoral
tissue is associated with response to BCNU therapy and
improved survival in patients with GBM.
A recent investigation of temozolamide for the treatment
of GBM found that epigenetic gene silencing of MGMT was
associated with a longer survival, independent of treatment.2,25 The study also demonstrated a survival advantage
among those patients treated with temozolomide and radiotherapy whose GBMs had a silenced MGMT gene. Since
temozolamide is now standard of care for the treatment of
GBM, testing for MGMT status is becoming an important
component of a complete diagnostic work-up.
IDH1
Mutations in isocitrate dehydrogenase 1 (IDH1) are frequent in grade 2 and 3 astrocytomas, oligodendrogliomas,
and oligoastrocytomas, as well as the GBMs that progress
from these lower-grade lesions (i.e., secondary GBMs).26,27
IDH2 mutations have also been noted in these same neoplasms, but at much lower frequency. Since IDH mutations
occur in diffuse gliomas with astrocytic, oligodendroglial,
and mixed histologies, it is believed that they are an early
event, preceding molecular alterations such as TP53 mutations and 1p/19q codeletion, which are associated with
astrocytic and oligodendroglial histologies, respectively.
Within all histologic types and grade of diffuse glioma, the
presence of IDH mutations is associated with prolonged
survival. In GBMs, the finding of IDH1 mutations is
strongly associated with patients with younger age and a
substantially longer survival.26,27
Importantly, over 90% of IDH1 mutations in the diffuse
gliomas occur at a specific site and are characterized by a
base exchange of guanine to adenine within codon 132,
resulting in an amino acid change from arginine to histidine
(R132H). Because of this consistent protein alteration, a

monoclonal antibody has been developed to the mutant
protein, allowing its use in paraffin-embedded specimens
(mIDH1R132H).28 The initial characterization of this antibody was performed on 186 gliomas of various histologies
and grades that had been sequenced for IDH1 mutations
and showed excellent sensitivity and specificity of the antibody for mutations. Moreover, the ability of the antibody to
detect only a minor component of the tissue as mutant may
give this method greater sensitivity than sequencing for
identifying R132H mutant gliomas.
The Cancer Genome Atlas Project
The Cancer Genome Atlas (TCGA) is a large-scale collaborative effort supported by the National Cancer Institute
and the National Human Genome Research Institute that
has provided an integrated platform for defining the molecular alterations of cancer that are associated with pathologic
and radiologic features, clinical behaviors, and response to
therapy. One of the first three malignancies targeted by the
TCGA pilot project was GBM, resulting in the identification
of specific genetic mutations, copy number variations, chromosomal translocations, gene and microRNA expression,
and DNA methylation patterns in nearly 500 tumor samples. All data is publically available for intensive correlative
analysis. The genomic component of this investigation confirmed frequent alterations in the p53 (TP53 mutations,
p14ARF deletion, MDM2 amplification), RB (CDKN2A
deletion, CDK4 amplification, RB1 deletion/mutation), and
receptor tyrosine kinase signaling pathways (EGFR,
PDGFRA, and ERBB2 amplification, PTEN and PI(3)K
mutation, NF1 deletion).29 The TCGA project also led to the
identification of four distinct molecular subtypes based on
transcriptional profiles: proneural, neural, classical, and
mesenchymal.30 These expression classes have variable associations with genomic alterations. For example, the proneural class has a high frequency of IDH1 mutations and
PDGFR amplifications; NF1 mutations and deletions are
most frequent in the mesenchymal class; and the classical
class has a high frequency of EGFR amplifications. However, these genetic associations are not absolute, and GBMs
with EGFR and PDGFR amplification, TP53 and PTEN
mutations, and CDKN2A deletions are noted in each transcriptional class. The proneural gene expression signature is
associated with improved clinical outcome. Much of this
survival advantage is due to the inclusion of tumors with
IDH mutations. This specific subset of GBMs, those within
the proneural expression class and with IDH mutations, is
tightly correlated with the CpG island methylator phenotype (G-CIMP).31 The prognostic difference among the other
three transcriptional classes in the TCGA data is not statistically significant. However, this robust transcriptional classification may lead to the identification of class-specific
therapeutic targets that are directed at underlying molecular mechanisms.
Author
Daniel J. Brat*
106
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Louis DN, Ohgaki H, Wiestler OD, et al. WHO Classification of Tumours
of the Central Nervous System. 4th ed. Lyon: Intl. Agency for Research; 2007.
2. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus
concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med.
2005;352:987-996.
3. Swanson KR, Bridge C, Murray JD, et al. Virtual and real brain tumors:
using mathematical modeling to quantify glioma growth and invasion. J Neurol Sci. 2003;216:1-10.
4. Bellail AC, Hunter SB, Brat DJ, et al. Microregional extracellular matrix
heterogeneity in brain modulates glioma cell invasion. Int J Biochem Cell
Biol. 2004;36:1046-1069.
5. Gupta M, Djalilvand A, Brat DJ. Clarifying the diffuse gliomas: an
update on the morphologic features and markers that discriminate oligodendroglioma from astrocytoma. Am J Clin Pathol. 2005;124:755-768.
6. Brat DJ, Prayson RA, Ryken TC, et al. Diagnosis of malignant glioma:
Role of neuropathology. J Neurooncol. 2008;89:287-311.
7. Giannini C, Scheithauer BW, Burger PC, et al. Cellular proliferation in
pilocytic and diffuse astrocytomas. J Neuropathol Exp Neurol. 1999;58:46-53.
8. Hsu DW, Louis DN, Efird JT, et al. Use of MIB-1 (Ki-67) immunoreactivity in differentiating grade II and grade III gliomas. J Neuropathol Exp
Neurol. 1997;56:857-65.
9. Moskowitz SI, Jin T, Prayson RA. Role of MIB1 in predicting survival in
patients with glioblastomas. J Neurooncol. 2006;76:193-200.
10. Brat DJ, Van Meir EG. Vaso-occlusive and prothrombotic mechanisms
associated with tumor hypoxia, necrosis, and accelerated growth in glioblastoma. Lab Invest. 2004;84:397-405.
11. Rong Y, Durden DL, Van Meir EG, et al. Pseudopalisading necrosis in
glioblastoma: a familiar morphologic feature that links vascular pathology,
hypoxia, and angiogenesis. J Neuropathol Exp Neurol. 2006;65:529-539.
12. Brat DJ, Castellano-Sanchez AA, Hunter SB, et al. Pseudopalisades in
glioblastoma are hypoxic, express extracellular matrix proteases, and are
formed by an actively migrating cell population. Cancer Res. 2004;64:920-927.
13. Dong S, Nutt CL, Betensky RA, et al. Histology-based expression
profiling yields novel prognostic markers in human glioblastoma. J Neuropathol Exp Neurol. 2005;64:948-955.
14. Holash J, Maisonpierre PC, Compton D, et al. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF. Science.
1999;284:1994-1998.
15. Tehrani M, Friedman T, Olson JJ, et al. Intravascular thrombosis in
central nervous system malignancies: a potential role in astrocytoma progression to glioblastoma. Brain Pathol. 2008;18:164-171.
16. Senger DR, Galli SJ, Dvorak AM, et al. Tumor cells secrete a vascular
permeability factor that promotes accumulation of ascites fluid. Science.
1983;219:983-985.
17. Plate KH. Mechanisms of angiogenesis in the brain. J Neuropathol Exp

Neurol. 1999;58:313-320.
18. Brat DJ, Bellail AC, Van Meir EG. The role of interleukin-8 and its
receptors in gliomagenesis and tumoral angiogenesis. Neuro-oncol. 2005;7:
122-133.
19. Straume O, Chappuis PO, Salvesen HB, et al. Prognostic importance of
glomeruloid microvascular proliferation indicates an aggressive angiogenic
phenotype in human cancers. Cancer Res. 2002;62:6808-6811.
20. Cairncross G, Berkey B, Shaw E, et al. Phase III trial of chemotherapy
plus radiotherapy compared with radiotherapy alone for pure and mixed
anaplastic oligodendroglioma: Intergroup Radiation Therapy Oncology Group
21. Smith JS, Alderete B, Minn Y, et al. Localization of common deletion
regions on 1p and 19q in human gliomas and their association with histological subtype. Oncogene. 1999;18:4144-4152.
22. Perry A, Fuller CE, Banerjee R, et al. Ancillary FISH analysis for 1p
and 19q status: preliminary observations in 287 gliomas and oligodendroglioma mimics. Front Biosci. 2003;8:a1-a9.
23. Brat DJ, Seiferheld WF, Perry A, et al. Analysis of 1p, 19q, 9p, and 10q
as prognostic markers for high-grade astrocytomas using fluorescence in situ
hybridization on tissue microarrays from Radiation Therapy Oncology Group
trials. Neuro-oncol. 2004;6:96-103.
24. Mellinghoff IK, Wang MY, Vivanco I, et al. Molecular determinants of
the response of glioblastomas to EGFR kinase inhibitors. N Engl J Med.
2005;353:2012-2024.
25. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and
benefit from temozolomide in glioblastoma. N Engl J Med. 2005;352:9971003.
26. Parsons DW, Jones S, Zhang X, et al. An integrated genomic analysis of
human glioblastoma multiforme. Science. 2008;321:1807-1812.
27. Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in gliomas.
N Engl J Med. 2009;360:765-773.
28. Capper D, Weissert S, Balss J, et al. Characterization of R132H
mutation-specific IDH1 antibody binding in brain tumors. Brain Pathol.
2010;20:245-254.
29. The Cancer Genome Atlas Network. Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature.
2008;455:1061-1068.
30. Verhaak RG, Hoadley KA, Purdom E, et al. Integrated genomic
analysis identifies clinically relevant subtypes of glioblastoma characterized
by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010;17:
98-110.
31. Noushmehr H, Weisenberger DJ, Diefes K, et al. Identification of a CpG
island methylator phenotype that defines a distinct subgroup of glioma.
Cancer Cell. 2010;17:510-522.
107
Future Directions in Glioblastoma Therapy

By Howard Colman, MD, PhD
Overview: The standard of care for both newly diagnosed and

recurrent glioblastoma (GBM) patients has changed significantly in the past 10 years. Surgery followed by radiation and
concurrent and adjuvant temozolomide is now the wellestablished standard treatment for newly diagnosed GBM.
More recently, bevacizumab has become a mainstay of treatment for recurrent GBM. However, despite these advances and
significant improvements in patient outcomes, the management and treatment of GBM patients remains a challenging
and frustrating endeavor. Difficulties in interpretation of imaging changes after initial treatment, as well as the effects of
antiangiogenic agents like bevacizumab on MRI characteristics, can make even the determination of disease progression
complicated in multiple situations. Although a high percentage of patients benefit from antiangiogenic therapy in terms
of radiographic response and progression-free survival, the
LIOBLASTOMA (GBM) IS the most common malignant brain tumor in adults, with approximately 15,000
new GBM cases diagnosed each year in the United States.
GBM is classified by the World Health Organization (WHO)
as a grade 4 astrocytic tumor and is differentiated from
lower grade astrocytomas by the pathologic features of
pseudopalisading necrosis and microvascular proliferation,
among others.
Standard Therapy for Newly Diagnosed GBM
Radiographically, these tumors typically present as ring

enhancing masses on contrast-enhance MRI imaging, often
associated with marked infiltration and edema of surrounding brain. Standard treatment for newly diagnosed GBM
includes maximal safe resection followed by chemoradiation,
as defined by the phase III study by Stupp and colleagues.1
This standard treatment includes fractionated external
beam radiation given over 6 weeks to a dose of 60 Gy
combined with daily temozolomide at a dose of 75 mg/m2.
This initial concurrent chemoradiation phase is followed by
adjuvant temozolomide treatment at a dose of 150 mg/m2
day 1 through 5 out of 28 for the first cycle, with dose
escalation to 200 mg/m2 day 1 through 5 out of 28 for
subsequent cycles. The duration of adjuvant therapy in the
initial study was 6 months, but subsequent studies have
continued adjuvant treatment out to 12 months or more.
Although this regimen demonstrated significant improvement in median survival compared with radiation alone
(14.6 months compared with 12.1 months, p 0.001) and a
benefit in 2-year survival rates (26.5% compared with
10.4%), long-term survival for GBM patients remains disappointing. In a follow-up analysis of the initial phase III
study, 2-, 3-, and 5-year survival rates were 27.2%, 16.0%,
and 9.8%, respectively in the temozolomide chemoradiation
group.2 More recently, a phase III study comparing two dose
schedules of adjuvant temozolomide after chemoradiation
(RTOG 0525) was reported with no significant difference in
overall survival between the standard and dose-dense temozolomide groups.3 Together, these studies demonstrate that
temozolomide chemoradiation has a significant benefit in
newly diagnosed GBM compared to radiation alone (and
prior chemotherapy/radiation combinations). However, de-
108
effects of bevacizumab on prolonging overall survival remain

controversial. Furthermore, tumor progression after treatment
with antiangiogenic agents carries a particularly poor prognosis and there is a general lack of effective therapies for this
group of patients. These limitations in terms of standard
treatments contrast with a relative wealth of new information
regarding the molecular underpinnings of GBM. Data from
several large-scale efforts to molecularly profile GBM tumors
including The Cancer Genome Atlas (TCGA) project have
helped define specific molecular subtypes of GBM with distinct biology and clinical outcomes. These findings are helping
to refine our understanding of the molecular heterogeneity
and pathogenesis of these tumors and provide a basis for the
future development of rational and targeted therapies for
specific tumor subtypes.
spite this improved standard of care, the majority of patients

still die of their disease before 2 years.
The Problem of Pseudoprogression
One aspect of the management of GBM patients that

causes significant uncertainty is the issue of interpretation
of MRI imaging and disease progression. Gadoliniumenhanced MRI remains the mainstay of assessment of GBM
status. However, changes in enhancement on MRI reflect
changes in vascular permeability, and are thus an imaging
surrogate for disease status rather than a reflection of true
tumor burden. Other causes of increased vascular permeability can increase the extent and degree of enhancement on
MRI, resulting in a potentially incorrect determination of
tumor progression. This need to differentiate pseudoprogression (a radiographic increase in MRI enhancement or
edema resulting from tissue injury or inflammation without
increased tumor activity) from true tumor progression represents a clinical and diagnostic challenge, particularly in
patients who have recently completed radiation with concurrent chemotherapy.4 Pseudoprogression typically stabilizes
or improves without alteration in treatment within 3 to 6
months. Pseudoprogression rates have been reported between 9% and 31% in patients with malignant glioma after
chemoradiation, with significantly higher incidence in tumors with MGMT promoter methylation.5,6
Since conventional MRI often cannot distinguish true
progression from pseudoprogression, a lot of attention has
been focused on other advanced imaging modalities. Although MRI perfusion, MRI spectroscopy, and PET can add
diagnostic information and aid in clinical decision making,
none of these are sufficiently sensitive or specific to definitively determine the underlying etiology.7 An incorrect determination of progression in these patients can lead to
From the Department of Neurosurgery and Huntsman Cancer Institute, University of

Utah, Salt Lake City, UT.
Address reprint requests to Howard Colman, MD, PhD, 175 North Medical Drive East,
Salt Lake City, UT, 84132; email: howard.colman@hci.utah.edu.
1092-9118/10/1-10
FUTURE DIRECTIONS IN GBM THERAPY
premature discontinuation of an effective therapy and/or

initiation of less effective therapies. In addition, treatment
of pseudoprogression as progression with an alteration in
therapy can lead to a misinterpretation of subsequent imaging improvement as a response to the new therapy. For
these reasons, a clear understanding regarding the issues
and limitations of current imaging is crucial for clinicians
managing these patients. To help guide these decisions, the
recent recommendations by the Response Assessment in
Neuro-Oncology (RANO) Working Group are that progression can only be determined with certainty in the 12 weeks
after initial chemoradiation by pathologic confirmation or if
the majority of the new enhancement is outside the original
high-dose radiation field.8 Patients in which the determination of progression cannot be made with certainty should not
be enrolled in clinical trials during this time period, and
patients who remain clinically stable or in whom pseudoprogression is suspected should continue with their current
therapy.
Standard Therapies for Recurrent GBM
Following disease progression, historic outcomes for treatment of recurrent disease have been disappointing. In
pooled analyses of multiple clinical trials for recurrent
GBM, response rates have ranged from 4% to 7%; 6-month
KEY POINTS
Radiation therapy with concurrent temozolomide and

adjuvant temozolomide is the established standard
treatment for newly diagnosed glioblastoma (GBM).
Pseudoprogression is an important complicating factor in the interpretation of MRI imaging of GBM
patients and practitioners need to have a good awareness of this issue and familiarity with current recommendations by the Response Assessment in NeuroOncology (RANO) Working Group.
Although multiple therapeutic options exist for recurrent GBM, bevacizumab monotherapy remains the
most common standard salvage therapy. However,
bevacizumab treatment is associated with modest
effects on overall survival, difficulties in imaging
interpretation of progression, and the development of
resistance and poor prognosis at progression.
Understanding of the molecular alterations associated with distinct molecular subtypes and clinical
outcome in GBM continues to increase. Key biomarkers with prognostic and/or tumor subtype associations include proneural and mesenchymal gene
expression patterns, MGMT promoter methylation,
IDH1 mutation, and CpG island methylation phenotype (CIMP).
The efficacy and evaluation of targeted therapy in
GBM can be complicated by issues of drug permeability and difficulty in assessing biologic endpoints in
the tumor, and these aspects have to be addressed in
current and future clinical trials to increase the
likelihood of further improvements in outcome.
progression-free survival (PFS) rates ranged from 9% to

15%; and median overall survival (OS) rates ranged from
57 months.9,10 Standard treatment options for recurrent
disease are more varied than for newly diagnosed patients,
with recent United States Food and Drug Administration
(FDA) approvals adding to the number and types of standard treatment options. Reresection alone or with placement
of polifeprosan 20 with carmustine is an option for patients
with recurrence in noneloquent areas of the brain.11,12
Treatment with additional cytotoxic chemotherapy is frequently considered, with several potential choices. Retreatment with alternative schedules of temozolomide has some
efficacy, with one study suggesting improved 6-month PFS
and one-year survival in patients who had been stable for
at least several months after completing standard adjuvant
temozolomide, but with poorer outcomes in patients who
were treated with an alternative temozolomide schedule at
the time of failure on standard dose temozolomide.13 Several
other alkylating agents (CCNU, BCNU) and other cytotoxic
chemotherapies (carboplatin, irinotecan, etoposide) are also
options for initial or subsequent progressions. Evaluation of
data from the control arm using CCNU at 100 130 mg/m2
every 6 weeks from a recent phase III study demonstrated
a modest radiographic response rate (4.3%) but a relatively
good PFS-6 rate of 19%,14 which was somewhat better than
the enzastaurin (experimental) arm (11.1%, p 0.13) and
highlighted the potential activity of CCNU in this situation.
Very recently, the FDA approved a device, the NovoTTF100A system, for treatment of recurrent GBM based on a
randomized study demonstrating similar survival outcomes
in patients treated with this system compared with chemotherapy of the physicians choice.
Antiangiogenic Therapies
Glioblastoma has long been recognized as a promising

tumor for antiangiogenic treatments because of its high
vascularity. Perhaps the most significant development in
the treatment of recurrent GBM in recent years has been
the development of antiangiogenic therapies and the FDA
granting of accelerated approval in March 2009 for bevacizumab. Although this agent has altered the standard options and approach for recurrent GBM, it has also become
increasingly clear that this treatment has limitations in
terms of disease control and requires clinicians to modify
their standard interpretation of MRI imaging and determination of disease progression. This approval was based on
two independent phase II studies demonstrating activity.15,16 Both studies showed a high radiographic response
rate and an improvement in 6-month PFS with bevacizumab
treatment compared to prior historic controls treated with
cytotoxic chemotherapies and other targeted agents. In the
multicenter BRAIN study, use of bevacizumab resulted in a
radiographic response rate of 28% (95% CI 18.5%, 40.3%)
and the median duration of response was 4.2 months (95%
CI 3.0%, 5.7%).15 Bevacizumab treatment also resulted in an
increase in the percentage of patients treated with stable or
decreasing corticosteroid dosages. A second study reported
by Kreisl and colleagues demonstrated similar results with
single agent bevacizumab in which 29% of patients achieved
6 months of PFS.16 Based on these results, two phase III
studies testing the addition of bevacizumab to temozolomide
chemoradiation in newly diagnosed GBM have been completed or are in progress, with results expected within the
109
HOWARD COLMAN
next 2 to 3 years. In addition to bevacizumab, several other

agents targeting vascular endothelial growth factor (VEGF)
or its receptor have been tested in clinical trials. In general,
reported radiographic response rates and PFS rates have
been similar or lower with these agents compared with
bevacizumab.17,18 In one recent phase III study testing the
VEGF receptor inhibitor cediranib as monotherapy or with
lomustine compared with lomustine alone in recurrent GBM
found no significant difference in median PFS between these
groups.19
The Problem of Bevacizumab Failure
Although bevacizumab and other antiangiogenesis agents

have demonstrated improved radiographic response rates
and PFS rates in recurrent GBM compared to prior cytotoxic
and targeted therapies, it is less clear that these agents are
really improving OS. In addition, multiple studies to date
have failed to demonstrate a benefit of combination therapy
with bevacizumab. In the setting of bevacizumab failure,
studies of various salvage therapies have demonstrated low
radiographic response rates and 6-month PFS rates in the
0% to 2% range.20,21 This dismal prognosis indicates that
progression in the setting of antiangiogenic therapy likely
represents a change in biology of the tumor. However, the
mechanisms of resistance and strategies for optimizing
treatment at this stage of disease remain an active area of
investigation. Because of concern regarding rebound edema
with discontinuation of bevacizumab, many practitioners
tend to continue bevacizumab in this setting despite radiographic progression. Although data related to the question
of continuation or discontinuation of bevacizumab at the
time of bevacizumab progression continues to evolve, a
recent single institution meta-analysis suggested an improved outcome with continuation of bevacizumab in this
situation.22
Opportunities and Barriers for Improved
Treatment of GBM
Molecular subtypes and candidate therapeutic targets in

GBM. One of the recent therapeutic themes in oncology is
that the identification of signature molecular alterations in
particular tumors in combination with drugs that potently
and specifically target that alteration can have dramatic
effects on patient outcome. The term oncogene addiction
describes the hypothesis that tumor growth and survival is
dependent on activity of one key gene or pathway23 and can
be seen in successes of targeted therapies in specific tumor
subtypes such as imatinib for BCR-ABL-positive CML or
BRAF inhibitors for metastatic melanoma with BRAF mutations. Although similar dramatic advances in GBM have
not been observed yet, notable increases in our knowledge of
the molecular underpinnings of GBM are raising hopes for
targeted breakthroughs in therapy.
Strong data indicate that the pathologic entity defined
under the single WHO diagnosis of GBM consists of multiple
molecular subtypes of tumors. These tumors all share the
histopathologic features of GBM, but demonstrate very
distinct biology and molecular ontology. These data may
point the way to more effective or personalized treatment
based on the molecular features of an individual patients
tumor.24 In particular, several studies have identified gene
expression differences between GBM subtypes. One prominent subtype, referred to as mesenchymal, is characterized
110
by increased levels of expression of genes associated with

mesenchymal differentiation, extracellular matrix, invasion,
and angiogenesis. These mesenchymal tumors are associated with worse prognosis and may be associated with
primary GBMs. Another robust gene expression GBM
subtype identified and validated in multiple data sets is
characterized by increased expression of genes associated
with normal neural tissues or neural development, and has
been called proneural. These proneural tumors are associated with better prognosis and secondary GBMs.25-27
The recent integrated molecular analysis from The Cancer
Genome Atlas Network (TCGA) effort in GBM further found
that specific somatic mutation and DNA copy number alterations were associated with specific gene expression subtypes.27 Proneural tumors were associated with significantly
higher rates of point mutations in IDH1 (p 0.01) or p53
genes (p 0.1), and amplification of PDGRA (p 0.01). In
contrast, mesenchymal tumors were characterized by higher
rates of chromosomal deletions involving NF1 or mutations in NF1. Another gene expression subtype called Classical was associated with higher rates of amplification of
chromosome 7 (including amplification of EGFR) and loss
of chromosome 10. Since these molecular alterations are
associated with differences in clinical behavior and prognosis, these and other less-dominant molecular alterations
are potential candidates for therapies targeting particular
tumor subtypes.
Predictive and Prognostic Biomarkers in GBM
In addition to a better understanding of the molecular

pathogenesis of gliomas, the growing molecular profiling
data have led to the identification of biomarkers that are
beginning to be used to help guide patient therapy. Clinically relevant biomarkers can be roughly divided into those
that are prognostic and predictive. Prognostic markers are
associated with patient outcome or natural history of the
disease in patients without treatment or in patients receiving nontargeted therapies. Predictive biomarkers are associated with differential outcome to treatment with a specific
targeted therapy. Recent studies have identified several
molecular markers prognostic of patient outcome to standard therapy (radiation and temozolomide) in GBM. These
include MGMT promoter methylation28 and a multigene
predictor based on gene expression levels of multiple
genes.25 Mutation in the genes IDH1 and IDH2 was recently
described in low- and high-grade gliomas.29 Mutation of
IDH1 or IDH2 is observed in a small subset of GBM tumors
and is strongly associated with the proneural gene expression phenotype and better prognosis. In addition, evidence
from TCGA has also identified an epigenetically defined
tumor subset that overlaps with IDH1 mutation. This
CIMP, is analogous to prior descriptions of CIMP in colon
cancer and other tumors, although the identity of the genes
methylated in GBM is distinct from these other tumors.30
CIMP-positive tumors are positively associated with IDH1
mutation and show improved prognosis, compared to CIMPnegative tumors. Future work is required to define the
causative role of CIMP, IDH mutation, and their relation to
gene expression phenotype and gliomagenesis in GBM and
lower grade gliomas.
FUTURE DIRECTIONS IN GBM THERAPY

Challenges for the Evaluation and Efficacy of
Targeted Therapies in GBM
As in other solid tumors, there are currently a large

number of ongoing trials in GBM testing various targeted
therapies and other approaches to overcoming resistance
including those targeting DNA repair, receptor tyrosine
kinases and other signaling pathways, and pathways involved in tumor stem-cell maintenance and regulation.
Unfortunately, the use of targeted therapies in brain tumors and their evaluation in clinical trials face some additional challenges that are different than with solid tumors
elsewhere in the body. First, many small molecule inhibitors have relatively poor penetration into the brain and
cerebrospinal fluid, resulting in lower effective dose. Second,
evaluating the biologic effects of a drug on its target in
brain tumors is challenging because of its anatomic location
and increased difficulty in doing repeat biopsies during or
after treatment compared to some other systemic cancers.

Nonetheless, if experience from other solid tumors is any
indication, future successful clinical trials in recurrent
GBM will likely need to have a larger emphasis on pharmacodynamics and biologic tissue effects as important endpoints.
In summary, the dramatic increase in our understanding
of the molecular underpinnings and subtypes of glioblastoma, along with increases in the number and type of
available small molecule inhibitors, indicates that improvement in patient outcome in GBM with appropriate targeted
therapy is indeed feasible. However, the success and rapidity of these developments will hinge on successful integration of molecular and preclinical data with the more efficient
and innovative clinical trial designs in order to identify
those key pathways and drugs that will have the highest
benefit for particular GBM subtypes.
Author
Howard Colman
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Castle
Biosciences
REFERENCES
2005;352:987-996.
2. Stupp R, Hegi ME, Mason WP, et al. Effects of radiotherapy with
concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the
EORTC-NCIC trial. Lancet. 2009;10:459-466.
3. Gilbert MR, et al. M.P. RTOG 0525: A randomized phase III trial
comparing standard adjuvant temozolomide (TMZ) with a dose-dense (dd)
schedule in newly diagnosed glioblastoma (GBM). Neuro-Oncology. 2011;
13:3s (suppl; abstr 46).
4. de Wit MC, de Bruin HG, Eijkenboom W, et al. Immediate postradiotherapy changes in malignant glioma can mimic tumor progression.
Neurology. 2004;63:535-537.
5. Brandsma D, van den Bent MJ. Pseudoprogression and pseudoresponse
in the treatment of gliomas. Curr Opin Neurol Neurosurg. 2009;22:633-638.
6. Brandes AA, Franceschi E, Tosoni A, et al. MGMT promoter methylation
status can predict the incidence and outcome of pseudoprogression after
concomitant radiochemotherapy in newly diagnosed glioblastoma patients.
J Clin Oncol. 2008;26:2192-2197.
7. Brandsma D, Stalpers L, Taal W, et al. Clinical features, mechanisms, and
management of pseudoprogression in malignant gliomas. Lancet. 2008;9:453-461.
8. Wen PY, Macdonald DR, Reardon DA, et al. Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology
working group. J Clin Oncol. 2010;28:1963-1972.
9. Wong ET, Hess KR, Gleason MJ, et al. Outcomes and prognostic factors
in recurrent glioma patients enrolled onto phase II clinical trials. J Clin
Oncol. 1999;17:2572-2578.
10. Lamborn KR, Yung WK, Chang SM, et al. Progression-free survival: an
important end point in evaluating therapy for recurrent high-grade gliomas.
Neuro-Oncology. 2008;10:162-170.
11. Barker FG 2nd, Chang SM, Gutin PH, et al. Survival and functional
status after resection of recurrent glioblastoma multiforme. Neurosurgery.
1998;42:709-720.
12. Brem H, Piantadosi S, Burger PC, et al. Placebo-controlled trial of
safety and efficacy of intraoperative controlled delivery by biodegradable
polymers of chemotherapy for recurrent gliomas. The Polymer-brain Tumor
Treatment Group. Lancet. 1995;345:1008-1012.
13. Perry JR, Belanger K, Mason WP, et al. Phase II trial of continuous
dose-intense temozolomide in recurrent malignant glioma: RESCUE study.
J Clin Oncol. 2010;28:2051-2057.
14. Wick W, Puduvalli VK, Chamberlain MC, et al. Phase III study of
enzastaurin compared with lomustine in the treatment of recurrent intracranial glioblastoma. J Clin Oncol. 2010;28:1168-1174.
15. Friedman HS, Prados MD, Wen PY, et al. Bevacizumab alone and in
combination with irinotecan in recurrent glioblastoma. J Clin Oncol. 2009;
27:4733-4740.
16. Kreisl TN, Kim L, Moore K, et al. Phase II trial of single-agent
bevacizumab followed by bevacizumab plus irinotecan at tumor progression
in recurrent glioblastoma. J Clin Oncol. 2009;27:740-745.
17. Norden AD, Drappatz J, Wen PY. Antiangiogenic therapies for highgrade glioma. Nat Rev Neurol. 2009;5:610-620.
18. Reardon DA, Turner S, Peters KB, et al. A review of VEGF/VEGFRtargeted therapeutics for recurrent glioblastoma. J Natl Compr Canc Netw.
2011;9:414-427.
19. Batchelor T, Mulholland P, Nyns B, et al. The efficacy of cediranib as
monotherapy and in combination with lomustine compared to lomustine alone
in patients with recurrent glioblastoma: A phase III randomized study. Neuro
Oncol. 2010;12(suppl 4).
20. Iwamoto FM, Abrey LE, Beal K, et al. Patterns of relapse and prognosis
after bevacizumab failure in recurrent glioblastoma. Neurology. 2009;73:1200-1206.
21. Quant EC, Norden AD, Drappatz J, et al. Role of a second chemotherapy
in recurrent malignant glioma patients who progress on bevacizumab. Neuro
Oncol. 2009;11:550-555.
22. Reardon DA, Vredenburgh JJ, Desjardins A, et al. Bevacizumab (BV)
continuation following BV progression: Meta-analysis of five consecutive recurrent glioblastoma (GBM) trials. J Clin Oncol. 2011;29:45s(suppl; abstr 2030).
23. Weinstein IB, Joe A. Oncogene addiction. Cancer Res. 2008;68:3077-3080.
24. Colman H, Aldape K. Molecular predictors in glioblastoma: toward
personalized therapy. Arch Neurol. 2008;65:877-883.
25. Colman H, Zhang L, Sulman EP, et al. A multigene predictor of
outcome in glioblastoma. Neuro-Oncology. 2010;12:49-57.
26. Phillips HS, Kharbanda S, Chen R, et al. Molecular subclasses of
high-grade glioma predict prognosis, delineate a pattern of disease progression, and resemble stages in neurogenesis. Cancer Cell. 2006;9:157-173.
27. Verhaak RG, Hoadley KA, Purdom E, et al. Integrated genomic analysis
identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010;17:98-110.
28. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit
from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-1003.
N Engl J Med. 2009;360:765-773.
30. Noushmehr H, Weisenberger DJ, Diefes K, et al. Identification of a CpG
Cancer Cell. 2010;17:510-522.
111
Establishing the Standard of Care for

Patients with Newly Diagnosed and
Recurrent Glioblastoma
By Mark R. Gilbert, MD
Overview: The current standard of care for patients with

newly diagnosed glioblastoma includes maximal safe tumor
resection followed by concurrent external-beam radiation with
daily low-dose temozolomide followed by 6 to 12 months of
adjuvant temozolomide, typically by using a cycle of 5 consecutive days out of 28. Efforts to improve on these results
from the European Organisation for Research and Treatment
of Cancer (EORTC)/National Cancer Institute of Canada (NCIC)
trial using either dose-dense chemotherapy strategies or
combinations with signal transduction modulators have, to
date, been unsuccessful. Two large international randomized
trials examining the efficacy of adding bevacizumab, an antiangiogenic agent, to the standard treatment have been completed, with expectations of results within in the next 2 years.
For recurrent glioblastoma, there are no firmly established
LIOBLASTOMA IS classified as a grade 4 glioma by

using the now universally accepted WHO grading
criteria. It is the most common malignant primary brain
neoplasm, accounting for 54% of all gliomas.1 The annual
incidence in the United States is estimated to be between
3 and 4 per 100,000 of the general population. Although
glioblastoma is found in all age groups, the highest incidence
is in middle age to elderly. Although distinct pathologic
characteristics such as microvascular proliferation and necrosis are used to define the disease, recent advances in
molecular characterization demonstrate that the uniform
histologic criteria belie tremendous genetic heterogeneity.
Primary glioblastoma, developing as a glioblastoma, has
distinct molecular characteristics from the secondary glioblastoma which evolves from the progression transformation
from lower grade gliomas.2 These differences may influence
treatment decisions and in the case of mutations in the
IDH1 gene which is found almost exclusively in secondary
glioblastoma, clear implications on prognosis.3
The prognosis for patients with glioblastoma remains
poor. Overall, the 1-year survival rate in population-based
studies is less than 40%, although patients enrolled on
clinical trials do have better reported outcomes.1 Several
clinical prognostic factors have been identified, including
age, performance status, and extent of initial tumor resection.4 Several of the important clinical prognostic factors
have been combined by using recursive partitioning analysis
into a six-class system, allowing easier stratification accounting in clinical trials.5 Median survival for glioblastoma
ranges from 6 months to 18 month according to clinical
prognostic factors in the recursive partitioning analysis.
This underscores the importance of clinical factors in outcome and the potential for misinterpretation of clinical trial
outcomes if there is an imbalance of these factors. Additionally, molecular prognostic factors have been identified that
are independent of the clinical factors. These include IDH1
mutation, presence of a glioma-specific CpG Island hypermethylation phenotype (G-CIMP), a gene expression profile,
and the methylation of the promoter region of the methylguanine methyltransferase (MGMT) gene.3,6-8
112
standards of care. Although intracavitary insertion of carmustineimpregnated polymers has been approved by the U.S. Food
and Drug Administration (FDA), this strategy is not widely
used. Bevacizumab has been FDA approved for recurrent
glioblastoma, but no randomized trial has clearly demonstrated a survival benefit. Alternative dosing schedules of
temozolomide (i.e., metronomic) has modest activity even in
patients with prior temozolomide exposure. Clinical trials
testing small-molecule signal transduction modulators have
been disappointing, although most report a small response
rate, suggesting that molecularly definable tumor subpopulations may help guide treatment decisions. Successful
implementation of marker-based treatment would lead to
personalized care and the creation of individualized standards
of care.
Treatment of Newly Diagnosed Glioblastoma
The seminal work of the Brain Tumor Study Group in

the 1970s established the efficacy of external-beam radiation
for glioblastoma.9 Their studies demonstrated a more than
doubling of survival from 3 to 4 months with surgery and
corticosteroids to more than 9 months with radiation treatment. The addition of chemotherapy, typically a nitrosourea
such as lomustine or carmustine, did not substantially
improve survival compared with radiation alone. In fact, a
large meta-analysis compiling 12 randomized trials and
including more than 3,000 patients found only a 6% improvement in 1-year survival when chemotherapy was
added to surgery and conventional radiation.10
However, a new standard of care for patients with glioblastoma was established with the 2005 publication of the
results of a clinical trial performed by the European Organisation for Research and Treatment of Cancer (EORTC) and
National Cancer Institute of Canada (NCIC).11 This was a
large, randomized phase III trial that compared regional
external-beam radiation (60 Gy in 2-Gy fractions) as the
standard arm with the experimental treatment comprising
concurrent regional radiation (60 Gy in 2-Gy fractions) with
daily temozolomide (75 mg/m2 for 42 consecutive days). The
chemoradiation was then followed by six cycles of singleagent temozolomide (150 to 200 mg/m2/day for 5 consecutive
days of every 28 day cycle). The treatment schema is
provided in Fig. 1. The clinical trial demonstrated statistically significant improvements in overall survival (14.6 vs.
12.1 months), progression-free survival, and the 2-year survival rate (26.5% vs. 10%). Long-term follow-up continued to
demonstrate improved outcomes with a 5-year survival rate
From the Department of Neuro-oncology, University of Texas M. D. Anderson Cancer

Center, Houston, TX.
Address reprint requests to Mark R. Gilbert, MD, Department of Neuro-oncology,
University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX
77030; email: mrgilbert@mdanderson.org.
1092-9118/10/1-10
ESTABLISHING TREATMENTS FOR GLIOBLASTOMA
Fig. 1. Treatment schema for the chemoradiation regimen from the EORTC/NCI-Canada Clinical Trial.
of 9.8% in the combined treatment patients compared with

1.9% in the radiation-only group.12
A companion laboratory correlative study to the EORTC/
NCIC trial evaluating patient outcomes on the basis of the
methylation status of the methylguanine methyltransferase
(MGMT) gene was performed by Hegi and colleagues.8
Hypermethylation of the promoter region was postulated to
decrease or eliminate gene expression and that with absent
or low intracellular MGMT, a major mechanism of repair of
alkylating agentmediated DNA damage. Therefore, in the
presence of the methylated MGMT promoter region, the
addition of temozolomide would result in a better response
rate. They were able to analyze 206 and found that 45% were
had MGMT promoter methylation. These studies did demonstrate an improved outcome for patients with tumors
harboring methylated MGMT gene promoter regions. However, although not as marked, the outcomes for patients
with unmethylated tumors experienced improvement with
the addition of temozolomide, suggesting that MGMT methylation status is clearly prognostic but not fully predictive of
treatment benefit.
These findings, correlating MGMT promoter methylation
with outcomes, suggested that the MGMT enzyme may be
an important therapeutic target and depletion of intracellular MGMT in tumor cells may enhance temozolomide
efficacy. Prolonged exposure to temozolomide has been demonstrated to decrease MGMT activity in peripheral-blood
mononuclear cells in patients receiving 21 consecutive days
of treatment.13 This approach, using a dose-dense treatment
schedule, was tested in Radiation Therapy Oncology Group
KEY POINTS
A standard of care consisting of concurrent radiation

and temozolomide followed by adjuvant temozolomide has been established for patients with newly
diagnosed glioblastoma.
Recently completed and ongoing randomized clinical
trials in newly diagnosed glioblastoma seek to enhance the current therapy by adding synergistic
therapies.
There are very few treatments established for recurrent glioblastoma despite extensive testing of many
signal transduction modulators.
Antiangiogenic therapies look promising for recurrent glioblastoma. Bevacizumab is administered frequently in this setting.
Given the molecular heterogeneity of glioblastoma,
further advances will likely require a comprehensive
effort merging tumor profiling with targeted
therapies.
(RTOG) 0525, an international collaborative phase III

trial.14 Enrollment criteria included adults older than 18
years with available tumor tissue blocks with more than
1 cm2 of tumor and Karnofsky performance score of 70 or
greater. All patients received concurrent radiation with
daily temozolomide and then were randomly assigned to
either standard adjuvant temozolomide (150 to 200 mg/m2/
day for 5 consecutive days of a 28-day cycle) or dose-dense
temozolomide (75 to 100 mg/m2/day for 21 consecutive days
of a 28-day cycle). Patients could receive treatment for up
to 12 cycles of adjuvant therapy. The study accrued 1,173
patients with 833 undergoing random assignment. The two
treatment arms were well balanced in all parameters including age, gender, performance status, extent of tumor resection and MGMT methylation status. The results, presented
at the 47th ASCO Annual Meeting (June 4 8, 2011, Chicago,
IL), demonstrated that there was no improvement in either
overall or progression-free survival with the use of the
dose-dense temozolomide schedule. Additionally, subset
analysis by MGMT methylation status did not show a
selective benefit for dose-dense treatment for tumors with
either methylated or unmethylated MGMT gene promoter
phenotype. However, this prospective study did demonstrate
that MGMT methylation status is clearly prognostic.
Other strategies have been evaluated as potential enhancers of the established efficacy of the chemoradiation regimen. A series of signal transduction modulators and other
similar agents have been evaluated, typically in single-arm
phase II trials. Agents such as erlotinib, talampanel, and
Poly ICLC have been tested with promising results compared with the historic controls provided by the EORTC
study.15,16 However, the validity of this comparison has been
questioned, and concerns regarding patient population differences and the more recent availability of effective salvage
regimens have limited interest in pursuing large-scale clinical trials.
Interest in the use of antiangiogenic strategies has generated several large-scale randomized clinical trials. Bevacizumab, a humanized monoclonal antibody targeting
vascular endothelial growth factor (VEGF) A, has demonstrated activity in patients with recurrent glioblastoma
(discussed in more detail later herein). This has led to the
development and completion of two randomized, doubleblind placebo-controlled trials for patients with newly diagnosed glioblastoma. The results of the two studies (RTOG
0825 and AvaGLIA) are expected within the next 2 years.
Additionally, data from a phase II trial adding the integrin
inhibitor cilengitide to the conventional chemoradiation regimen showed an improved outcome in newly diagnosed,
MGMT-methylated glioblastoma compared with historic
controls and led to a phase III placebo-controlled randomized trial.17 This study recently completed accrual and
outcome results are expected soon. Other studies including
randomized phase II trials evaluating mammalian target
113
MARK R. GILBERT
of rapamycin (mTOR) inhibition, VEGFR inhibition (cediranib), or poly (ADP-ribose) polymerase (PARP) inhibition
are either underway or in planning.
In summary, the pivotal clinical trial performed by the
EORTC and NCIC using radiation and temozolomide established the current standard of care for patients with newly
diagnosed glioblastoma. Efforts to improve on these results
with intensification of the chemotherapy were unsuccessful.
Clinical studies have been performed or are underway to
determine whether there are synergistic therapies that can
further enhance outcomes for this patient population and
thereby establish a new standard of care.
Treatment of Recurrent Glioblastoma
The treatment of recurrent glioblastoma remains unsatisfactory. Tumor resection for recurrent disease does not
appear to substantially affect either the 6-month progressionfree survival rate nor overall survival from the time of
progression.18 However, relief of tumor-induced mass effect
may be clinically beneficial and the confirmation of true
recurrence (rather than treatment-related necrosis or pseudoprogression) may be extremely helpful in treatment decisions.
A wide variety of approaches have been studied for recurrent glioblastoma. These include local strategies often requiring a surgical procedure and systemic chemotherapy
treatments either as single agent or in combination.
Locoregional Treatment Strategies
A variety of treatment approaches have been tested targeting the tumor directly, recognizing that spread of glioblastoma outside of the nervous system is rare. Use of a
carmustine-impregnated bioerodable polymer was tested for
recurrent disease and in a placebo-controlled randomized
trial showed a modest, but statistically significant improvement in the 6-month survival rate.19 These findings led to
FDA approval, but this treatment approach is not used
widely today.
Attempts to improve delivery of agents directly to tumor
led to the institution of convectional enhanced delivery.
Intratumoral catheters delivering small volumes of fluid
under pressure have been shown to spread widely from the
point of infusion. This method was used to deliver cintredekin besudotox (interleukin [IL]-13-PE38QQR, IL-13
pseudomona extoxin) in a series of clinical trials. Despite
strong preclinical data, there was no improvement in outcome compared with the carmustine wafer in a phase III
clinical trial.20 Alternative strategies are looking at using
convection-enhanced delivery for chemotherapies, particularly those like topotecan that do not cross the blood-brain
barrier with systemic delivery.21
Direct injection of vectors containing gene therapies has
also been evaluated. Preclinical studies using transfection of
the herpes thymidine kinase gene into tumor cells by using
a retroviral vector demonstrated high efficacy after administration of acyclovir or ganciclovir. However, clinical trials
in both recurrent and newly diagnosed glioblastoma failed
to demonstrate efficacy, potentially the consequence of
poor vector delivery by using direct injection into the walls
of the tumor cavity.22 More recently, replication-competent
viruses are being evaluated in clinical trials, potentially
reducing the problem of delivery with continued local pro-
114
duction and spread of the virus. For example, the delta-24

virus, a replication-competent adenovirus that requires amplification of the Rb pathway to replicate, is currently being
studied in a clinical trial.23
Systemic Chemotherapy
Metronomic and dose-dense temozolomide. Dose-dense temozolomide has also been investigated in patients with
recurrent glioblastoma.24 Two alternative dosing schedules,
alternating weekly schedule or 21 consecutive days of a
28-day cycle, have been used most commonly. Several phase
II trials have evaluated the 21 of 28-day regimen with
6-month progression-free survival rates ranging from 18% to
30% even in patients with prior temozolomide exposure.
Similarly, retrospective analyses of patients treated with
the week on, week off dose-dense schedule with temozolomide report a 6-month progression-free survival rate as high
as 36%. Both schedules are associated with a high rate of
lymphopenia, although this toxicity may be more pronounced with the 21 of 28-day schedule.
Metronomic schedules of temozolomide have been evaluated in a prospective clinical trial.25 Temozolomide was
administered at low dose (50 mg/m2/day) continuously to
patients who had previously been treated with standard
dosing of temozolomide as a component of first-line treatment (as described previously herein for newly diagnosed
glioblastoma). Interestingly, patients rechallenged after
early failure ( 6 months of adjuvant temozolomide) or after
a treatment-free interval experienced 6-month progressionfree survival rates of 27% and 36%, respectively. Patients
with treatment failure during adjuvant treatment but more
than six cycles fared poorly. These results suggest that in
select subpopulations, re-treatment with temozolomide may
be effective.
Resistance modulation with PARP inhibitors. There is
increasing interest in exploring inhibitors of PARP as a
novel strategy to enhance the efficacy of temozolomide and
address the recent finding that recurrent glioblastomas
acquire either MSH6 mismatch geneinactivating mutations or hypermethylation of the promoter region leading to
reduced expression.26 Several PARP inhibitors are in early
clinical trials, typically in combination with a variety of
temozolomide dosing schedules. The dose-limiting toxicity
is likely to be myelosuppression, supporting the need for
pharmcodynamic studies confirming maximal synergy of
PARP inhibition with temozolomide-induced tumor DNA
damage.
Nitrosoureas and other chemotherapy agents. Nitrosoureas such as carmustine and lomustine were the standard treatment for recurrent malignant gliomas but their
use steadily declined with the introduction of temozolomide,
which is better tolerated and rarely causes cumulative
myelotoxicity. There has been a recent resurgence of nitrosourea use after the report of a 6-month progression-free
survival rate of 19% in patients with prior temozolomide
exposure.27 Other chemotherapy agents such as irinotecan,
carboplatin, cisplatin and procarbazine are used but the
response rates in recurrent disease have been modest.
Targeted agents. As discussed previously, glioblastoma
has been classified into two types primarily on the basis of
genetic features. Although there is some similarity in genetic alterations, such as loss of phosphotase and tensin
homolog on chromosome 10 (PTEN), loss of cyclin-dependent

Table 1. Single-Agent Targeted Therapies for Recurrent Glioblastoma
Agent
Erlotinib
Gefitinib
Imatinib
Pazopanib
Vorinostat
Tipifarnib
Temsirolimus
Enzastaurin
Study
van den Bent

Rich30
Raymond31
Iwamoto32
Galanis33
Cloughesy34
Galanis35
Chang36
Wick27
Kreisl37
29
Year
Target
2009
2004
2008
2010
2009
2006
2005
2005
2010
2010
EGFR
EGFR
C-ABL, C-KIT, PDGFR
VEGFR, PDGFR
HDAC
Farnesyltransferase
Trial
Phase
No. of
Patients
6-Month
Progression-Free
Survival (%)
II
II
II
II
II
II
II
110
53
51
35
66
67
65
41
174
72
11
13; 14
16
3
15
12
8
3
11
7
mTOR
III
I/II
PKC
Abbreviations: EGFR, epidermal growth factor receptor; C-ABL, a non-receptor protein tyrosine kinase; C-KIT, a cell surface protein that binds stem cell factor;
C-MET, met proto-oncogene; PDGFR, platelet-derived growth factor receptor; VEGFR, vascular endothelial growth factor receptor; HDAC, histone deacetylase; mTOR,
mammalian target of rapamycin; PKC, protein kinase C.
kinase inhibitor p16INK4A and amplification of CDK 4,

primary glioblastoma typically has higher incidence of epidermal growth factor receptor (EGFR) amplification with
inactivation of PTEN and p16 tumor suppression genes. In
contrast, secondary glioblastoma multiforme is characterized by mutation of TP53 gene and more recently, almost
exclusively demonstrate mutations in the isocitrate dehydrogenase (IDH) 1 and 2 genes.
Epidermal growth factor (EGF) and its receptor EGFR,
platelet-derived growth factor (PDGF) A and B and their
receptors PDGFR and , VEGF and its receptor (VEGFR),
insulin-like growth factor (IGF)-1 and IGF receptor (IGFR),
transforming growth factor (TGF)-, fibroblast growth factor
(FGF), and hepatocyte growth factor (HGF) are thought to
be critical to the pathogenesis and survival of glioblastoma.
Activation of these tyrosine kinase receptors triggers three
major downstream pathways: mitogen-activated protein
kinase (MAPK); phosphoinositide 3 kinase (PI3K)/Akt; and
phospholipase C (PLC) and protein kinase C (PKC). These
signaling pathways regulate cell proliferation and differentiation and prevent apoptosis, and are therefore logical
targets of treatment for glioblastoma.
Unfortunately, despite these molecular findings in a high
percentage of glioblastoma, as indicated in Table 1, studies
with single-agent signal transduction modulators demonstrated only modest results at best. In particular, even in the
presence of amplification or mutation of EGFR, treatment
with a potent EGFR inhibitor such as erlotinib did not result
in a high response rate. Subsequent analyses suggested that
response to erlotinib occurred only when the downstream
component of the pathway was not already constitutively
activated as indicated by the presence of a functional PTEN
gene regulating Akt.28 These results underscore the complex
nature of signal transduction modulation strategies with
pathway overlap and downstream effectors. Attempts and
combination regimens of signal transduction modulators has
been complicated by overlapping toxicities.
Antiangiogenic agents. One of the hallmarks of glioblastoma is prominent angiogenesis, making this component of
tumor biology a logical target. The newly formed blood
vessels are often poorly developed, with incomplete tight
junction between endothelial cells and tortuous paths with
blind loops. These features lead to peritumoral edema and
account for much of the imaging enhancement by leakage of
systemic administration of contrast material before either
computed tomography or magnetic resonance imaging. A
variety of antiangiogenic agents have been tested, as outlined in Table 2.

Bevacizumab, a humanized monoclonal antibody, works
by sequestering the circulating ligand VEGF-A, resulting in
angiogenesis inhibition. Several phase II studies have been
performed in patients with recurrent glioblastoma. Most
studies demonstrate a response rate ranging from 25% to
40% and a 6-month progression-free survival in the same
range. Importantly, nearly all patients receiving bevacizumab are able to either reduce or stop corticosteroid use.
These findings, most notably in a multicenter randomized
noncomparative phase II trial, led to the accelerated approval of bevacizumab for patients with recurrent glioblastoma.39
A variety of other antiangiogenic agents have been evaluated including aflibercept (VEGF-Trap), a decoy VEGFR
fused to the Fc portion of an immunoglobulin molecule.
However, results of a phase II trial in recurrent glioblastoma
revealed only modest activity. Cediranib is a small-molecule
tyrosine kinase inhibitor of the spectrum of VEGFR that
showed early efficacy, but a subsequent phase III trial with
lomustine failed to demonstrate added benefit from the
cediranib. Carbozantinib (XL184), a small-molecule agent
that targets both VEGFR and c-Met also showed efficacy in
early studies, but systemic toxicities often precluded extended use. Cilengitide is a novel antiangiogenic agent that
targets the v3 integrin that is required for endothelial cell
migration for neovascularization. There may be additional
tumor signal pathway effects that further enhance efficacy.
Single-agent activity has been modest, but as described
herein, synergistic efficacy with chemoradiation is being
tested in newly diagnosed glioblastoma.
Table 2. Antiangiogenic Therapies Investigated for

Recurrent Glioblastoma
Agent
Study
Cediranib
Batchelor38
Bevacizumab Friedman39
Cilengitide
Reardon40
Gilbert41
XL-184
Wen42
Year
Target
2010
2009
2008
2011
2010
pan-VEGFR
VEGF-A
v3 integrin
v5 integrin
VEGFR, C-MET
6-Month
Trial
No. of Progression-Free
Survival (%)
Phase Patients
II
II
II
II
31
85
81
26
124
26
36
15
12
21
Abbreviations: VEGFR, vascular endothelial growth factor receptor; C-MET,

met proto-oncogene; EGFR, epidermal growth factor receptor; VEGF, vascular
endothelial growth factor.
115
MARK R. GILBERT
Conclusion
A standard of care has been established for patients with

newly diagnosed glioblastoma on the basis of level 1 evidence for a randomized clinical trial. Current investigations
are exploring agents to add to the established chemoradiation regimen that would augment activity without a marked
increase in toxicity. Many of these trials are also including
molecular characterization of the tumors in an effort to
define patient subpopulations likely to benefit (or not benefit) from the new regimen, thereby enhancing the risk to
benefit for individual patients.
A similar standard does not currently exist for recurrent

glioblastoma, although antiangiogenic agents particularly
bevacizumab, demonstrate tumor response and tumor control although this is often short-lived and to date, there are
no salvage regimens after bevacizumab failure. Given the
marked molecular heterogeneity of recurrent glioblastoma,
future advances will likely require a major initiative to
perform clinical trials encompassing correlations or patient
selection on the basis of tumor characteristics. This initiative will require both an increase in resource allocation and
a widespread collaborative effort to be successful.
Author
Mark R. Gilbert
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Abbott
Laboratories;
Genentech;
GlaxoSmithKline;
Merck
Honoraria
Research
Funding
Genentech;
Merck
Genentech;
Merck
Expert
Testimony
Other
Remuneration
REFERENCES
1. 2005-2006 Statistical Report: Primary Brain Tumors in the United
States Statistical Report, 1998-2002 (Years Data Collected). Hinsdale, IL:
Central Brain Tumor Registry of the United States; 2006.
2. Wen PY, Kesari S. Malignant gliomas in adults. N Engl J Med.
2008;359:492-507.
N Engl J Med. 2009;360:765-773.
4. Weller M, Felsberg J, Hartmann C, et al. Molecular predictors of
progression-free and overall survival in patients with newly diagnosed
glioblastoma: a prospective translational study of the German Glioma Network. J Clin Oncol. 2009;27:5743-5750.
5. Curran WJ Jr, Scott CB, Horton J, et al. Recursive partitioning analysis
of prognostic factors in three Radiation Therapy Oncology Group malignant
glioma trials. J Natl Cancer Inst. 1993;85:704-710.
6. Noushmehr H, Weisenberger DJ, Diefes K, et al: Identification of a CpG
Cancer Cell 17:510-522.
7. Colman H, Zhang L, Sulman EP, et al. A multigene predictor of outcome
in glioblastoma. Neuro Oncol. 2010;12:49-57.
8. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit
from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-1003.
9. Walker MD, Green SB, Byar DP, et al. Randomized comparisons of
radiotherapy and nitrosoureas for the treatment of malignant glioma after
surgery. N Engl J Med. 1980;303:1323-1329.
10. Stewart LA. Chemotherapy in adult high-grade glioma: a systematic
review and meta-analysis of individual patient data from 12 randomised
trials. Lancet. 2002;359:1011-1018.
2005;352:987-996.
12. Stupp R, Hegi ME, Mason WP, et al. Effects of radiotherapy with
concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the
EORTC-NCIC trial. Lancet Oncol. 2009;10:459-466
13. Tolcher AW, Gerson SL, Denis L, et al. Marked inactivation of O6alkylguanine-DNA alkyltransferase activity with protracted temozolomide
schedules. Br J Cancer. 2003;88:1004-1011.
14. Gilbert MR, Wang M, Aldape K, et al. RTOG 0525: a randomized phase
III trial comparing standard adjuvant temozolomide (TMZ) with a dose-dense
(dd) schedule in newly diagnosed glioblastoma (GBM). J Clin Oncol 2011;29
(suppl; abstr 141s).
15. Prados MD, Chang SM, Butowski N, et al. Phase II study of erlotinib
plus temozolomide during and after radiation therapy in patients with newly
diagnosed glioblastoma multiforme or gliosarcoma. J Clin Oncol. 2009;27:
579-584.
16. Grossman SA, Ye X, Piantadosi S, et al. Survival of patients with newly
116
diagnosed glioblastoma treated with radiation and temozolomide in research

studies in the United States. Clin Cancer Res. 2010;16:2443-2449.
17. Stupp R, Goldbrunner B, Neyns B, et al. Phase I/IIa trial of cilengitide
(EMD121974) and temozolomide with concomitant radiotherapy, followed
by temozolomide and cilengitide maintenance therapy in patients (pts) with
newly diagnosed glioblastoma (GBM). J Clin Oncol. 2007;25 (suppl; abstr
2000).
18. Clarke JL, Ennis MM, Yung WK, et al. Is surgery at progression a
prognostic marker for improved 6-month progression-free survival or overall
survival for patients with recurrent glioblastoma? Neuro Oncol. 2011;13:
1118-1124.
19. Brem H, Piantadosi S, Burger PC, et al. Placebo-controlled trial of
safety and efficacy of intraoperative controlled delivery by biodegradable
polymers of chemotherapy for recurrent gliomas. The Polymer-brain Tumor
Treatment Group. Lancet. 1995;345:1008-1012.
20. Kunwar S, Prados MD, Chang SM, et al. Direct intracerebral delivery
of cintredekin besudotox (IL13-PE38QQR) in recurrent malignant glioma: a
report by the Cintredekin Besudotox Intraparenchymal Study Group. J Clin
Oncol. 2007;25:837-844.
21. Anderson RC, Elder JB, Brown MD, et al. Changes in the immunologic
phenotype of human malignant glioma cells after passaging in vitro. Clin
Immunol. 2002;102:84-95.
22. Rainov NG. A phase III clinical evaluation of herpes simplex virus type
1 thymidine kinase and ganciclovir gene therapy as an adjuvant to surgical
resection and radiation in adults with previously untreated glioblastoma
multiforme. Hum Gene Ther. 2000;11:2389-2401.
23. Jiang H, Gomez-Manzano C, Lang FF, et al. Oncolytic adenovirus:
preclinical and clinical studies in patients with human malignant gliomas.
Curr Gene Ther. 2009;9:422-427.
24. Wick W, Platten M, Weller M. New (alternative) temozolomide regimens for the treatment of glioma. Neuro Oncol. 2009;11:69-79.
25. Perry JR, Rizek P, Cashman R, et al. Temozolomide rechallenge in
recurrent malignant glioma by using a continuous temozolomide schedule:
the rescue approach. Cancer. 2008;113:2152-2157.
26. Yip S, Miao J, Cahill DP, et al. MSH6 mutations arise in glioblastomas
during temozolomide therapy and mediate temozolomide resistance. Clin
Cancer Res. 2009;15:4622-4629.
27. Wick W, Puduvalli VK, Chamberlain MC, et al. Phase III study of
enzastaurin compared with lomustine in the treatment of recurrent intracranial glioblastoma. J Clin Oncol. 2010;28:1168-1174.
28. Haas-Kogan DA, Prados MD, Tihan T, et al. Epidermal growth factor
receptor, protein kinase B/Akt, and glioma response to erlotinib. J Natl
Cancer Inst. 2005;97:880-887.
29. van den Bent MJ, Brandes AA, Rampling R, et al. Randomized phase II
trial of erlotinib versus temozolomide or carmustine in recurrent glioblastoma: EORTC brain tumor group study 26034. J Clin Oncol. 2009;27:1268-1274.

30. Rich JN, Reardon DA, Peery T, et al. Phase II trial of gefitinib in
recurrent glioblastoma. J Clin Oncol. 2004;22:133-142.
31. Raymond E, Brandes A, Van Oosterom A, et al. Multicentre phase II
study of imatinib mesylate in patients with recurrent glioblastoma: an
EORTC/NDDG/BTG Intergroup Study. J Clin Oncol. 2004;22 (suppl; abstr
1501).
32. Iwamoto FM, Lamborn KR, Kuhn JG, et al. A phase I/II trial of the
histone deacetylase inhibitor romidepsin for adults with recurrent malignant
glioma: North American Brain Tumor Consortium Study 03-03. Neuro Oncol.
2010;13:509-516.
33. Galanis E, Jaeckle KA, Maurer MJ, et al. Phase II trial of vorinostat in
recurrent glioblastoma multiforme: a North Central Cancer Treatment Group
34. Cloughesy TF, Wen PY, Robins HI, et al. Phase II trial of tipifarnib in
patients with recurrent malignant glioma either receiving or not receiving
enzyme-inducing antiepileptic drugs: A North American Brain Tumor Consortium Study. J Clin Oncol. 2006;24:3651-3656.
35. Galanis E, Buckner JC, Maurer MJ, et al. Phase II trial of temsirolimus
(CCI-779) in recurrent glioblastoma multiforme: a North Central Cancer
Treatment Group Study. J Clin Oncol. 2005;23:5294-5304.
36. Chang SM, Wen P, Cloughesy T, et al. Phase II study of CCI-779 in
patients with recurrent glioblastoma multiforme. Invest New Drugs. 2005;23:

357-361.
37. Kreisl TN, Kotliarova S, Butman JA, et al. A phase I/II trial of
enzastaurin in patients with recurrent high-grade gliomas. Neuro Oncol.
2010;12:181-189.
38. Batchelor TT, Duda DG, di Tomaso E, et al. Phase II study of cediranib,
an oral pan-vascular endothelial growth factor receptor tyrosine kinase
inhibitor, in patients with recurrent glioblastoma. J Clin Oncol. 2010;28:
2817-2823.
27:4733-4740.
40. Reardon DA, Fink KL, Mikkelsen T, et al. Randomized phase II study
of cilengitide, an integrin-targeting arginine-glycine-aspartic acid peptide, in
recurrent glioblastoma multiforme. J Clin Oncol. 2008;26:5610-5617.
41. Gilbert MR, Kuhn J, Lamborn KR, et al. Cilengitide in patients with
recurrent glioblastoma: the results of NABTC 03-02, a phase II trial with
measures of treatment delivery. J Neurooncol. 2011;106:147-153.
42. Wen PY, Prados M, Schiff D, et al: Phase II study of XL814 (BMS
907351), an inhibitor of MET, VEGFR2 and RET in patients (pts) with
progressive glioblastoma (GB). J Clin Oncol. 2010;28 (suppl; abstr 2006).
117
IMAGING IN NEURO-ONCOLOGY: PRACTICAL

PRIMER FOR THE PRACTICING PHYSICIAN
CHAIR
Susan M. Chang, MD
San Francisco, CA
SPEAKERS
Whitney B. Pope, MD, PhD
University of California, Los Angeles
Los Angeles, CA
Andrew D. Norden, MD, MPH
Boston, MA
Current Concepts in Brain Tumor Imaging

By Andrew D. Norden, MD, MPH, Whitney B. Pope, MD, PhD, and Susan M. Chang, MD
Overview: Magnetic resonance imaging (MRI) is the most

useful imaging tool in the evaluation of patients with brain
tumors. Most information is supplied by standard anatomic
images that were developed in the 1980s and 1990s. More
recently, functional imaging including diffusion and perfusion
MRI has been investigated as a way to generate predictive and
prognostic biomarkers for high-grade glioma evaluation, but
additional research is needed to establish the added benefits
of these indices to standard MRI. Response critieria for
high-grade gliomas have recently been updated by the Response Assessment in Neuro-Oncology (RANO) working
HE FIRST published magnetic resonance image (MRI)

was from a paper in the journal Nature by Nobel Prize
Laureate Paul Lauterbur in 1973.1 By 1981, magnetic resonance had been used for imaging the brain, demonstrated
the pathologic appearance of glioblastoma (GBM), and
compared favorably to computed tomography (CT) as the
posterior fossa was visualized with substantially less artifact. . . .2 MRI was noted to detect tumors not seen on CT as
early as 1982, and this led quickly to an explosion of articles
evaluating MRI of brain tumors and other intracranial
pathology. T1- and T2-weighted images were quickly adopted as standard imaging along with multiple planar scanning. Gadolinium-based contrast agents were introduced
around 1984,3 as were high-field strength superconducting
(1.5 Tesla) scanners.4 Another advance in brain tumor
imaging occurred in the late 1990s with the introduction of
fluid-attenuated inversion recovery (FLAIR) sequences that
generated strongly T2-weighted images although signal associated with cerebrospinal fluid (CSF) was suppressed.5
Today, MRI remains the imaging modality of choice for
tumor diagnosis, characterization, and assessment of treatment response.
Conventional MRI in Neuro-Oncology
MRI has traditionally been used to evaluate tumor location, size and extent, mass effect, involvement of critical
structures such as adjacent blood vessels, and compromise of
the blood-brain barrier (which results in contrast enhancement). The typical MR scan for a patient with glioma
includes sagittal T1, axial T1, T2, FLAIR and postcontrast
axial and coronal T1-weighted images. Recently pulse sequences sensitive to physiology, rather than just anatomy,
are being more commonly used (see following). Changes in
enhancing tumor size based on bidimensional measurements of postcontrast T1-weighted images are the basis for
both the Macdonald and later RANO criteria for evaluating
tumor response.6 Conversely, nonenhancing tumor is assessed qualitatively in RANO, but not at all in the Macdonald criteria. Nonenhancing tumor is typified by areas of
increased T2 signal intensity associated with mass effect
and architectural distortion such as blurring of the graywhite interface.7 Edema and treatment effect including
gliosis also result in increased T2 signal, which can make
nonenhancing tumor difficult to quantify. FLAIR is more
sensitive to T2 signal abnormalities, as a result of the
nulling of CSF, thereby overcoming the limitation of partial
volume averaging in the cortical and periventricular regions
group. The new criteria account for nonenhancing tumor in

addition to the contrast-enhancing abnormalities on which
older criteria relied. This issue has recently come to the fore
with the introduction of the antiangiogenic agent bevacizumab
into standard treatment for recurrent glioblastoma. Because
of its potent antipermeability effect, contrast enhancement is
markedly reduced in patients who receive bevacizumab. The
RANO criteria also address the phenomenon of pseudoprogression, in which there may be transient MRI worsening of a
glioblastoma following concurrent radiotherapy and temozolomide.
as can be seen in standard T2 images. However, FLAIR also

reduces gray-white differentiation in comparison to typical
T2-weighted images, which can diminish the image readers
ability to distinguish the T2 changes that are a result of
tumor compared with T2 changes that are the result of
edema and/or gliosis. Thus, T2 and FLAIR images can
provide complementary information, and both should be
acquired for evaluation of patients with brain tumors.
Although relying on changes in enhancing tumor previously worked well for evaluating treatment response, the
widespread adoption of bevacizumab therapy for recurrent
GBM highlights the limitations of this approach. This limitation stems largely from bevacizumabs antipermeability
effect. GBMs are characterized by extensive abnormal vasculature with a leaky blood-brain barrier.8 As a result,
contrast material extravasates out of tumor vessels, leading
to increased signal on gadolinium-enhanced T1-weighted
images. Bevacizumab sequesters vascular endothelial
growth factor (VEGF), a potent permeability factor and
promoter of angiogenesis, and thereby acts to diminish
contrast enhancement. Therefore a reduction in contrast
enhancement following bevacizumab infusion may not necessarily reflect a cytotoxic tumor effect. Relying on the
change in contrast enhancement alone can thus misrepresent treatment response, a phenomenon known as pseudoresponse.9,10
Although the RANO criteria include FLAIR or T2 hyperintensity changes as potentially indicative of nonenhancing
tumor progression, no quantification of nonenhancing tumor
is performed. This is because of difficulties in determining
the borders of T2 abnormal regions as well as differentiating
gliosis and other treatment effects from tumor. This has
spurred interest in physiologic imaging as a way of obtaining
quantitative data on tumor burden, although to date, this
goal has not been fully realized.
From Dana-Farber Cancer Institute and Brigham and Womens Hospital, Boston, MA;
David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, CA;
University of California, San Francisco, Department of Neurological Surgery, San Francisco, CA.
Address reprint requests to Susan M. Chang, MD, University of California, San
Francisco, Department of Neurological Surgery, 400 Parnassus Ave., A808, San Francisco,
CA 94143-0372; email: changs@neurosurg.ucsf.edu.
1092-9118/10/1-10
119
NORDEN, POPE, AND CHANG

Functional Imaging in Neuro-Oncology
Magnetic Resonance Perfusion
Given the antiangiogenic effects of bevacizumab, perfusion imaging is intuitively an appealing technique for assessing the effect of drug treatment. Several methods of
obtaining perfusion data have been developed, the two most
common of which are dynamic susceptibility contrast
(DSC) imaging and dynamic contrast enhanced (DCE)
imaging. DSC is used to generate maps of relative cerebral
(or tumor) blood volume (rCBV), and relative cerebral blood
flow (rCBF), among other metrics. DCE is generally used to
measure the permeability constant Ktrans, which is a metric
of capillary leakiness. Many groups have investigated the
role of perfusion imaging in the evaluation of gliomas. For
instance, maximal rCBV has prognostic value in astrocytoma, even when controlling for tumor grade.11 A number of
studies have shown that high rCBV or increasing rCBV is
associated with a worse prognosis across tumor grades.12
Perfusion imaging has been used to assess response to
standard (radiation and cytotoxic chemotherapy) as well as
antiangiogenic treatment. For patients with GBM who receive standard therapy, the percentage change in rCBV from
pre- to post-treatment measurements is predictive of 1-year
survival.13 In patients with recurrent GBM who are treated
with antiangiogenic therapy, a vascular normalization index that is generated by combining Ktrans, microvessel
KEY POINTS
120
Magnetic resonance imaging (MRI) is the imaging

modality of choice to evaluate brain tumor location,
size and extent, mass effect, involvement of critical
structures such as adjacent blood vessels, and compromise of the blood-brain barrier.
Perfusion and diffusion magnetic resonance imaging
tools are examples of physiologic imaging modalities
that may provide quantitative data on tumor burden,
although further investigation is required to define
their roles in routine care and clinical trials.
In approximately 20% to 30% of glioblastoma cases,
the first MRI obtained after radiation therapy and
concurrent temozolomide meets the criteria for progressive disease, but subsequent follow-up scans
show lesion shrinkage or stability. This has been
termed pseudoprogression and is among the most
common causes of misdiagnosed tumor recurrence.
Because of its antipermeability effect, antiangiogenic
therapy with bevacizumab results in marked reduction of tumor enhancement, which reduces the sensitivity of magnetic resonance imaging for diagnosing
tumor recurrence.
Response critieria for high-grade gliomas have recently been updated by the Response Assessment in
Neuro-Oncology working group. The new criteria account for nonenhancing tumor in addition to the
contrast-enhancing abnormalities on which older criteria relied.
volume, and circulating collagen IV is predictive of survival,

even though it is acquired only 1 day after treatment
initiation.14 Yet many studies that assess treatment response to bevacizumab therapy have not been particularly
successful in relating changes in CBV or CBF to outcomes.15,16 Thus challenges continue in the application of
perfusion indices as a biomarker of treatment response in
this setting. The use of perfusion as an early response (1-day
posttreatment initiation marker14) could have a clinical
effect, however.
Perfusion imaging has also been applied to the issue of
differentiating true from pseudoprogression. In general,
true progression has higher perfusion than radiationinduced changes, although there can be overlap in
values.17-20 Standardized protocols may help improve the
use of perfusion imaging in multicenter trials, but overall
accuracy also must be increased to have the desired clinical
affect.
Magnetic Resonance Diffusion
Diffusion-weighting imaging (DWI) is another potential

physiology-based magnetic resonance biomarker for the
evaluation of gliomas. Based on the movement of water
molecules, diffusion data are typically reported as the
apparent diffusion coefficient (ADC); decreased water motion corresponds to lower ADC values. Since water molecules are generally more motion-restricted intracellularly
compared to extracellularly, necrosis and cell death or
lysis can result in increased ADC. Similarly, edema increases ADC by expanding the interstitial, extracellular
fluid volume, which also allows for freer movement of
water molecules. Conversely, lower ADC is associated
with increased cell density.23 Because of these properties,
DWI has the potential to indicate treatment response in
gliomas.
Functional diffusion maps (fDMs) are used to assess
the voxel-wise changes in ADC measured in the same
patient over time.21-24 This can increase the sensitivity in
detecting subtle changes in tumor cell density. Initially
fDMs were used to predict response to cytotoxic chemotherapy and radiotherapy within the contrast-enhancing tumor
bed,22-24 but recent studies have demonstrated their effective use outside regions of contrast enhancement21,25,26 and
as tools for studying the effects of antiangiogenic treatment.27,28
Another approach to using diffusion imaging data is to
construct histograms of ADC values. ADC histogram analysis has been investigated as a predictive biomarker of
bevacizumab treatment response in the setting of recurrent
GBM.29 For this application, tumors with low ADC values
before initiation of bevacizumab were more likely to progress by 6 months compared with tumors with high ADC
values. These results have since been confirmed in a retrospective analysis of a large multicenter trial.30 However,
prospectively validated biomarkers for response to antiangiogenesis treatment are not currently available for clinical
use.
Imaging of Recurrent Gliomas
The Macdonald criteria defined progressive disease, or

recurrence, as . . . 25% increase in size of enhancing
CONCEPTS IN BRAIN TUMOR IMAGING
tumor or any new tumor on CT or MRI scans, or neurologically worse, and steroids stable or increased.31 Although
these were the best available criteria for nearly 20 years,
their focus on enhancing tumor limited their usefulness
because low-grade diffuse gliomas typically lack an enhancing component and high-grade gliomas often contain nonenhancing elements. Furthermore, processes other than tumor
frequently influence the extent of contrast enhancement.
Examples include treatment-induced inflammation or
necrosis, postoperative changes, seizures, and infarction.
Medications may also affect contrast enhancement. Corticosteroids, for example, perhaps the most commonly used class
of medication in neuro-oncology, are known to reduce contrast enhancement in a large portion of patients with gliomas.32
False Positives
A number of phenomena may lead to an inaccurate diagnosis of tumor recurrence on posttreatment MRI scans.
Since publication of a large, randomized trial in 2005,
standard-of-care therapy for GBM includes involved-field
radiation therapy with concurrent and adjuvant temozolomide.33 Most neuro-oncologists obtain an initial posttreatment MRI scan 4 weeks following the radiation therapy
phase. In approximately 20% to 30% of cases, this MRI
meets the criteria for progressive disease, but without
further treatment other than adjuvant temozolomide, subsequent follow-up scans show lesion shrinkage or stability.34,35 This has been termed pseudoprogression and is
among the most common causes of misdiagnosed tumor
recurrence.
Knowledge of this phenomenon has changed practice for
many neuro-oncologists, who now accept the first postradiation MRI scan as a new baseline; so long as a patient is
not highly symptomatic from apparent progression, a
reasonable approach is to proceed with one-to-three cycles
of adjuvant temozolomide before deciding whether the
changes reflect pseudoprogression or true progression. In
general, pseudoprogression that is related to radiation
therapy occurs within 3 months following treatment, but
cases have been described as late as 6 months following
treatment. For patients who develop severe or symptomatic worsening in this window of time, surgery may be
required. Although the pathologic substrate of pseudoprogression remains to be determined with certainty, in one
article seven patients underwent surgical resection for pseudoprogression and subsequent histopathology showed only
necrosis.36
Given an increasing recognition of pseudoprogression in
the temozolomide treatment era and the knowledge that
temozolomide has radiosensitizing properties, many have
suggested that combining temozolomide with radiation therapy increases the risk of pseudoprogression compared with
radiation therapy alone.36 However, the rates of pseudoprogression reported in recent studies34,35 are similar to the
rates reported before the widespread use of temozolomide.37
Another potentially important risk factor for pseudoprogression is the DNA repair gene O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation. Recent data
demonstrate that MGMT promoter methylation status is
predictive of response to temozolomide and a favorable
prognosis in patients with GBM.38 This finding is intuitive
because MGMT is an endogenous DNA repair enzyme that

mediates resistance to alkylating chemotherapeutics like
temozolomide. In a study of 103 patients newly diagnosed
with GBM, pseudoprogression was diagnosed in 32 (31%)
patients. Ninety-one percent of subjects whose tumors had
methylated MGMT promoters developed pseudoprogression
as compared with 41% of patients whose tumors had unmethylated MGMT promoters, and the difference was statistically significant.35 Also of interest is that overall
survival was higher in patients whose tumors developed
pseudoprogression. The findings suggest that pseudoprogression may occur when concurrent radiation therapy and
temozolomide are particularly active against residual tumor, although this remains to be confirmed in prospective
studies.
Pseudoprogression is problematic from the perspective of
clinical trial end points in addition to the clinical challenges
it creates. Accrual of patients whose tumors spontaneously
respond falsely elevates response rates and prolongs
progression-free survival in clinical trials for recurrent disease. For this reason, the recently adopted RANO criteria
(Table 1) stipulate that progression cannot be diagnosed
within the first 12 weeks after radiation therapy unless the
majority of new enhancement is outside the radiation field
or there is histopathologic confirmation of tumor progression.6
Treatments other than radiation therapy are less common
causes of misdiagnosed progression. It is well known, for
example, that enhancement at the margin of an operative
cavity often develops 48 to 72 hours after surgery and may
persist for weeks or months. In some cases, perioperative
infarction may result in new enhancement that is visually
indistinguishable from tumor recurrence. In a study of 50
GBM resections, diffusion-weighted MRI sequences showed
restricted diffusion consistent with infarction in 35 (70%)
cases. Of these, 15 (43%) showed enhancement that might
have been confused with recurrent disease.39 Finally, local
therapies may provoke MRI changes that are suspicious for
recurrence. Examples include carmustine wafers and experimental therapies administered by direct intracerebral injection or convection-enhanced delivery.
False Negatives
As noted above, the Macdonald criteria were not designed

to assess nonenhancing tumors. This is problematic because
the majority of low-grade gliomas and an important minority of high-grade gliomas do not enhance, particularly in
older adults.40 Furthermore, nonenhancing elements are
present in almost all enhancing high-grade gliomas. Another challenge in assessing nonenhancing tumor is that
these abnormalities typically appear hyperintense on T2- or
FLAIR sequences and cannot be readily distinguished from
edema, gliosis, toxic leukoencephalopathy, microangiopathy, or radiation-induced white matter disease.
Unfortunately, the widespread use of antiangiogenic
therapy has only compounded the problem. The prototypic
antiangiogenic agent is bevacizumab, a humanized monoclonal antibody against VEGF that received accelerated
United States Food and Drug Administration approval
for recurrent high-grade glioma in 2009. Although bevacizumab is responsible for high radiographic response rates
and marked prolongation of progression-free survival,41,42
121

Table 1. RANO Criteria for Response Assessment Incorporating MRI and Clinical Factors
Complete response
Partial response
Stable disease
Progressive disease
Requires all of the following: complete disappearance of all enhancing measurable and nonmeasurable disease sustained for at least 4
wk; no new lesions; stable or improved nonenhancing (T2/FLAIR) lesions; patients must be off corticosteroids (or on physiologic
replacement doses only); and stable or improved clinically. Note: Patients with nonmeasurable disease only cannot have a complete
response; the best response possible is stable disease
Requires all of the following: 50% decrease compared with baseline in the sum of products of perpendicular diameters of all
measurable enhancing lesions sustained for at least 4 wk; no progression of nonmeasurable disease; no new lesions; stable or
improved nonenhancing (T2/FLAIR) lesions on same or lower dose of corticosteroids compared with baseline scan; the corticosteroid
dose at the time of the scan evaluation should be no greater than the dose at time of baseline scan; and stable or improved clinically.
Note: Patients with nonmeasurable disease only cannot have a partial response; the best response possible is stable disease
Requires all of the following: does not qualify for complete response, partial response, or progression; stable nonenhancing (T2/FLAIR)
lesions on same or lower dose of corticosteroids compared with baseline scan. In the event that the corticosteroid dose was increased
for new symptoms and signs without confirmation of disease progression on neuroimaging, and subsequent follow-up imaging shows
that this increase in corticosteroids was required because of disease progression, the last scan considered to show stable disease will
be the scan obtained when the corticosteroid dose was equivalent to the baseline dose
Defined by any of the following: 25% increase in sum of the products of perpendicular diameters of enhancing lesions compared with
the smallest tumor measurement obtained either at baseline (if no decrease) or best response, on stable or increasing doses of
corticosteroids*; significant increase in T2/FLAIR nonenhancing lesion on stable or increasing doses of corticosteroids compared with
baseline scan or best response after initiation of therapy* not caused by comorbid events (eg, radiation therapy, demyelination,
ischemic injury, infection, seizures, postoperative changes, or other treatment effects); any new lesion; clear clinical deterioration not
attributable to other causes apart from the tumor (eg, seizures, medication adverse effects, complications of therapy, cerebrovascular
events, infection, and so on) or changes in corticosteroid dose; failure to return for evaluation as a result of death or deteriorating
condition; or clear progression of nonmeasurable disease
Abbreviations: MRI, magnetic resonance imaging; FLAIR, fluid-attenuated inversion recovery.

All measurable and nonmeasurable lesions must be assessed using the same techniques as at baseline.
* Stable doses of corticosteroids include patients not on corticosteroids.
Reprinted with permission. 6 2010 by American Society of Clinical Oncology.
the results of ongoing studies are needed to determine

whether it improves overall survival. Some data suggest
that anti-VEGF therapies may prolong survival by virtue of
potently reducing vascular permeability.43 This is a wellknown effect of antiangiogenic treatment that likely accounts for the reduction in contrast enhancement typically
seen within days of the first dose of bevacizumab or similar
agents.44
Although antiangiogenic drugs are generally welltolerated and lack the toxicities that are typical of cytotoxic
agents, substantial preclinical data obtained since the late
1990s suggest that blocking VEGF signaling promotes an
infiltrative tumor growth pattern based on co-option of
existing cerebral vasculature.45-48 Several uncontrolled, retrospective clinical studies reported that high-grade gliomas
treated with bevacizumab showed increased infiltrative tumor growth on T2/FLAIR sequences.49-52 Indeed, patients
for whom there is MRI evidence of prolonged disease control
on bevacizumab sometimes experience clinical progression
that seems consistent with this phenomenon. In a small
study that pathologically characterized the areas of nonenhancing, infiltrative tumor, there were thin-walled, relatively normal-appearing blood vessels and elevated levels of
insulin-like growth factor binding protein-2 and matrix
metalloprotease-2.53
Some recent data challenge the assumption that antiangiogenic therapy promotes an infiltrative growth pattern. In
one study, patients treated with bevacizumab-treated were
compared with matched pairs of patients treated with conventional therapies.54 The rates of distant or diffuse recurrence were approximately 20% in both groups with no
significant difference detected. Another study examined preand posttreatment recurrence patterns in patients with
recurrent GBM treated with bevacizumab or bevacizumab/
irinotecan on the BRAIN trial.55 Among patients treated
122
with bevacizumab alone, 16% experienced a change in recurrence pattern from local to diffuse. For reasons that are
unclear, a higher proportion (39%) of patients treated with
bevacizumab and irinotecan experienced this change. In a
retrospective report that examined recurrence patterns in
80 patients with GBM who were treated with bevacizumab,
there was no significant difference in recurrence pattern
after bevacizumab treatmentas compared with before bevacizumab treatment.56 Approximately 70% to 80% of recurrences were described as local, which is consistent with older
data.57
Whether antiangiogenic therapy actually promotes infiltrative tumor growth or merely unmasks it is an unanswered question. By controlling peritumoral edema and
durably reducing contrast enhancement, bevacizumab may
lead to the false impression of increased nonenhancing
tumor progression compared with the prebevacizumab era.
Further data are required to address this. Regardless, there
is no question that bevacizumab therapy renders the radiological diagnosis of progressive disease more challenging
than before. There does appear to be a subset of patients
with recurrent high-grade glioma whose tumors progress
primarily in a nonenhancing fashion who would be missed
when the conventional Macdonald criteria is applied. The
RANO criteria for progressive disease therefore include
. . . significant increase in T2/FLAIR nonenhancing lesion
on stable or increasing doses of corticosteroids not caused by
comorbid events (eg, radiation therapy, demyelination, ischemic injury, infection, seizures, postoperative changes, or
other treatment effects).6 Although the ability to determine
the etiology of T2/FLAIR changes is challenging, these
criteria represent an important step toward optimally assessing progressive disease in the current era of antiangiogenic therapy.
CONCEPTS IN BRAIN TUMOR IMAGING
Author
Andrew D. Norden*
Whitney B. Pope
Susan M. Chang
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Genentech
Novartis;
Schering-Plough
REFERENCES
1. Lauterbur PC. Image formation by induced local interactions. Examples
employing nuclear magnetic resonance. 1973. Clin Orthop Relat Res. 1989;
3-6.
2. Doyle FH, Gore JC, Pennock JM, et al. Imaging of the brain by nuclear
magnetic resonance. Lancet. 1981;2:53-57.
3. Carr DH, Gadian DG. Contrast agents in magnetic resonance imaging.
Clin Radiol. 1985;36:561-568.
4. Bilaniuk LT, Zimmerman RA, Wehrli FW, et al. Cerebral magnetic
resonance: comparison of high and low field strength imaging. Radiology.
1984;153:409-414.
5. Kates R, Atkinson D, Brant-Zawadzki M. Fluid-attenuated inversion
recovery (FLAIR): clinical prospectus of current and future applications. Top
Magn Reson Imaging 1996;8:389-396.
6. Wen PY, Macdonald DR, Reardon DA, et al. Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology
working group. J Clin Oncol. 2010;28:1963-1972.
7. Pope WB, Sayre J, Perlina A, et al. MR imaging correlates of survival in
patients with high-grade gliomas. AJNR Am J Neuroradiol. 2005;26:24662474.
8. Rees JH, Smirniotopoulos JG, Jones RV, et al. Glioblastoma multiforme:
radiologic-pathologic correlation. Radiographics. 1996;16:1413-1438; quiz
1462-1463.
9. Clarke JL, Chang S. Pseudoprogression and pseudoresponse: challenges
in brain tumor imaging. Curr Neurol Neurosci Rep. 2009;9:241-246.
10. Brandsma D, van den Bent MJ. Pseudoprogression and pseudoresponse in the treatment of gliomas. Curr Opin Neurol. 2009;22:633-638.
11. Hirai T, Murakami R, Nakamura H, et al. Prognostic value of perfusion
MR imaging of high-grade astrocytomas: long-term follow-up study. AJNR
Am J Neuroradiol. 2008;29:1505-1510.
12. Law M, Young RJ, Babb JS, et al. Gliomas: predicting time to progression or survival with cerebral blood volume measurements at dynamic
susceptibility-weighted contrast-enhanced perfusion MR imaging. Radiology.
2008;247:490-498.
13. Mangla R, Singh G, Ziegelitz D, et al. Changes in relative cerebral blood
volume 1 month after radiation-temozolomide therapy can help predict
overall survival in patients with glioblastoma. Radiology. 2010;256:575-584.
14. Sorensen AG, Batchelor TT, Zhang WT, et al. A vascular normalization index as potential mechanistic biomarker to predict survival after a
single dose of cediranib in recurrent glioblastoma patients. Cancer Res.
2009;69:5296-5300.
15. Bidault F, Sahnoun, M., Rousseau, V., et al.: High-Grade Gliomas
Treated with Bevacizumab: Assessment of Tumor Response with Dynamic
Susceptibility Contrast Perfusion MRI (DSC-MRI) presented at the Radiological Society of North America Annual Meeting, Chicago, IL., 2011.
16. Sawlani RN, Raizer J, Horowitz SW, et al. Glioblastoma: A method for
predicting response to antiangiogenic chemotherapy by using MR perfusion
imaging-pilot study. Radiology. 2010;255:622-8.
17. Barajas RF Jr, Chang JS, Segal MR, et al. Differentiation of recurrent
glioblastoma multiforme from radiation necrosis after external beam radiation therapy with dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging. Radiology. 2009;253:486-496.
18. Hu LS, Baxter LC, Smith KA, et al. Relative cerebral blood volume
values to differentiate high-grade glioma recurrence from posttreatment
radiation effect: direct correlation between image-guided tissue histopathology and localized dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging measurements. AJNR Am J Neuroradiol. 2009;30:552-558.
19. Sugahara T, Korogi Y, Tomiguchi S, et al. Posttherapeutic intraaxial
brain tumor: the value of perfusion-sensitive contrast-enhanced MR imaging
for differentiating tumor recurrence from nonneoplastic contrast-enhancing
tissue. AJNR Am J Neuroradiol. 2000;21:901-909.
20. Lev MH, Ozsunar Y, Henson JW, et al. Glial tumor grading and
outcome prediction using dynamic spin-echo MR susceptibility mapping
compared with conventional contrast-enhanced MR: confounding effect of
elevated rCBV of oligodendrogliomas [corrected]. AJNR Am J Neuroradiol.

2004;25:214-221.
21. Ellingson BM, Malkin MG, Rand SD, et al. Validation of functional
diffusion maps (fDMs) as a biomarker for human glioma cellularity. J Magn
Reson Imaging 2010;31:538-548.
22. Moffat BA, Chenevert TL, Meyer CR, et al. The functional diffusion
map: an imaging biomarker for the early prediction of cancer treatment
outcome. Neoplasia. 2006;8:259-267.
23. Hamstra DA, Chenevert TL, Moffat BA, et al. Evaluation of the
functional diffusion map as an early biomarker of time-to-progression and
overall survival in high-grade glioma. Proc Natl Acad Sci U S A. 2005;102:
16759-16764.
24. Hamstra DA, Galban CJ, Meyer CR, et al. Functional diffusion map as
an early imaging biomarker for high-grade glioma: correlation with conventional radiologic response and overall survival. J Clin Oncol. 2008;26:33873394.
25. Ellingson B, Malkin M, Rand S, et al. Functional diffusion maps
applied to FLAIR abnormal areas are valuable for the clinical monitoring of
recurrent brain tumors. Proc Intl Soc Mag Reson Med. 2009;102:17:285.
26. Ellingson BM, Rand SD, Malkin MG, et al. Utility of functional
diffusion maps to monitor a patient diagnosed with gliomatosis cerebri.
J Neurooncol. 2010;97:419-423.
27. Ellingson BM, Malkin MG, Rand SD, et al. Comparison of cytotoxic and
anti-angiogenic treatment responses using functional diffusion maps in
FLAIR abnormal regions. Proc Intl Soc Mag Reson Med. 2009;102:17:1010.
28. Ellingson BM, Malkin MG, Rand SD, et al. Volumetric analysis of
functional diffusion maps is a predictive imaging biomarker for cytotoxic and
anti-angiogenic treatments in malignant gliomas. J Neurooncol. 2011:102;95103.
29. Pope WB, Kim HJ, Huo J, et al. Recurrent glioblastoma multiforme:
ADC histogram analysis predicts response to bevacizumab treatment. Radiology. 2009;252:182-189.
30. Pope WB, et al. Title forthcoming. J Neurooncol. In press.
31. Macdonald DR, Cascino TL, Schold SC Jr, et al. Response criteria for
phase II studies of supratentorial malignant glioma. J Clin Oncol. 1990;8:
1277-1280.
32. Watling CJ, Lee DH, Macdonald DR, et al. Corticosteroid-induced
magnetic resonance imaging changes in patients with recurrent malignant
glioma. J Clin Oncol. 1994;12:1886-1889.
2005;352:987-996.
34. Brandes AA, Franceschi E, Tosoni A, et al. MGMT promoter methylation status can predict the incidence and outcome of pseudoprogression after
concomitant radiochemotherapy in newly diagnosed glioblastoma patients.
J Clin Oncol. 2008;26:2192-2197.
35. Taal W, Brandsma D, de Bruin HG, et al. Incidence of early pseudoprogression in a cohort of malignant glioma patients treated with chemoirradiation with temozolomide. Cancer. 2008;113:405-410.
36. Chamberlain MC, Glantz MJ, Chalmers L, et al. Early necrosis following concurrent Temodar and radiotherapy in patients with glioblastoma.
J Neurooncol. 2007;82:81-83.
37. de Wit MC, de Bruin HG, Eijkenboom W, et al. Immediate postradiotherapy changes in malignant glioma can mimic tumor progression.
Neurology. 2004;63:535-537.
38. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and
benefit from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-1003.
39. Ulmer S, Braga TA, Barker FG 2nd, et al. Clinical and radiographic
features of peritumoral infarction following resection of glioblastoma. Neurology. 2006;67:1668-1670.
40. Scott JN, Brasher PM, Sevick RJ, et al. How often are nonenhancing
supratentorial gliomas malignant? A population study. Neurology. 2002;59:
947-949.
123

27:4733-4740.
42. Kreisl TN, Kim L, Moore K, et al. Phase II trial of single-agent
bevacizumab followed by bevacizumab plus irinotecan at tumor progression
in recurrent glioblastoma. J Clin Oncol. 2009;27:740-745.
43. Kamoun WS, Ley CD, Farrar CT, et al. Edema control by cediranib, a
vascular endothelial growth factor receptor-targeted kinase inhibitor, prolongs survival despite persistent brain tumor growth in mice. J Clin Oncol.
2009;27:2542-2552.
44. Pope WB, Lai A, Nghiemphu P, et al. MRI in patients with high-grade
gliomas treated with bevacizumab and chemotherapy. Neurology. 2006;66:
1258-1260.
45. Ebos JM, Lee CR, Cruz-Munoz W, et al. Accelerated metastasis after
short-term treatment with a potent inhibitor of tumor angiogenesis. Cancer
Cell. 2009;15:232-239.
46. Keunen O, Johansson M, Oudin A, et al. Anti-VEGF treatment reduces
blood supply and increases tumor cell invasion in glioblastoma. Proc Natl
Acad Sci U S A. 2011;108:3749-3754.
47. Lamszus K, Kunkel P, Westphal M. Invasion as limitation to antiangiogenic glioma therapy. Acta Neurochir Suppl. 2003;88:169-177.
48. Pa`ez-Ribes M, Allen E, Hudock J, et al. Antiangiogenic therapy elicits
malignant progression of tumors to increased local invasion and distant
metastasis. Cancer Cell. 2009;15:220-231.
124
49. Fischer I, Cunliffe CH, Bollo RJ, et al. High-grade glioma before and
after treatment with radiation and Avastin: initial observations. Neuro Oncol.
2008;10:700-708.
after bevacizumab failure in recurrent glioblastoma. Neurology. 2009;73:
1200-1206.
51. Norden AD, Young GS, Setayesh K, et al. Bevacizumab for recurrent
malignant gliomas: efficacy, toxicity, and patterns of recurrence. Neurology.
2008;70:779-787.
52. Narayana A, Kunnakkat SD, Medabalmi P, et al. Change in pattern of
relapse after antiangiogenic therapy in high-grade glioma. Int J Radiat Oncol
Biol Phys. 2012;82:77-82.
53. de Groot JF, Fuller G, Kumar AJ, et al. Tumor invasion after treatment
of glioblastoma with bevacizumab: radiographic and pathologic correlation in
humans and mice. Neuro Oncol. 2010;12:233-242.
54. Wick A, Dorner N, Schafer N, et al. Bevacizumab does not increase the
risk of remote relapse in malignant glioma. Ann Neurol. 2011;69:586-592.
55. Pope WB, Xia Q, Paton VE, et al. Patterns of progression in patients
with recurrent glioblastoma treated with bevacizumab. Neurology. 2011;76:
432-437.
56. Chamberlain MC. Radiographic patterns of relapse in glioblastoma.
J Neurooncol. 2011;101:319-323.
57. Hochberg FH, Pruitt A. Assumptions in the radiotherapy of glioblastoma. Neurology. 1980;30:907-911.
PERSPECTIVES ON HEADLINE-MAKING NEWS IN

NEURO-ONCOLOGY
CHAIR
Timothy F. Cloughesy, MD
Los Angeles, CA
SPEAKERS
E. Antonia Chiocca, MD, PhD
The Ohio State University Medical Center
Columbus, OH
Philip H. Gutin, MD
New York, NY
Noninvasive Application of Alternating

Electric Fields in Glioblastoma: A Fourth
Cancer Treatment Modality
By Philip H. Gutin, MD, and Eric T. Wong, MD
Overview: Tumor treating fields (TTF) therapy is a novel

antimitotic, electric field based treatment for cancer. This
nonchemical, nonablative treatment is unlike any of the established cancer treatment modalities, such as surgery, radiation, and chemotherapy. Recently, it has entered clinical use
after a decade of intensive translational research. TTF therapy
is delivered to patients by a portable, battery-operated, medical device using noninvasive transducer arrays placed on the
skin surface surrounding the treated tumor. TTF therapy is
HE DEFINITION of the electric field is attributed to

Michael Faraday in the 1820s and was later formulated
by James Clerk Maxwell in his electromagnetic theory in
1865.1 It is a field of electric forces that surround a source
charge. When a test charge is placed within an electric field,
a force acts on it. Negative charges attract positive charges,
while similar signed charges repel each other. As seen in
Fig. 1A, an electric field surrounding a source charge can be
described using diverging lines of force. The closer the test
charge is to the source charge, the closer the lines of force are
to each other, which represents higher field intensity.
To understand the effects of electric fields within cells, it is
important to introduce three definitions. First, electric fields
can be uniform or nonuniform. A uniform electric field is
represented by parallel lines of force (Fig. 1B). A nonuniform
electric field is represented by converging or diverging lines
of force (Fig. 1A and 1D). Second, an electric field can be a
constant field or a time-varying field, resulting in electrostatic or electrodynamic phenomena, respectively. In a constant field, the source charges remain the same over time. A
test charge will move in one direction within a constant
electric field toward the oppositely charged source (Fig. 1B).
In a time-varying or alternating electric field, the charge of
the sources alternates over time (Fig. 1C). Third, the test
charge can be an electric charge or an electric dipole (an
element with a positive charge on one end and a negative
charge on the opposite end). An electric charge will move
back and forth, while a dipole will rotate within an alternating uniform electric field and align with the direction of the
field. In a nonuniform converging electric field, both dipoles
and charges move in the direction of the higher field intensity through a process known as dielectrophoresis (Fig. 1D).
Mechanism of Action of TTF Therapy
Over 100 years after Maxwells original publication,

Yoram Palti, MD, PhD, hypothesized that properly tuned
alternating electric fields at physiological intensities (i.e.,
13 V/cm) would disrupt the mitotic process of dividing
cancer cells.2,3 Dr. Palti hypothesized and subsequently
demonstrated in vitro that at frequencies between 100 and
300 kHz, alternating electric fields disrupt the formation of
the mitotic spindle during metaphase and lead to dielectrophoretic movement of charged and/or polar molecules and
organelles during anaphase and telophase, disrupting normal cytokinesis and leading to apoptosis.2,3 According to this
model, the first mechanism of action is explained by the fact
126
now a U.S. Food and Drug Administration (FDA)approved

treatment for patients with recurrent glioblastoma (GBM) who
have exhausted surgical and radiation treatments. This article
will introduce the basic science behind TTF therapy, its
mechanism of action, the preclinical findings that led to its
clinical testing, and the clinical safety and efficacy data
available to date, as well as offer future research directions on
this novel treatment modality for cancer.
that the tubulin subunits are one of the most polar molecules in the cell. These tubulin subunits align in the direction of the applied electric field (Fig. 2A), interfering with
the normal polymerization of the mitotic spindle, which
results in formation of abnormal mitotic figures in vitro.3
The second mechanism of action is explained by examining
the change in shape of the electric field within a dividing cell
from anaphase to telophase. When the cell division axis is
aligned with the direction of the electric field, the field lines
that enter the cell at one end converge at the cytokinetic
furrow between the developing daughter cells and then
diverge on the opposite side (Fig. 2B). This nonuniform
electric field within the cell generates dielectrophoretic
forces that act on polar and charged elements in the cell,
pushing them toward the cytokinetic furrow leading to
violent blebbing of the plasma membrane.3 This finding was
also validated by researchers from Beth Israel Deaconess
Medical Center and may be mediated by improper placement of the contractile elements that form the cytokinetic
ring on anaphase entry.4
Preclinical Studies of the Antitumor Effects of
TTF Therapy
Between 2004 and 2010, a series of publications and

conference presentations addressed the issue of the applicability range of TTF therapy to different in vitro and in vivo
cancer models either alone or in combination with standard
chemotherapy.3,5-8 Tables 1 and 2 summarize the state-ofthe-art preclinical research with TTF therapy. TTF therapy
has been shown to effectively inhibit cancer cell growth in
various cell lines in vitro (Table 1). This effect was clearly
dose (field intensity) dependent in the range of 1 to 3 V/cm.5
The optimal frequency for the inhibitory effect of TTF
therapy differed between cell types and was inversely related to cell size (Table 1; e.g., glioma cell cultures at 200
kHz3,5). In addition, based on the directional nature of TTF
From the Department of Neurosurgery, Memorial Sloan-Kettering Cancer Center Brain

Tumor Center, New York, NY; and Brain Tumor Center and Neuro-Oncology Unit, Beth
Israel Deaconess Medical Center, Boston, MA.
Address reprint requests to Philip H. Gutin, MD, Department of Neurosurgery, C-703,
Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY 10065; email:
gutinp@mskcc.org.
1092-9118/10/1-10
TTF THERAPY IN GLIOBLASTOMA
Fig. 1. Electric field theory. (A) Opposite charges attract. (B) A constant, uniform, electric field. (C) Charges
and dipoles in a time-varying, uniform electric field. (D) A
dipole in a time-varying, nonuniform electric field (dielectrophoresis).
therapy, its antimitotic effect in cultures was enhanced by

sequentially applying more than one field direction to the
treated cells.5 The combination of TTF therapy with different chemotherapeutic agents has been shown to have at
least additive if not synergistic effects.7,9 Specifically, the
combination of TTF therapy with temozolomide in glioma
cell lines was shown to be additive. Interestingly, in breast
cancer cells, TTF therapy showed overt synergism with
taxanes (e.g., paclitaxel), probably a result of the temporal
KEY POINTS
Tumor treating fields (TTF) therapy is an emerging,

low-toxicity treatment modality for solid tumors
based on the delivery of antimitotic alternating electric fields to the tumor, which interfere with cytokinesis and microtubule assembly that eventually lead
to cell death.
As a monotherapy, TTF therapy is at least as effective as currently available active chemotherapy and
biologic therapies for the treatment of recurrent glioblastoma (GBM).
The efficacy of this noninvasive treatment modality is
achieved with significantly less toxicity and a better
quality of life compared with chemotherapy.
Preliminary data suggest TTF therapy acts synergistically with temozolomide and other chemotherapy in
both preclinical and clinical trials.
Future research should focus on integrating TTF
therapy into the treatment of GBM in the adjuvant
and maintenance settings, as well as in the treatment
of other solid tumor malignancies.
proximity of taxanes effect in metaphase and TTF therapys

mitotic interference on cell entry into anaphase.5
TTF therapy has been tested in numerous in vivo cancer
models (Table 2).3,5,8,10 Noninvasive application of TTF
therapy to animals was performed using electrically insulated transducer arrays placed on the head or torso surrounding the region of the tumor. Inhibition of tumor growth
was seen in each of these models when the correct frequency
of TTF therapy was applied. Specifically, 200 kHz TTF
therapy applied in two sequential and perpendicular field
directions lead to significant (p 0.01) inhibition of a
syngeneic, orthotopic F-98 glioma in rats after 7 days of
treatment.5 An additional syngeneic, orthotopic model of
non-small cell lung cancer in mice showed that 150 kHz TTF
therapy significantly (p 0.01) inhibited tumor growth
within 7 days of treatment.8,11 Furthermore, the additive
effect of TTF therapy with chemotherapy seen in vitro was
recapitulated in different in vivo models.5,8 Finally, in a
metastatic tumor model using a squamous carcinoma tumor
implanted in the kidney capsule of rabbits, TTF therapy
applied to the abdomen blocked metastatic spread of tumor
from the kidney to the lungs.10,27
Translating TTF Therapy into Clinical Use
Since TTF therapy is a physical antimitotic modality with

no half-life, its application should be continuous. Kinetic
modeling was used to predict the minimal treatment duration needed with TTF therapy.12 Based on these data, a
minimal treatment course of 4 weeks was defined and
implemented in clinical studies. In vivo animal experiments
and pilot clinical data subsequently verified the 4-week
minimal treatment duration.12 Such continuous delivery
was made possible by the development of a portable, batteryoperated, medical device that patients can use at home
(NovoTTF-100A, Novocure, Haifa, Israel). Finally, extensive
toxicity studies of TTF therapy were performed in healthy
127
GUTIN AND WONG
Fig. 2. Effects of tumor treating fields therapy on

intracellular structures during mitosis. (A) During metaphase, tubulin dimers align with the external electric
field, interfering with the formation of the mitotic spindle. (B) During cytokinesis, the nonuniform electric field
formed within the dividing cell drives charged and polar
macro-molecules and organelles toward the cleavage
furrow.
mice, rats, and rabbits.5,9 Clinical, laboratory, and pathologic analyses showed that TTF therapy is well tolerated and
does not lead to systemic toxicity in animals. As expected by
the frequency range of TTF therapy (100 300 kHz), these
electric fields do not have any effect on excitable tissues
(neural, muscular, or cardiac), nor do they cause significant
heating.13-15
Clinical Testing of TTF Therapy as a Monotherapy
The NovoTTF device was first applied to patients in a

small feasibility trial in Switzerland in 2003.16 In 2004, TTF
therapy was tested in a pilot clinical trial in patients with
recurrent GBM (Table 3).5 This single-center, single-arm
trial included patients with favorable prognostic character-
Table 1. In Vitro Evidence Overview

Histology
High-grade glioma
Breast adenocarcinoma
Non-small cell lung cancer (adenocarcinoma)
Colorectal adenocarcinoma
Malignant melanoma
Prostate
Cervical cancer
Cell Line
F-98; C-6; RG-2

U-118; U-87
Normal:
MDA-MB-231
MCF7
Multiple drug resistant:
MDA-MB-231Dox
AA8/EmtR1
MCF7/Mx
H1299
LLC
CT-26
B16F1 Patricia
PC-3
HeLa
Optimal/Effective
TTF Frequency (kHz)
Additive/Synergistic
with Chemotherapy
200
Temozolomide (dacarbazine)
120
Cyclophosphamide
Can Res, 20043

Proc Natl Acad Sci U S A, 20075
Can Res, 20043
120
Doxorubicin
Paclitaxel
Neuro Oncol, 20114

BMC Cancer, 20107
150
100*
100
100*
200*
Abbreviations: TTF, tumor treating fields; NA, not available (was not reported by the authors).
* Effect seen at this frequency; additional frequencies were not tested.
128
Doxorubicin
Paclitaxel
Paclitaxel
Pemetrexed
NA
NA
NA
NA
Reference
ERS, 20108
AACR, 20076
Can Res, 20043
Can Res, 20043
Can Res, 20043
Can Res, 20043
Neuro Oncol, 20114

Table 2. In Vivo Evidence Overview
Tumor Type
Anatomic Location
Animal Model
Frequency (kHz)
GBM
Non-small cell lung cancer
Right hemisphere
Lung parenchyma
Rat
Mouse
200
150
Malignant melanoma
Intradermal
Mouse
100
Malignant melanoma
VX-2 (anaplastic)
Intravenous
Kidney capsule
Mouse
Rabbit
100
150200
Effect of TTF
References
Tumor growth inhibition with 2 and 3 field directions

1. Tumor growth inhibition with 2 field directions
2. Additive tumor inhibition with pemetrexed
Tumor growth inhibition with 1 and 2 field directions
Inhibition of metastatic seeding in the lungs
1. Tumor growth inhibition seen with 2 field directions
2. Increase in median survival
3. Inhibition of metastatic seeding in the lungs
4. Additive tumor inhibition with paclitaxel

ERS, 20108
Can Res, 20043
Clin Exp Metastasis, 200910
Clin Exp Metastasis, 200910
AACR, 200927
Neuro Oncol, 201012
Abbreviation: GBM, glioblastoma.
ASCO Annual Meeting and were updated at the 2011

Society for Neuro-Oncology (SNO) Annual Meeting.18,19
Baseline characteristics of patients were balanced between
the two treatment groups. In both groups, patients had poor
prognostic predictors compared with previous clinical trials
of recurrent GBM (90% of patients were at their second or
subsequent recurrence; 20% had failed bevacizumab before
entering the trial; and the average tumor diameter was
above 5 cm). In the conservative intent-to-treat (ITT) analysis, the study showed that patients with recurrent GBM
treated with NovoTTF alone had comparable OS to that of
patients who received chemotherapy and/or bevacizumab
(6.6 months vs. 6.0 months, respectively; p 0.26; hazard
ratio [HR] 0.86; Table 3). Although NovoTTF did not show
superiority over active chemotherapies, it was clear that it
was at least as effective as these treatments. Secondary
endpoints in the trial were supportive: blinded radiology
review showed that PFS at 6 months was 21.4% in the
NovoTTF group compared with 15.2% in the chemotherapy
group (p 0.24). There were more radiological responses
seen in the NovoTTF group compared with the chemotherapy group (12% vs. 6%, respectively; p 0.07), including
istics. Treatment with the device was well tolerated, and no

treatment-related serious adverse events were reported.
Most patients developed grade 1 to 2 contact dermatitis
beneath the transducer arrays on the scalp. Efficacy endpoints were very encouraging with a 20% objective response
rate, progression-free survival (PFS) at 6 months of 50%,
median time to progression (TTP) of 26 weeks, and median
overall survival (OS) of 62.2 weeks (14.4 months). Compared
to the historic results of salvage chemotherapy, these results
showed clear activity of TTF therapy when used as a
monotherapy in recurrent GBM.17
Based on the results of this pilot trial, a pivotal phase III,
multicenter, randomized (1:1) clinical study was initiated in
patients with recurrent GBM (Table 3). The randomized
study, which recruited 237 patients between 2006 and 2009,
compared the efficacy and safety of monotherapy with the
NovoTTF device to that of the best available active chemotherapy according to physicians choice. Thirty-six patients
received bevacizumab, 36 received nitrosureas, 12 received
temozolomide, and 33 received other agents. This was the
largest randomized study in recurrent GBM to be completed
to date. The results of the study were presented at the 2010
Table 3. Clinical Evidence Overview
Indication (Analysis Group)
Trial Phase
(# of Subjects)
Analysis
Phase I-II (n 10)

ITT Analysis
Recurrent GBM (at second and
Phase III (n 237)
ITT analysis
fourth relapse)
Recurrent GBM (treated patients only)
Phase III (n 210)
PP Analysis
Recurrent GBM (KPS 80, age 61)
Phase III (n 110)
Subgroup analysis
Recurrent GBM (after bevacizumab failure) Phase III (n 43)
Subgroup analysis
Recurrent GBM (TTF versus bevacizumab) Phase III (n 156)
Subgroup analysis
Newly diagnosed GBM
I-II (n 10)
ITT Analysis
(together with temozolomide)
Relapsed advanced NSCLC
I-II (n 42)
ITT Analysis
(together with pemetrexed)
Recurrent GBM (at first relapse)
Overall Survival
(Months)
TTF
Hazard
Ratio (p)
Chemo
Progression-Free
Survival (PFS)
at 6 Months or
Median PFS (Weeks)
TTF
Chemo
14.5 m 6.0 m*
Non-randomized 50%
15%*
6.6 m
6.0 m
21.4%
15.2%
7.8 m
6.0 m
26.2%
15.2%
8.8 m
6.6 m
25.6%
7.7%
4.4 m
3.1 m
HR 0.86
(p 0.26)
HR 0.67
(p 0.012)
HR NA
(p 0.01)
(p 0.02)
NA
NA
21%
21%
90%
155 w
28 w
50%*
26 w
12 w*
HR 0.65
(p 0.048)
39 m 14.7 m* (p 0.002)
6.6 m
5.0 m
13.8 m 8.2 m*
NA
P value
References
NA
Proc Natl Acad Sci U S A,

20075
p 0.24 J Clin Oncol, 201018
Neuro Oncol, 201119
p 0.03 J Clin Oncol, 201018
Neuro Oncol, 201119
NA
Neuro Oncol, 201019
NA
Neuro Oncol, 201020
p 0.05 Neuro Oncol, 201121

NA
BMC Med Phys, 20099

ESMO, 201025
ERS, 20108
Expert Opin Investig Drugs,
201011
Abbreviations: GBM, glioblastoma; ITT, intention to treat; NA, not available (was not reported by the authors); HR, hazard ratio; PP, per protocol; KPS, Karnofsky
performance status; TTF, tumor treating fields; NSCLC, non-small cell lung cancer.
* Single-arm trials with literature control.
129
GUTIN AND WONG
three sustained complete responses in the NovoTTF group

compared with none in the chemotherapy group. These
results were accompanied by significantly (p 0.05) less
treatment-related adverse events with NovoTTF compared
with chemotherapy. Patients in the NovoTTF group reported a higher quality of life compared with patients
treated with chemotherapy. This analysis was based on the
European Organisation for Research and Treatment of Cancer QLQ-C30 and mirrored the lack of chemotherapy-related
toxicities in the NovoTTF group. Interestingly, patients in
the NovoTTF group reported better cognitive and emotional
functioning and much less pain than patients in the chemotherapy group, although these domains of the questionnaires are not related to known side effects of chemotherapy.
To date, several exploratory analyses of the study data
have been performed. The first analysis compared patients
who received the same amount of therapy in both groups.
This prospectively defined per-protocol analysis excluded
patients from both groups who received less than one predefined treatment course. The analysis demonstrated superior survival in the NovoTTF group compared with the
chemotherapy group (7.8 months vs. 6.0 months; p 0.012,
HR 0.67).18,19 The rationale behind this analysis is that
TTF is a physical modality with no half-life, so that the
moment the therapy is stopped, its antimitotic effect stops
as well. In contrast, chemotherapies have measurable
plasma and tissue half-life, which results in continued
efficacy and toxicity long after a dose has been given.
Therefore, to achieve pharmacokinetic balance in the
amount of treatment in both groups, this analysis used a
simplified criterion that one course of chemotherapy (e.g., 1
day of carmustine or 5 days of temozolomide) is equivalent to
four weeks of continuous TTF therapy.
Two more analyses of the study data were presented at
the 2010 and 2011 SNO Annual Meetings.20,21 The first
study analyzed known clinical prognostic factors of age and
Karnofsky performance status (KPS). This analysis demonstrated that in patients age 60 and younger with a KPS
greater than 70, treatment with NovoTTF resulted in superior OS compared with chemotherapy (8.8 months vs. 6.6
months; p 0.01). This survival advantage could be attributed to better compliance with TTF therapy in this group of
patients. In support of this finding, a statistically significant
correlation was seen in the NovoTTF group between treatment compliance (as measured by the device computerized
log file) and OS (p 0.0475).
The second analysis is a post hoc, exploratory analysis of
the treatment of 120 patients with NovoTTF compared with
36 patients with bevacizumab. Although without a prespecified analysis in the trial, patients in the study treated with
NovoTTF lived significantly longer than those treated with
bevacizumab (6.6 months vs. 5.0 months, respectively; p
0.048, HR 0.65).21 This analysis included all ITT patients
who received either bevacizumab or NovoTTF. Patient characteristics were almost identical and, in fact, favored the
bevacizumab group prognostically. Clearly, this analysis
cannot be taken as final evidence of superiority of NovoTTF
over bevacizumab; however, it should be treated as
hypothesis-generating data for future clinical studies. Finally, in the 43 patients who entered the study after bevacizumab therapy failure (approximately 20% of patients in
both groups), OS was significantly longer with TTF therapy
130
than with chemotherapy (4.4 months vs. 3.1 months, respectively; p 0.02). The data for the chemotherapy-treated
group is in line with previous publications, which showed
that following bevacizumab failure, the survival of patients
with recurrent GBM is limited.22
Based on the results of this pivotal phase III study, the
FDA approved the NovoTTF-100A device on April 8, 2011,
through the premarket approval (PMA) regulatory pathway.
The PMA pathway is reserved for class III (high-risk)
medical devices and requires preclinical, clinical, and manufacturing evidence, including review of both efficacy and
safety data by a panel of independent experts. The FDA
concluded that the study results showed NovoTTF to be
comparable in efficacy to active chemotherapy, without
many of the side effects associated with chemotherapies and
with a better quality of life.23
Clinical Trials Evaluating TTF Therapy in Combination
with Chemotherapy
Two studies of combined TTF therapy and chemotherapy

have been published to date. The first was a single-arm,
single-center trial performed in 2006 in patients with newly
diagnosed GBM.9 Patients received the Stupp protocol with
TTF therapy added to maintenance temozolomide.24 This
trial showed promising PFS and OS data (PFS 14 months;
OS 39 months; Table 3) and served as the basis for an
ongoing, multicenter, pivotal phase III, randomized clinical
study comparing TTF therapy and temozolomide with temozolomide alone in the maintenance stage of the Stupp
protocol.
The second study tested TTF therapy together with pemetrexed in 42 patients with pretreated, advanced non-small
cell lung cancer.8,11,25 Efficacy and safety with this combined treatment paradigm were promising. Time to local
disease progression in the lungs and liver (where TTF was
applied) was 28 weeks, and OS was 13.8 months. In contrast, TTP and OS for pemetrexed alone were previously
reported to be 12 weeks and 8.3 months, respectively.26
TTF therapy is still in its early days. However, it has an
established mechanism of action, and a growing body of
preclinical evidence has shown its wide applicability in solid
tumor malignancies either alone or in combination with
standard chemotherapies. Objective antitumor activity and
an unprecedented safety profile of this treatment modality
have been seen in patients with recurrent GBM. Although
TTF monotherapy has been shown to be at least as effective
as the best available chemotherapies today for recurrent
GBM, in-depth analysis of the phase III study data identified at least two subgroups where TTF therapy was superior
to chemotherapy and could be offered to patients as an
alternative to chemotherapy: younger patients with a better
functional status and patients in whom bevacizumab treatment has failed in the past.
Conclusion
The approval of TTF therapy for recurrent GBM ushers in

a fourth modality of cancer treatment. More importantly,
TTF treatment has a superior safety profile, and its minor
side effects do not appear to overlap with those of cytotoxic
chemotherapies, targeted agents, or antiangiogenesis drugs.
Therefore, the rational combination of TTF therapy with
specific pharmacologic agents may enhance tumor cell death
because of potential additive or synergistic effects. First, as

demonstrated in preclinical and clinical models, chemotherapy administered together with TTF therapy may result in
additive or synergistic tumor control without increasing
systemic toxicities. Second, TTF treatment could be combined with targeted agents that block survival signaling
within the tumor cell. This block may be sufficiently strong
to enhance the cytotoxic effect of TTF therapy or vice versa.
Third, the combination of TTF and antiangiogenesis agents

may be another promising path that combines different
antitumor treatments to improve tumor control. Lastly, the
proper scheduling of TTF therapy with other agents is
unknown. Additional research may shed light on the optimal
scheduling that may achieve a synergistic effect on tumor
growth leading to long-term tumor control and enhanced
patient survival.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Philip H. Gutin
Novocure
Eric T. Wong
Novocure
Expert
Testimony
Other
Remuneration
Novocure
REFERENCES
1. Maxwell JC. A Dynamical Theory of the Electromagnetic Field. Royal
Society Transactions. CLV:1865.
2. Palti Y, Schneiderman R, Gurvich Z, et al. Cell proliferation arrest and
tumor cell destruction by low intensity, frequency tuned electric fields. 2004
AACR Meeting Abstracts. (abstr 1000).
3. Kirson ED, Gurvich Z, Schneiderman R, et al. Disruption of cancer cell
replication by alternating electric fields. Cancer Res. 2004;64:3288-3295.
4. Lee SX, Wong ET, Swanson KD. Mitosis Interference of Cancer Cells
During Anaphase By Electric Field from NovoTTF-100A. Neuro Oncol.
2011;13:iii10-iii25 (suppl 3; abstr CB-17).
5. Kirson ED, Dbaly V, Tovarys F, et al. Alternating electric fields arrest
cell proliferation in animal tumor models and human brain tumors. Proc Natl
Acad Sci U S A. 2007;104:10152-10157.
6. Schneiderman R, Shmueli E, Kirson E, et al. Synergism between
chemotherapy and alternating electric fields in the inhibition of cancer cell
proliferation in-vitro. 2007 AACR Meeting Abstracts. (abstr 2276).
7. Schneiderman RS, Shmueli E, Kirson ED, et al. TTFields alone and in
combination with chemotherapeutic agents effectively reduce the viability of
MDR cell sub-lines that over-express ABC transporters. BMC Cancer. 2010;
10:229.
8. Weinberg U, Fresard I, Kueng M, et al. An Open Label Pilot Study of
Tumor Treating Fields (TTFields) in Combination with Pemetrexed for
Advanced Non-small Cell Lung Cancer (NSCLC). 2010 ERS Annual Congress.
(abstr 363).
9. Kirson ED, Schneiderman RS, Dbaly V, et al. Chemotherapeutic treatment efficacy and sensitivity are increased by adjuvant alternating electric
fields (TTFields). BMC Med Phys. 2009;9:1.
10. Kirson ED, Giladi M, Gurvich Z, et al. Alternating electric fields
(TTFields) inhibit metastatic spread of solid tumors to the lungs. Clin Exp
Metastasis. 2009;26:633-640.
11. Pless M, Weinberg U. Tumor treating fields: Concept, evidence and
future. Expert Opin Investig Drugs. 2010;20:1099-1106.
12. Kirson ED, Wasserman Y, Izhaki A, et al. Modeling tumor growth
kinetics and its implications for TTFields treatment planning. Neuro Oncol.
2010;12:iv36-iv57 (suppl 4; abstr NO-54).
13. Palti Y. Stimulation of muscles and nerves by means of externally
applied electrodes. Bull Res Counc Isr Sect E Exp Med. 1962;10:54-56.
14. Shizgal P, Mathews G. Electrical stimulation of the rat diencephalon:
Differential effects of interrupted stimulation on on- and off-responding.
Brain Res. 1977;129:319-333.
15. Yearwood TL, Hershey B, Bradley K, et al. Pulse width programming in

spinal cord stimulation: A clinical study. Pain Physician. 2010;13:321-335.
16. Salzberg M, Kirson E, Palti Y, et al. A pilot study with very lowintensity, intermediate-frequency electric fields in patients with locally advanced and/or metastatic solid tumors. Onkologie. 2008;31:362-365.
17. Wong ET, Hess KR, Gleason MJ, et al. Outcomes and prognostic factors
in recurrent glioma patients enrolled onto phase II clinical trials. J Clin
Oncol. 1999;17:2572-2578.
18. Stupp R, Kanner A, Engelhard H, et al. A prospective, randomized,
open-label, phase III clinical trial of NovoTTF-100A versus best standard of
care chemotherapy in patients with recurrent glioblastoma. J Clin Oncol.
2010;28:18s (suppl; abstr LBA2007).
19. Wong ET, Ram Z, Gutin PH, et al. Updated survival data of the phase
III clinical trial of NovoTTF-100A versus best standard chemotherapy for
recurrent glioblastoma. Neuro Oncol. 2011;13:iii85-iii91 (suppl 3; abstr OT09).
20. Ram Z, Gutin PH, Stupp R. Subgroup and quality of life analyses of the
phase III clinical trial of NovoTTF-100A versus best standard chemotherapy
for recurrent glioblastoma. Neuro Oncol. 2010;12:iv36-iv57 (suppl 4; abstr
NO-55).
21. Ram Z, Gutin PH, Wong ET. Comparing the effect of NovoTTF to
Bevacizumab in Recurrent GBM: A Post-Hoc Sub-Analysis of the Phase III
Trial Data. Neuro Oncol. 2011;13:iii41-iii68 (suppl 3; abstr NO-50).
after bevacizumab failure in recurrent glioblastoma. Neurology. 2009;73:
1200-1206.
23. FDA: NovoTTF-100A Information for Use, 2011. http://www.access
data.fda.gov/cdrh_docs/pdf10/P100034c.pdf. Accessed February 23, 2012.
2005;352:987-996.
25. Pless M, Betticher DC, Buess M, et al. A phase II study of tumor
treating fields (TTFields) in combination with pemetrexed for advanced non
small cell lung cancer (NSCLC). Ann Oncol. 2010:viii122-viii161.
26. Hanna N, Shepherd FA, Fossella FV, et al. Randomized phase III trial
of pemetrexed versus docetaxel in patients with non-small-cell lung cancer
previously treated with chemotherapy. J Clin Oncol. 2004;22:1589-1597.
27. Kirson E, Gurvich Z, Izhaki A, et al. Alternating electric fields
(TTFields) inhibit metastatic spread of solid tumors to the lungs in-vivo. 2009
AACR Meeting Abstracts. (abstr 151).
131
ADAPTIVE CLINICAL TRIAL DESIGNS:

WHAT ARE THEY AND SHOULD THEY BE USED?
CHAIR
Mary Weber Redman, PhD
Fred Hutchinson Cancer Research Center
Seattle, WA
SPEAKERS
J. Jack Lee, PhD, DDS
Houston, TX
Marc Buyse, ScD
International Drug Development Institute
Louvain-la-Neuve, Belgium
Limitations of Adaptive Clinical Trials

By Marc Buyse, ScD
Overview: Adaptive designs are aimed at introducing flexibility in clinical research by allowing important characteristics
of a trial to be adapted during the course of the trial based
on data coming from the trial itself. Adaptive designs can be
used in all phases of clinical research, from phase I to phase
III. They tend to be especially useful in early development,
ITH THE large number of promising new molecules

that are currently available for clinical testing, clinical trials must detect a drugs benefit (or harm) as quickly
as possible. In parallel to the explosion in the number of
drugs awaiting clinical testing, the costs of clinical trials
have sky-rocketed, which adds to the pressure of optimizing
trials, to the extent possible, in terms of sample sizes,
timelines, and risk of failure. A new class of designs has
emerged to address these challenges, collectively known as
adaptive designs. In this chapter, we review different types
of adaptive designs and briefly mention some situations in
which such designs can be useful. Much of this chapter,
however, is devoted to a discussion of the limitations and
drawbacks of adaptive designs, which might partly explain
why these designs have not been commonly used and might
in the future have less of an effect on clinical research than
claimed by their advocates.
Types of Adaptive Designs
One of the difficulties surrounding adaptive designs is

that the term is used to encompass different situations. For
clarity, we divide group adaptive designs into three broad
categories.
The first category is treatment effectindependent adaptive designs. In these designs, some of the design features
can be adapted on observation of predefined patient characteristics (such as baseline prognostic factors) or outcomes
(such as response rate or hazard rate overall or in the control
group) but in ignorance of the treatment effect.
The second category is treatment effect dependent adaptive designs. In these designs, one or more of the design
features (such as the sample size, the patient inclusion
criteria, the treatment groups being compared, the treatment allocation ratio, or even the primary endpoint) can be
adapted, depending on the observed treatment effect.
The third category is other types of adaptive designs,
which include the continual reassessment method (CRM) for
phase I trials and seamless phase II/III designs.
Treatment EffectIndependent Adaptive Designs
In these designs, adaptations do not depend on the treatment effect. As such, these adaptations raise few issues and
have little effect on the statistical inference; in particular,
they do not inflate the type I error. In fact, such adaptations are so mild that trials using them are not referred to
as adaptive. We provide two examples but do not discuss
these designs in detail.
Covariate-Adaptive Randomization
One instance of treatment effectindependent adaptation

is covariate-adaptive randomization, for which the probabil-
when the paucity of prior data makes their flexibility a key

benefit. The need for adaptive designs lessened as new
treatments progress to later phases of development, when
emphasis shifts to confirmation of hypotheses using fully
prespecified, well-controlled designs.
ity of allocating the next patient to one of the trials treatment groups is computed dynamically to ensure good
balance among the treatment groups with respect to important prognostic factors (center or country, clinicopathologic
features, and, increasingly, biomarkers measured at baseline). A common implementation of this approach is minimization, for which a predefined algorithm is used to minimize
the imbalance between the distributions of important prognostic factors at baseline among treatment groups. When
minimization is used, the treatment group the next patient
is allocated to can depend on the baseline characteristics of
previously accrued patients but not on their outcome.1
Sample Size Increases
Another type of treatment effectindependent adaptation

consists of a sample size increase if the incidence of the event
of interest is much lower than expected in the control group
(to preserve the power of the trial) or if the event rate is
much lower than expected overall (to preserve the timelines
of event-driven analyses when the outcome of interest is a
time-to-event, such as disease-free survival or overall survival). Again, these sample size increases are implemented
in ignorance of the treatment effect; hence, they generally
have no effect on type I error and, if implemented appropriately, raise no special statistical concerns.2
Treatment EffectDependent Adaptive Designs
These designs are truly adaptive insofar as adaptations

depend on the observed treatment effect, which requires
caution to be exercised and a proper statistical approach to
be used. If, for instance, the sample size of a trial was
increased (or decreased) simply because the observed treatment effect was smaller (or larger) than anticipated, the
final results of the trial could be biased.3 For instance, a
randomly large treatment effect could lead to a reduction
in sample size even though the true effect were as expected.
Note that a group sequential design is not subject to this
problem because its sample size is fixed and can only be
decreased if the trial is stopped for efficacy or futility at an
interim analysis.4
From the International Drug Development Institute, Houston, TX; Interuniversity Institute for Biostatistics and Statistical Bioinformatics, Hasselt University, Diepenbeek,
Belgium.
Address reprint requests to Marc Buyse, ScD, IDDI Inc, 363 N. Sam Houston Pkwy. E.,
Suite 1100, Houston, TX 77060; email: marc.buyse@iddi.com.
1092-9118/10/1-10
133
MARC BUYSE
Sample Size Recalculation
The most obvious adaptation to consider for an ongoing

comparative (phase III) trial is a sample size recalculation if
the treatment effect turns out to be vastly different from
that assumed to design the trial. It might make sense, for
example, to consider an increase in sample size if the
assumed treatment effect was grossly overestimated (e.g., as
a result of highly promising phase II results), but there are
serious theoretical and practical objections to doing so (Table 1).
Theoretically, the treatment effect assumed in the protocol was the smallest effect considered to be clinically worthwhile; hence, there is no reason to lower this effect merely to
reach statistical significance. In practice, however, the sample size is often based on a larger treatment effect than the
minimum considered worthwhile, especially when the findings of phase II trials suggest that a larger treatment effect
can potentially be achieved. Commonly, other considerations
come into play, including budgetary constraints that tend to
drive sample sizes down. It is tempting, especially for small
companies with limited financial resources, to shoot for large
treatment effects, hope for the best, and increase the sample
size only if needed. Great caution must, however, be exercised when a sample size increase is triggered by the
observation of a smaller than expected treatment benefit,
insofar as such an adaptation can lead to changes in the
types of patients accrued into the trial so that the trial after
the adaptation is no longer the same as before the adaptation. It has also been proven that for any adaptive design one
might consider, there exists a group sequential design that
uniformly outperforms it (i.e., has better power for any given
KEY POINTS
134
Adaptive designs can be useful in specific situations

(e.g., in phase I trials aimed at determining a maximum tolerated dose or in dose-finding trials when a
wide range of doses needs to be investigated).
Despite the hype surrounding adaptive designs, they
have been used infrequently so far; the flexibility
afforded by adaptive designs might not be compensated for by the potential loss in credibility associated
with their use.
Although adaptive designs can be useful to rescue
trials that are based on incorrect assumptions, they
should not generally replace well-designed trials that
use conventional approaches (e.g., group sequential
designs that are more statistically efficient).
Generally speaking, treatment effectindependent
adaptations raise few difficulties, whereas treatment
effect dependent adaptations are to be implemented
with caution and use of appropriate statistical methods.
Covariate-adaptive randomization is useful to minimize imbalances in prognostic factors among treatment groups; in contrast, outcome-adaptive
randomization is unhelpful statistically, difficult logistically, and unnecessary ethically.
Table 1. Pros and Cons of Adaptive Sample Size Increases

Pros
Potential to rescue a trial that would

miss statistical significance
Flexibility if design considerations
were inadequate
Cons
Statistically inefficient
Emphasis on statistical significance
rather than clinical relevance
Changes in patient population or other
temporal trends
Can often be substituted by nonadaptive
sample size increases that do not affect
the type I error
treatment effect).5 In other words, adaptive designs are

inefficient compared with group sequential designs. This
mathematical result has profound consequences because it
implies that even the most astute adaptive design can, in
theory, be replaced by a group sequential counterpart that
has better statistical power and none of the dangers created
by adaptations made to an ongoing trial. In fairness to
adaptive designs, the superior efficiency of group sequential
designs assumes that the appropriate spending function can
be prespecified and that interim analyses come at no cost,
which is not the case in practice.6
Figure 1 compares a group sequential design with an
adaptive design in the same situation. This figure illustrates
the putative advantage of the adaptive design that starts
with a small sample size based on a large treatment effect
and increases the sample size if the observed effect is
smaller than anticipated, whereas the group sequential
design starts with a large sample size based on a small
treatment effect and stops early if the observed effect is
larger than anticipated. Stated differently, adaptive designs
take an optimistic view and adapt if required, whereas
group sequential designs take a pessimistic view and stop
early if indicated. In the example of Fig. 1, the group
sequential design is based on a difference in proportions of
0.05 and requires 1,400 patients but can stop after 350, 700,
or 1,050 patients if the difference is equal to approximately
0.15, 0.075, or 0.05, respectively. In contrast, the adaptive
design is based on a difference in proportions of 0.10 and
requires only 400 patients, but the sample size can be
increased if the conditional power at 200 patients or 400
patients is too low.
The conditional power is the power that the trial is
expected to have at its final analysis, given the data from the
already accrued patients and assuming that the protocolspecified difference will apply to all patients still to be
accrued. Note that at the beginning of the trial, the conditional power is simply equal to the power because no data
are available yet. As stated above, it is essential, when
recalculating the sample size in an adaptive manner, to use
appropriate statistical methods. Several methods for this
exist; Tsiatis and Mehta provide a review and examples of
these methods.5
Outcome-Adaptive Randomization
In outcome-adaptive randomization, the probability of

allocating the next patient to one of the trials treatment
groups is computed dynamically to favor the treatment
group with the best outcome so far.7 For instance, if response (tumor shrinkage) was the outcome of interest,
patients would be allocated preferentially to the treatment
group with the highest response rate. This approach has
LIMITATIONS OF ADAPTIVE CLINICAL TRIALS
Fig. 1.
Comparison between a group sequential design and an adaptive design.
been called a play-the-winner strategy.8 A crucial distinction needs to be made between covariate-adaptive randomization and outcome-adaptive randomization. Indeed,
although the former raises no particular issue, the latter is
fraught with problems.
First, the outcome that is used to adapt the randomization has to be observed early and reliably, and it must be
reasonably predictive of important clinical endpoints for the
adaptation to succeed at placing more patients in the better
treatment group.
Second, adaptive randomization can result in major imbalances among treatment arms, which in turn negatively
affects the statistical power of the trial.
Third, the statistical inference is complicated because the
treatment assignments and the responses are correlated; as
a consequence, rerandomization tests must be used instead
of traditional likelihood-based tests.
Fourth, adaptive randomization can cause accrual bias (if
patients wait for the probability of receiving the better
treatment to increase) and/or selection bias (if patients are
aware of the emerging difference among the treatment
groups).
Last but not least, it is incorrect to claim that adaptive,
randomization is ethically superior to fixed randomization
because equipoise mandates that allocation to any of the
treatment groups be considered equally desirable. It might
make sense to allocate more patients to the experimental
group than to the control group, but the justification for
doing so is that more information is needed about a new
treatment than about a well-established standard treatment. When such is the case, a fixed unequal allocation ratio
(such as a 2:1 ratio in favor of the experimental group) will
do just as well as adaptive randomization, without being
subject to the problems listed above.9
Other Types of Adaptive Designs
There are many other types of adaptive designs, some of

which do not fall into the two clear-cut categories discussed.
We briefly mention two types of designs that have attractive
properties but have also been rarely used in practice.
CRM in Phase I Trials
Classic phase I cancer trials are aimed at determining the

maximum tolerated dose (MTD) of a new drug or combination of drugs.10 They are usually performed according to a
fixed design called the 3 3 design. The design proceeds in
cohorts of three patients, with the first cohort being treated
at the minimum dose of interest and the next cohorts being
treated at increasing dose levels according to a predetermined dose escalation scheme. The dose escalation proceeds
until at least one dose-limiting toxic effect is observed in a
cohort of three patients, in which case a second cohort of
three patients is treated at the same dose level. The dose
escalation stops as soon as at least two patients experience
a dose-limiting toxic effect, either in the first cohort of three
patients treated at that dose level or in the two cohorts of
three patients treated at that dose level. Although this
design is used in almost every phase I cancer trial today, it
has several limitations.11 First, too many patients may be
treated at low doses, with virtually no chance of efficacy.
Second, dose escalation may be too slow because of an
excessive number of escalation steps, resulting in trials that
take longer than needed to get to the MTD. Third, too few
patients may be treated near the MTD, resulting in substantial residual uncertainty about the dose recommended for
further trials, which raises ethical concerns. Indeed, if the
recommended dose is chosen too low, it can fail to have
efficacy in phase II trials, whereas if it is chosen too high, it
can put patients at unacceptable risk in phase II trials. Last
but not least, the 3 3 design makes no allowance for
patient variability. An adaptive design known as the CRM
was originally proposed by OQuigley et al12 in the early
1990s. This design involves a statistical approach based on
an assumed dose-response relationship, which is described
through a mathematical function that links the probability
of a dose-limiting toxic effect and the dose level. The CRM
design is adaptive insofar as the dose to use for the next
patient is determined from the toxic effects experienced by
all the patients already treated so far. Many modifications to
the CRM have been proposed, and simulation studies have
135
MARC BUYSE
Table 2. Comparisons of Three Approaches for Late Clinical Development
Approach
Benefits
Phase II trial followed by phase III trial
Seamless phase II/III trial
Phase III trial with interim analyses
Costs and Risks
Standard approach
No regulatory risk
Phase III trial provides independent confirmation of
phase II results
Can adapt the phase III based on the phase II results
Statistical method well established
Upfront commitment for phase II only
Much experience with group sequential designs
No regulatory risk
Optimal statistical approach if several interim looks
are planned
shown that it generally outperforms the 3 3 design.13,14

Ironically, however, the CRM is not as popular as it should
be given its attractive properties, probably because it requires calculations to be performed for the next dose to be
determined, whereas the 3 3 design does not.
Seamless Phase II/III Designs
It is possible to embed a phase II trial into a phase III trial

so that the transition between the two phases is operationally seamlessas opposed to performing a randomized
phase II trial followed by a separate phase III trial. The
simplest version of this approach consists of using a classic
phase II design to screen for activity based on response and
to calculate the sample size required of the phase III trial
based on the final outcome of interest, such as time to
progression or survival. Because the purpose of the phase II
trial is only to stop for futility on the basis of lack of activity,
there is no inflation in type I error. One-stage or two-stage
phase II designs can be used, as well as a selection design if
several experimental arms are simultaneously screened. In
all cases, the phase II and phase III portions of the trial are
designed independently of each other. Table 2 contrasts a
seamless transition from phase II to phase III with two other
approaches. Some authors have proposed to use a Bayesian
approach to expand the phase II trial to a phase III trial.15
A different approach that is particularly useful in selecting one or more doses of a new investigational agent is to use
Longest development time

Largest total sample size
Negotiation with regulatory agencies essential

Less experience with adaptive designs
Requires large upfront commitment
Difficult to design or start phase III based on scanty
early data
an inferentially seamless design in which several doses are

tested in the phase II portion of the trial and to select only
the most promising one to continue in the phase III portion.16 Various designs have been proposed that control the
overall significance level of the trial, whether or not there
are adaptations of some design aspects at the end of the
phase II trial. Jennison and Turnbull offer a review and
useful guidance.17,18
Conclusion
There is little doubt that clinical research is in need of

reengineering to provide efficient readouts of efficacy and
safety on the large number of treatments currently developed by pharmaceutical, biotechnology, and medical device
companies. Running large-scale trials that have a high
probability of failure is clearly undesirable. Many innovative
methods have been proposed, including adaptive designs,
Bayesian designs, and biomarker-based designs. Some recent trial designs combine all of these ideas and constitute,
as such, exciting models for further developments.19,20 The
future will tell which of these innovations are useful and
when they constitute a definite improvement over classic
approaches. In the meantime, oncologists involved in clinical
research must be aware of the limitations of each approach
and adjust their expectations accordingly. Clinical trial
sponsors might wish to consult regulatory guidances already
available on adaptive designs.21,22
Author
Marc Buyse
Employment or
Leadership
Positions
IDDI
Consultant or
Advisory Role
Stock
Ownership
IDDI
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Rosenberger WF, Lachin JM. Randomization in Clinical Trials: Theory
and Practice. New York, NY: Wiley; 2002.
2. Gould AL. Planning and revising the sample size for a trial. Stat Med.
1995;14:1039-1051.
3. Mehta CR. Sample size reestimation for confirmatory clinical trials. In:
Harrington D (ed). Designs for Clinical Trials: Perspectives on Current Issues.
New York, NY: Springer; 2011.
4. Kim KM. Sequential designs for clinical trials. In: Harrington D (ed).
Designs for Clinical Trials: Perspectives on Current Issues. New York, NY:
Springer; 2011.
5. Tsiatis AA, Mehta C. On the inefficiency of the adaptive design for
monitoring clinical trials. Biometrika. 2003;90:367-e78.
6. Brannath W, Bauer P, Posch M. On the efficiency of adaptive designs for
flexible interim decisions in clinical trials. J Stat Planning Infer. 2006;136:
1956-1961.
136
7. Hu F, Rosenberger WF. The Theory of Response-Adaptive Randomization in Clinical Trials. New York, NY: Wiley; 2006.
8. Wei LJ, Durham S. The randomized play-the-winner rule in medical
trials. J Am Stat Assoc. 1978;73:840-843.
9. Korn EL, Freidlin B. Outcome-adaptive randomization: is it useful?
J Clin Oncol. 2003;90:367-378.
10. Eisenhauer E, Twelves C, Buyse M. Phase I Clinical Trials in Cancer.
Oxford, England: Oxford University Press; 2006.
11. Cheung YK. Designs for phase I trials. In: Harrington D (ed). Designs
for Clinical Trials: Perspectives on Current Issues. New York, NY: Springer;
2011.
12. OQuigley J, Pepe M, Fisher L. Continual reassessment method: a
practical design for phase 1 clinical trials in cancer. Biometrics. 1990;46:33-48.
13. Chevret S. The continual reassessment method in cancer phase I
clinical trials: a simulation study. Stat Med. 1993;12:1093-1108.
LIMITATIONS OF ADAPTIVE CLINICAL TRIALS

14. Goodman SN, Zahurak ML, Piantadosi S. Some practical improvements in the continual reassessment method for phase I studies. Stat Med.
1995;14:1149-1161.
15. Inoue LY, Thall PF, Berry DA. Seamlessly expanding a randomized
phase II trial to phase III. Biometrics. 2002;58:823-831.
16. Bretz F, Koenig F, Brannath W, et al. Adaptive designs for confirmatory clinical trials. Stat Med. 2009;28:1181-1217.
17. Jennison C, Turnbull BW. Confirmatory seamless Phase II/III clinical
trials with hypotheses selection at interim: opportunities and limitations.
Biom J. 2006;48:650-655.
18. Jennison C, Turnbull BW. Adaptive seamless designs: selection and
prospective testing of hypotheses. J Biopharm Statist. 2007;17:1135-1161.
19. Zhou X, Liu S, Kim ES, Herbst RS, Lee JJ. Bayesian adaptive design
for targeted therapy development in lung cancera step toward personalized
medicine. Clin Trials. 2008;5:181-193.
20. Barker AD, Sigman CC, Kelloff GJ, Hylton NM, Berry DA, Esserman
LJ. I-SPY 2: an adaptive breast cancer trial design in the setting of neoadjuvant chemotherapy. Clin Pharmacol Ther. 2009;86:97-100.
21. European Medicines Agency (EMA), Committee for Medicinal Products
for Human Use (CHMP). Reflection Paper on Methodological Issues in
Confirmatory Clinical Trials Planned With an Adaptive Design. http://home.
att.ne.jp/red/akihiro/emea/245902enadopted.pdf. Accessed February 10,
2012.
22. U.S. Department of Health and Human Services, Food and Drug
Administration (FDA). Draft Guidance for Industry: Adaptive Design Clinical Trials for Drugs and Biologics. http://www.fda.gov/downloads/Drugs/
GuidanceComplianceRegulatoryInformation/Guidances/UCM201790.pdf. Accessed February 10, 2012.
137
ENDPOINTS FOR CANCER TRIALS IN 2012:

STATISTICAL, REGULATORY, AND CLINICAL
PERSPECTIVES (eQ&A)
CHAIR
Daniel J. Sargent, PhD
Mayo Clinic
Rochester, MN
SPEAKERS
Ann T. Farrell, MD
U. S. Food and Drug Administration
Silver Spring, MD
Deborah Watkins Bruner, PhD, RN
Abramson Cancer Center, University of Pennsylvania
Philadelphia, PA
Capturing the Patient Perspective:

Patient-Reported Outcomes as Clinical
Trial Endpoints
Deborah Watkins Bruner, RN, PhD, Benjamin Movsas, MD, and Ethan Basch, MD
Overview: Just as clinical trial design and rigor have evolved

with improvements in methods and processes, so too have
methods for capturing patient data in clinical trials. Substantial evidence suggests that standard physician reporting of
symptoms for which we lack objective diagnostics (e.g., pain)
is often discordant with patient self-report. Current reporting
using the National Cancer Institute Common Terminology
Criteria for Adverse Events (CTC[AE]) for symptom capture
relies on a filtering system, from patient to physician to
medical record to medical record abstraction to data entry,
with each step requiring interpretation and the possibility of
error. In contrast, patient-reported outcomes (PROs) eliminate
the filter and rely on direct report. Furthermore, the lack of
validation and training in use of the CTC(AE) creates an
inadequate data capture system. Inadequacies might be observed as underreporting or overreporting symptom preva-
HE TITLE of this chapter, Capturing the Patient

Perspective: Patient-Reported Outcomes as Clinical
Trial Endpoints, implies that the patient perspective is
separate from other clinical trial endpoints. All endpoints
involve patient data. Why then the emphasis on the patient
perspective? Currently, all cancer clinical trials use the
National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTC[AE]), a descriptive system
that is used for adverse event reporting. The CTC(AE) uses
a grading (severity) scale for each adverse event term.1
Grading is performed by either the physician or research
assistants, who abstract information from the patients
medical records. Either method of grading requires interpreting or filtering the patient experience. In contrast, A
patient-reported outcome (PRO) is a measurement of any
aspect of a patients health status that comes directly from
the patient (i.e., without interpretation of the patients
responses by a physician or anyone else).2
The Pros and Cons for Use or Nonuse of PROs in
Clinical Trials
What do we gain over standard CTC(AE) reporting by

including the patient perspective in clinical trials? Table 1
lists the complementary benefits of the CTC(AE) and PROs.
However, given limited resources, what is the downside of
experienced physicians and research associates grading a
patients symptoms? The downside is they might not be
correctly grading symptoms and/or they might be missing
key information; either can occur for a host of reasons.
lence and severity compared with patient self-report.

Inaccuracies in symptom reporting can lead to missing important prognostic information, lack of understanding of patient
adherence with therapies, and lack of information for patient
decision making. They can also lead to opportunities lost in
terms of labeling claims and comparative effectiveness analyses. New developments in patient-reported outcome (PRO)
reporting, including the PRO-CTC(AE) and models for incorporation of PROs in clinical trials, might facilitate routine
PRO reporting complementary to CTC(AE) in clinical trials. In
addition, the cadre of validated PRO instruments already in
existence allows for more in-depth, hypothesis-driven evaluations. For standard toxicity reporting, the time has come for
mandatory routine PRO symptom reporting complementary to
the CTC(AE).
is evident in the rare instances when the CTC(AE) has been

tested for interrater reliability, which is modest at best.
Kabaet al4 evaluated the reliability of CTC version 2.0 with
five experienced research associates independently reviewing 17 different medical records and grading toxicities.
Agreement among raters was lowest for symptoms that are
difficult to directly observe and highest for symptoms that
have an associated laboratory test, for example, agreement
for nausea was 0.47 (0.23 0.71, 95% CI) and agreement for
febrile neutropenia was 0.88 (0.731, 95% CI).4 Atkinson
and colleagues5 assessed interrater reliability of multiple
CTC(AE) symptoms. In a study at an NCI-designated comprehensive cancer center, patients receiving active chemotherapy routinely had toxicity assessments performed by
two clinicians. The ratings were performed independently,
without access to the other physicians reports, on a medical
record form that included the name of each symptom, a
checkbox to note the grade of the toxicity, and a key with
definitions of each CTC(AE) grade for each symptom. Agreement among raters was low for most symptoms (e.g. diarrhea [0.58], constipation [0.5], nausea [0.52], and vomiting
[0.46]). In addition to poor interrater reliability, no assessments of validity of the CTC(AE) have been published,
meaning it is unknown whether grade 1 is clinically different from grade 2, etc. By contrast, the use of PROs in clinical
trials is held to stringent and rigorous development and
psychometric validation as described in detail in the U.S.
Food and Drug Administration (FDA) guidance for use of
PROs.2
Lack of Reliability and Validity of the Current

Grading System
The CTC(AE) was not designed as a physician checklist or

patient interview, and there are a variety of ways in which
the grading is performed and by whom it is performed,
which are described elsewhere.3 This variation in grading
leads to questions about training to use the grading system,
and the answer is there is no training. This lack of training
From the Nell Hodgson Woodruff School of Nursing, Winship Cancer Institute, Emory
University, Atlanta, GA; Radiation Oncology Department, Henry Ford Health System,
Detroit, MI; Memorial Sloan-Kettering Cancer Center, New York, NY.
Address reprint requests to Deborah Watkins Bruner, RN, PhD, Nell Hodgson Woodruff
School of Nursing, Winship Cancer Institute, Emory University, 1520 Clifton Rd., Room
232, Atlanta, GA 30322-4201; email: dwbrune@emory.edu.
1092-9118/10/1-10
139
BRUNER, MOVSAS, AND BASCH

Table 1. Complementary Use of Common Terminology Criteria for Adverse Events (CTC[AE]) and Patient-Reported Outcomes (PROs)
in Clinical Trials
Primary use
Most useful for
Best captures
Valid
Reliable
Data capture method
Time of data capture
a
CTC(AE)
PROs
Toxic effects reporting

Objective assessment (e.g., diagnostic test, imaging,
overt sign, such as bleeding)
Severity, need for physician intervention
Not tested
No
Through layers of interpretation
As it occurs/as physician picks it up
Health status reporting

Subjective assessment (e.g., cannot be seen, felt, heard, observed,
or clinically tested by physician)
Severity, function, effect on quality of life and treatment adherence
Yesa
Yesa
Directly from the patient
At designated time points
Legacy instruments psychometrically tested to varying degrees; for current U.S. Food and Drug Administration use, must conform to stringent guidelines
Underreporting Prevalence of Symptoms
Data have demonstrated that PROs often identify more

symptoms than standard CTC(AE) reporting. In a phase III
trial of a cream compared with a placebo to prevent breast
cancer radiation therapyrelated dermatitis, PROs delineated a wider spectrum of toxicity than CTC(AE) and provided more information on rash, redness, pruritus, and
annoyance measures compared with CTC(AE) findings of
rash and pruritus.6
Underreporting and Overreporting Levels of Severity
A burgeoning body of evidence indicates that physicians

often underreport or overreport the level of severity of
symptoms compared with patient self-report. WatkinsBruner and colleagues,7 in the NCI-sponsored Radiation
Therapy Oncology Group (RTOG), reported discordance between physician and patient reporting in RTOGs first published PRO report (RTOG 90-20), A Phase II Trial of
Radiotherapy Plus a Radiosensitizer for Locally Advanced
Prostate Cancer. In a comparison of patient self-report of
symptoms on a validated measure with physician ratings
KEY POINTS
140
The current cancer clinical trial reporting system

using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTC[AE]) is
inadequate, by itself, to capture subjective symptoms.
Patient-reported outcome (PRO) measures, rigorously developed and validated, provide important and
accurate information required for robust clinical trials reporting.
The risk of not including PROs in routine reporting of
clinical trials is substantial inaccuracies in reporting
of prevalence and severity of symptoms related to
treatments, missing important prognostic information, lack of understanding of patient adherence with
therapies, and lack of information for patient decision
making.
Opportunities lost in not including PROs in routine
clinical reporting include lack of information for labeling claims and comparative effectiveness.
Validated and reliable PROs exist that are easily
available for both routine reporting of symptoms and
more complex evaluations of hypotheses-driven endpoints.
of the same symptoms on a toxicity scale, disagreement

between patient and physician reports of severity of dysuria,
diarrhea, and erectile dysfunction ranged from 13% to 45%.
The RTOG determined that the findings warranted routine
use of PROs in clinical trials. A larger initiative to assess
how well PROs correspond to reports of the same symptoms
as rated by the CTC(AE) across NCI cancer clinical trial
cooperative groups included a pooled analysis of 12 lung
cancer clinical trials conducted by three cooperative groups.8
Data were assessed in 1,013 patients who had both adverse
events and PROs recorded. Agreement between incidence of
any grade adverse event and a substantial decrease in PROs
ranged from 27% to 61% for grade 2 and higher and 36% to
61% for grade 3 and higher toxicities. Agreement between
the incidence of a specific grade adverse event and a substantial decrease in a corresponding PRO ranged from 0% to
38% for grade 2 and higher and 0% to 14% for grade 3 and
higher toxicities.8
Missing Prognostic Information
Data from clinical trials have shown validated PROs to be

prognostic for survival in almost every cancer disease site,
especially in more advanced stage disease. For example, in
an RTOG clinical trial of a radioprotectant for radiationinduced esophagitis in the treatment of nonsmall cell lung
cancer, baseline patient-reported scores on the European
Organisation for Research and Treatment of Cancer
Quality-of-Life Questionnaire replaced known prognostic
factors (performance status, stage, sex, age, race, marital
status, histologic features, and tumor location) as the sole
predictor of long-term overall survival. A 10-point higher
baseline quality-of-life score corresponded to a decrease in
the hazard of death by approximately 10% (p 0.004).9 A
Gynecologic Oncology Group analysis of a phase III trial
of cisplatin with or without topotecan in advanced cervical
cancer found the patient-reported Functional Assessment of
Cancer TherapyGeneral (FACT-G) plus the FACT-Cervix
subscale had a similar association with survival as the
RTOG study even though different PRO measures were used
and different disease sites were involved. A 10-point higher
baseline score on FACT-Cervix corresponded to a decrease
in hazard of death by approximately 10% (p 0.002).10 A
pharmaceutical companysponsored, phase III trial demonstrated efficacy of sunitinib, which prolonged progressionfree survival for patients with metastatic renal cell cancer.
Three baseline PRO measures were individually predictive
of progression-free survival. Higher scores at baseline on
FACT-G, the FACTKidney Symptom Index disease-related
symptoms subscale, and EuroQol Groups visual analog
scale were significantly associated with longer progression-
PATIENT-REPORTED OUTCOMES AS TRIAL ENDPOINTS
free survival (hazard ratios, 0.93, 0.89, and 0.91; p 0.001,

p 0.001, and p 0.008, respectively). These data indicate
that the risk of tumor progression or death was approximately 7%, 11%, and 9% lower, respectively, for every higher
change score of 5 points on FACT-G, 2 points on the
FACTKidney Symptom Index disease-related symptoms
subscale, and 10 points on EuroQol Groups visual analog
scale.11
Gotay et al12 confirmed this trend of the prognostic significance of PROs, Gotay et al in a systematic literature review
of cancer clinical trials in which they found that 36 (92%) of
39 studies had at least one PRO that was prognostic of
survival (p 0.05) in multivariate analysis. Of the 39
clinical trials that met inclusion criteria, most were lung (12
trials) and breast (eight trials) cancer trials, with a total of
13,874 patients. The European Organisation for Research
and Treatment of Cancer Quality-of-Life Questionnaire was
the most common PRO used, found in 56% of the trials.
Effect sizes varied, with 24 of 36 (66%) having a small effect
size, eight (23%) having a moderate effect size, and 4 (11%)
having a large effect size. Adjusting for performance status,
treatment arm, stage, weight loss, and serum markers,
PROs still provided distinct prognostic information for survival beyond standard clinical measures.
Lack of Understanding of Patient Adherence
Aromatase inhibitors provide an excellent example of how

not including PROs or not including them at the right time
points can mislead our understanding of patient adherence.
A large pivotal clinical trial of 5,187 patients demonstrated
the efficacy of an aromatase inhibitor to decrease local or
metastatic recurrence or new contralateral breast cancer.
The toxicities associated with aromatase inhibitors found
using CTC version 2.0 were reported as primarily grade 1 or
2, and those that were higher in the aromatase inhibitor
group compared with the placebo group were hot flashes,
arthritis, arthralgia, and myalgia. At a median of 2.4 years
of follow-up, no significance was found between discontinued
use of the aromatase inhibitor compared with placebo (4.5%
vs. 3.6%, p 0.11).13 In the quality-of-life analysis of the
same trial, Whelan and colleagues14 found small but meaningful worsening in mean change scores from baseline in the
PRO domains of vasomotor effects, bodily pain, vitality,
physical functioning, and sexual function, with the last two
domains having no corresponding detection using CTC reporting. However, the findings of studies in which adherence
was evaluated over shorter durations suggest better adherence than studies with longer follow-up. One-year aromatase inhibitor adherence rates range from 77% to 88%,
whereas 4- to 5-year prescription-based adherence rates
were 27% to 49% in other studies.15 PROs can help provide
the answer to decreasing adherence over time. PRO data
from two pivotal phase IV aromatase inhibitor trials found
high prevalence rates for joint pain (59.6%), hot flushes
(52%), lost interest in sexual intercourse (51.4%), and lack
of energy (40.3%). PROs resulted in higher prevalence rates
compared with physician ratings for most symptoms.16 A
recent study found that 36% of women who were prescribed
aromatase inhibitors for 5 years reported ending treatment
because of symptom bother.16
Lack of Information for Patient and

Clinical Decision Making
Patients often seek information that would help them

make decisions regarding treatments and symptom management. Without systematic PRO assessments in clinical trials, patients might seek out less rigorously gathered data.
For example, www.patientslikeme.com/ is a popular Webbased tool that seeks to provide users with information to
help make treatment decisions and manage symptoms. An
analysis of approximately 2,000 unique PRO reports on
modafinil use across five disease conditions indicated how
patients are easily able to gather information on the use of
the drug for management of symptoms. However, less than
1% of members who reported taking modafinil reported
taking it for an approved purpose (narcolepsy and excessive
daytime sleepiness resulting from sleep apnea). Patients are
able to access information on the other self-reported uses of
the drug, for example, general fatigue (68%), excessive
daytime sleepiness (16%), and difficulty concentrating (3%),
as well as perceived effectiveness of ameliorating each
symptom.18 In addition to the off-label use of the drug,
patients report adverse effects they attribute to the drug.
Routine reporting of PROs in clinical trials and a commitment to lay broadcasting beyond academic publication
would provide more accurate information for decision making.
Lack of Information for Labeling Claims
Few cancer drugs have received FDA approval for a

symptomatic benefit. A recent exception is the November
2011 approval of ruxolitinib for the treatment of myelofibrosis. FDA approval for symptomatic benefit before 2011 goes
back to 1998. However, between 1998 and 2011 almost 70
anticancer drugs have been approved. Incyte Corp. obtained
this rare symptom benefit labeling claim through close
interaction with the FDA. The primary endpoint was reduction in spleen volume, and the secondary PRO endpoint was
the total symptom score using a specifically developed
measure following the FDA PRO guidelines. The company
worked with the FDA for 2 years to finalize instrument
approval. The PRO measured six symptoms: night sweats,
itching, abdominal discomfort, pain under the ribs on the
left side, early satiety, and bone or muscle pain. Along with
rigorous instrument development, the clinical trial hypothesized a large effect of 50% or greater reduction in total
symptom score from baseline to week 24. Forty-six percent
of patients in the ruxolitinib arm met the PRO threshold
compared with 5.3% in the placebo group (p 0.0001).
Richard Pazdur, the director of FDAs Office of Hematology
Oncology Products, was quoted as saying, [The PRO] was a
secondary endpoint, but in our mind this is why we gave the
application full approval. One could quibble about the importance of reduction in spleen size, but with reduction in all
the symptoms, full approval was warranted.19
Lack of Information for Comparative Effectiveness
Without PRO data in clinical trials, we would have to rely

solely on toxicity data for comparative effectiveness of treatment modalities, but we have a preponderance of literature
showing that physician-reported toxicities do not tell the full
story. For example, in RTOG 0126, a phase III clinical trial,
patients were randomly assigned to high-dose (79.2-Gy) or
141

Table 2. Barriers and Solutions to Standard Use of PROs in Clinical Trials
Barrier
Solution
Lack of off-the-shelf validated instruments

Inadvertent unmasking
Need to make PROs publically available

Require a large effect size
Substantiate symptomatic benefits via objective measures (such as radiographic
or serum biomarker responses)
e-products for data capture, real-time reminders, real-time monitoring, telephone
interviews when patient is too ill to complete PRO
PROs better identify baseline symptoms related to disease or prior treatment than
current standard physician toxic effect reporting, making it clearer during a
trial which symptoms are attributable to study drug compared with preexisting
causes
Standard PRO reporting should be viewed no differently than standard CTC(AE)
reporting and routinely incorporated as such
Missing data
There may be concerns that systematic assessment will document more low
grade toxicities
Additional resources for routine inclusion of PROs in clinical trials for symptom
assessment may be needed since the NCI generally only funds PROs that
answer a specific hypothesis
Abbreviations: CTC(AE), Common Terminology Criteria for Adverse Events; NCI, National Cancer Institute; PRO, patient-reported outcome.
standard-dose (70.2-Gy) radiation therapy for localized prostate cancer. A preliminary analysis of the high-dose arm,
comparing 3-dimensional conformal radiotherapy (3DCRT)
and intensity-modulated radiotherapy (IMRT), where each
institution declared its choice of 3DCRT or IMRT before trial
participation, was conducted. A total of 763 patients were
in the high-dose arm and included in the toxicity analysis,20
and 499 patients (65%) were included in the PRO analysis.21
Although the CTC version 2.0 toxicity report indicated a
benefit in favor IMRT for grade 2 or higher gastrointestinal
toxicities (22% 3DCRT vs. 15% IMRT, p 0.039) and in
favor of IMRT for combined gastrointestinal and genitourinary grade 2 or higher toxicities (15.1% 3DCRT vs. 9.7%
IMRT, p 0.042), PROs did not corroborate a benefit
patients could note. Corresponding PRO gastrointestinal
and genitourinary variables indicated no substantial differences between 3DCRT and IMRT. An important difference
that PROs demonstrated in favor of IMRT not found on
physician reporting was erectile dysfunction; of note, the

IMRT benefit was primarily in men younger than 70 years
(p 0.04). Given the higher cost of IMRT over 3DCRT for
the treatment of prostate cancer, the PRO data add an
important dimension to the comparative effectiveness equation.
Although the benefits of inclusion of PROs in clinical trials
and the risks of noninclusion have been delineated, there
are some barriers to routine inclusion of PROs in clinical
trials. However, solutions to these barriers exist (Table 2).
Numerous validated PRO instruments with the breadth and
depth to answer hypothesis-specific clinical trial questions
are available, but a more parsimonious method would help
routine inclusion in clinical trials.
A major step forward has been the development of the
PRO version of the CTC-AE (PRO-CTC[AE]).22 The PROCTC(AE) has been designed to capture a comprehensive
range of symptoms and functioning that enhances monitor-
Fig. 1. Radiation Therapy Oncology

Group (RTOG) outcomes model for guiding inclusion of patient-reported outcomes in clinical trials.
Abbreviations: RT, radiation therapy;
PSA, prostate specific antigen; QALYs,
quality-adjusted life years; PROs, patient reported outcomes.
142
PATIENT-REPORTED OUTCOMES AS TRIAL ENDPOINTS
ing of adverse events in cancer clinical trials. Approximately

78 CTC(AE) items have been converted to PROs, including,
for example, items purely or with a large subjective component, such as pain, fatigue, nausea or vomiting, elimination,
sexual function, swallowing, and dyspnea. This NCI investment lays the groundwork for standard inclusion of PROs in
clinical trials as a companion to the CTC(AE).
In addition to having publicly accessible validated measures for PRO assessments, a framework for incorporation
of PROs into clinical trials would help guide consistent
assessments. To that end, the RTOG developed an outcomes
model for guidance of clinical trials concept development.
This model was first developed as a triad of outcomes,
including clinical (e.g., standard survival and toxicity reporting), humanistic (patient-reported symptoms, quality
of life, functional status, and preferences), and economic
(quality-adjusted life-years and comparisons of effectiveness).23 The outcomes model has evolved with the increasing
focus on the complexity and interrelationships of additional
variables, including biomarkers and physical parameters
(particular to radiation treatment planning) that can influence and be influenced by patient-reported outcomes. Figure
1 depicts an example of the most recent iteration of this
model used to guide inclusion of endpoints on an active trial
(RTOG 0543), A Phase III Trial of Short Term Androgen
Deprivation With Pelvic Lymph Node or Prostate Bed Only
Radiotherapy in Prostate Cancer Patients With a Rising

PSA After Radical Prostatectomy.
Further advances in the area of PROs have recently
emerged from the September 2011 NCI-sponsored clinical
trials planning meeting, Identification of Core Symptoms
and Health-Related Quality of Life Domains for Use in
Cancer Research. An extensive process of literature and
source data review identified a lengthy list of symptoms
ranked by prevalence. A panel of experts narrowed the list to
a core set of symptoms that can serve as a guide to common
reporting across all cancer clinical trials. If adopted, this
core set of PROs could facilitate comparison and combination of data across cancer clinical trials around the world.
The publications from this meeting are in development.
Conclusion
This chapter provides a review of what we stand to lose

without the routine incorporation of PROs in clinical trials.
It is important to also consider what we stand to gain,
including more comprehensive reporting of prevalence of
symptoms, improved accuracy in reporting of levels of severity, increased prognostic specificity, greater understanding
of patient adherence, better information for patient and
clinical decision making, additional targets for labeling
claims, and information for comparative effectiveness. Most
importantly, we give voice to the one who counts most, the
patient.
Author
Deborah Watkins Bruner
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Endo
Pharmaceuticals
Bristol-Myers
Squibb
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Benjamin Movsas*
Ethan Basch*
REFERENCES
1. National Cancer Institute. Common Terminology Criteria for Adverse
Events (CTCAE) and Common Toxicity Criteria (CTC). 2009. http://evs.nci.
nih.gov/ftp1/CTCAE/About.htm. Accessed March 20, 2012.
2. U.S. Food and Drug Administration. Guidance for Industry: PatientReported Outcome Measures: Use in Medical Product Development to Support Labeling Claims. 2009. http://www.fda.gov/downloads/Drugs/Guidance
ComplianceRegulatoryInformation/Guidances/UCM193282.pdf. Accessed March
20, 2012.
3. Bruner DW. Should patient-reported outcomes be mandatory for toxicity
reporting in cancer clinical trials? J Clin Oncol. 2007;25:5345-5347.
4. Kaba H, Fukuda H, Yamamoto S, Ohashi Y. Reliability at the National
Cancer Institute-Common Toxicity Criteria version 2.0 [in Japanese]. Gan To
Kagaku Ryoho. 2004;31:1187-1192.
5. Atkinson TM, Li Y, Coffey CW, et al. Reliability of adverse symptom
event reporting by clinicians. Qual Life Res. Epub 2011 Oct 8.
6. Neben-Wittich MA, Atherton PJ, Schwartz DJ, et al. Comparison of
provider-assessed and patient-reported outcome measures of acute skin
toxicity during a Phase III trial of mometasone cream versus placebo during
breast radiotherapy: the North Central Cancer Treatment Group (N06C4).
7. Watkins-Bruner D, Scott C, Lawton C, et al. RTOGs first quality of life
study-RTOG 90-20: a phase II trial of external beam radiation with etanidazole for locally advanced prostate cancer. Int J Radiat Oncol Biol Phys.
1995;33:901-906.
8. Bruner DW, Gotay C, Moinpour C, et al. Value-added of patient-reported
outcomes (pro) in cooperative group oncology clinical trials: a pooled analysis.
In: Proceedings of the 15th Annual Meeting of the International Society for
Quality of Life Research; October 22-25. 2008; Montevideo, Uruguay. Abstract 28.
9. Movsas B, Moughan J, Sarna L, et al. Quality of life supersedes the
classic prognosticators for long-term survival in locally advanced non-smallcell lung cancer: an analysis of RTOG 9801. J Clin Oncol. 2009;27:5816-5822.
10. Monk BJ, Huang HQ, Cella D, Long HJ III. Quality of life outcomes
from a randomized phase III trial of cisplatin with or without topotecan in
advanced carcinoma of the cervix: a Gynecologic Oncology Group Study.
J Clin Oncol 2005;23:4617-4625.
11. Cella D, Cappelleri JC, Bushmakin A, et al. Quality of life predicts
progression-free survival in patients with metastatic renal cell carcinoma
treated with sunitinib versus interferon alfa. J Oncol Pract. 2009;5:66-70.
12. Gotay CC, Kawamoto CT, Bottomley A, Efficace F. The prognostic
significance of patient-reported outcomes in cancer clinical trials. J Clin
Oncol. 2008;26:1355-1363.
14. Whelan TJ, Goss PE, Ingle JN, et al. Assessment of quality of life in
MA.17: a randomized, placebo-controlled trial of letrozole after 5 years of
tamoxifen in postmenopausal women. J Clin Oncol. 2005;23:6931-6940.
15. Gotay C, Dunn J. Adherence to long-term adjuvant hormonal therapy
for breast cancer. Exp Rev Pharmacoecon Outcomes Res. 2011;11:709-715.
16. Oberguggenberger A, Hubalek M, Sztankay M, et al. Is the toxicity of
adjuvant aromatase inhibitor therapy underestimated? Complementary information from patient-reported outcomes (PROs). Breast Cancer Res Treat.
2011;128:553-561.
17. Wagner L, Zhao F, Chapman J, et al. Patient-reported predictors of
143

treatment discontinuation: quality of life among postmenopausal women with
primary breast cancer randomized to exemestane or anastrozole. Cancer Res.
2011;71(suppl):S6.
18. Frost J, Okun S, Vaughan T, et al. Patient-reported outcomes as a
source of evidence in off-label prescribing: analysis of data from Patients
LikeMe. J Med Internet Res. 2011;13:e6.
19. McCallister E, Usdin S. A professional trial. BioCentury. 2011;19:A1A4.
20. Michalski JM, Yan Y, Watkins-Bruner D, et al. Preliminary analysis of
3D-CRT vs. IMRT on the high dose arm of the RTOG 0126 prostate cancer
trial: toxicity report. Int J Radiat Oncol Biol Phys. 2011;81(suppl):S1-S2.
144
21. Watkins-Bruner D, Hunt D, Michalski J, et al. Preliminary analysis

of 3DCRT vs IMRT on the high dose arm of the RTOG 0126 prostate cancer trial: patient reported outcomes. Int J Radiat Oncol Biol Phys. 2011;
81(suppl):S44.
22. National Cancer Institute. Patient-Reported Outcomes (PRO) version
of the Common Terminology Criteria for Adverse Events (PRO-CTCAE).
2010. http://outcomes.cancer.gov/tools/pro-ctcae.html. Accessed March 20,
2012.
23. Bruner DW. Outcomes research in cancer symptom management trials:
the Radiation Therapy Oncology Group (RTOG) conceptual model. J Natl
Cancer Inst Monogr. 2007;37:12-15.
NEW LOOKS AND CHALLENGES FOR

COOPERATIVE GROUPS
CHAIR
Stephen S. Grubbs, MD
Helen F. Graham Cancer Center at Christiana Care
Newark, DE
SPEAKERS
Robert Leo Comis, MD
Coalition of Cancer Cooperative Groups
Philadelphia, PA
Gregory H. Reaman, MD
Childrens National Medical Center
Washington, DC
The New National Cancer Institute National

Clinical Trials Network
By Stephen S. Grubbs, MD
Overview: The National Cancer Institute (NCI) Cooperative

Group Program has been reviewed by three published studies
in the last 7 years evaluating the efficiency and effectiveness
of this national oncology clinical trials system. The recommendations for improvement from these reports have
prompted NCI to transform the Cooperative Group Program
HE NCI Cooperative Group Program has been an

essential contributor to the discovery of new cancer
therapies over the last 50 years. The program comprises
networks of cancer centers and community oncology practices that develop and conduct large-scale multicenter
clinical trials and promote discovery of correlative cancer
science. The cooperative groups have established the therapies now routinely used to diagnose, prevent, and treat
cancers.
In 2007, the NCI director requested the Institute of
Medicine (IOM) conduct a consensus study of clinical trials
and the NCI Cooperative Group Program and provide recommendations to improve the system. This evaluation was
preceded by an NCI review of the system by the Clinical
Trials Working Group (CTWG) in June 2005. The IOM
convened a 17-member committee representing a broad
range of the oncology research community. The activity
was supported by funding from the NCI and other cancer
research organizations, including ASCO. A report was
issued in April 2010 titled, A National Cancer Clinical
Trials System for the 21st Century. Key features of the
report are listed in Table 1.
The NCI has responded to the IOM, CTWG, and the
March 2010 NCI Operational Efficiency Work Group
(OEWG) reports with a proposed transformation and reorganization of the cooperative groups into a new NCI Clinical
Trials Network (NCTN). Network goals are listed in Table 2.
To achieve these goals, the NCI has embarked on cooperative group consolidation and reconfiguration of the network
evaluation process and funding.
The NCI has recommended the 10 cooperative groups
consolidate to four adult-oriented groups and one pediatric
group. The existing nine adult-oriented groups have elected
to merge or affiliate into the four entities listed in Table 3.
The NCI is in the process of developing a new funding
opportunity announcement (FOA) to be published in July
2012 that will establish the new evaluation and funding of
the network. The NCI has articulated the goals of the new
system evaluation (Table 4). NCI oversight and prioritization of the network will utilize the existing Disease Steering
Committees and a newly proposed Cross-Disease/Trials
Oversight Panel. More efficient clinical trials development
and activation have already been implemented by strict
timelines established in the OEWG report.
The new FOA is expected in July 2012, and competing
applications for the new FOA are to be submitted by November 2012, with NCI review in February 2013. Network
awards are expected in late 2013 with funding to begin in
2014.
146
into a new NCI National Clinical Trials Network (NCTN) to

improve the efficiency of large clinical trials and increase the
speed of cancer translational research. The new NCTN offers
community-based clinical investigators new opportunities to
advance cancer research in their community setting but also
presents challenges in promoting community-based research.
KEY POINTS
The Institute of Medicine, the Clinical Trials Working Group, and the Operational Efficiency Working
Group have published evaluations recommending improvements of the National Cancer Institute (NCI)
Cooperative Group Program.
The NCI is transforming the Cooperative Group Program into a National Clinical Trials Network by
group consolidation and a new evaluation and funding model.
Nine adult disease oriented cooperative groups have
responded by consolidating into four groups, joining
the existing single pediatric-oriented cooperative
group.
Community clinical trial sites and investigators will
potentially gain access to a greater variety of trials,
will benefit from operational standardization between groups, and may achieve increase per case
reimbursement.
Community sites, however, will be challenged by
diminished volunteer investigator time and pressure
to produce high accrual volume and remain at risk for
per case funding below actual cost.
Community Needs from the NCI NCTN
Community clinical investigators are essential participants in a successful national oncology clinical trials system.
They have access to large numbers of newly diagnosed
patients and contribute significant accruals to phase III and
cancer control and prevention trials. The Community Clinical Oncology Program and cancer center community affiliate sites contribute over 50% of the patients annually to
the cooperative group total accrual. The new NCTN offers
opportunities and challenges for the community clinical
investigators.
A successful community clinical trials site requires five
From the Helen F. Graham Cancer Center, Delaware Christiana Care CCOP Newark,
DE.
Address reprint requests to Stephen S. Grubbs, MD, Helen F. Graham Cancer Center,
Delaware Christiana Care CCOP, 4701 Ogletown-Stanton Rd., Newark, DE, 19713-2055;
email: ssgrubbs@cbg.org.
1092-9118/10/1-10
NCI NATIONAL CLINICAL TRIALS NETWORK

Table 1. IOM Committee Recommendations
Goal I. Improve the speed and efficiency of the design, launch, and conduct of
clinical trials.
Review and consolidate some front-office operations of the cooperative groups
on the basis of peer review.
Consolidate back-office operations and improve processes.
Streamline and harmonize government oversight.
Improve collaboration among stakeholders.
Goal II. Incorporate innovative science and trial design into cancer clinical trials.
Support and use biorepositories.
Develop and evaluate novel trial designs.
Develop standards for new technologies.
Goal III. Improve the means of prioritization, selection, support, and completion of
cancer clinical trials.
Reevaluate the National Cancer Institutes role in the clinical trials system.
Increase the accrual volume, diversity, and speed of clinical trials.
Increase funding for the cooperative group program.
Goal IV. Incentivize the participation of patients and physicians in clinical trials.
Support clinical investigators.
Cover the cost of patient care in clinical trials.
Table 2. National Cancer Institute Clinical Trials Network

Transformation Goals
Prioritize molecular characterization resources and develop molecularly driven
trial design.
Improve prioritization of phase III portfolio across disease entities.
Remove disincentives to study less common diseases.
Create a shared IT structure.
Harmonize procedures.
Integrate imaging technology.
Integrate national tissue banking resources.
Grant open access to the network for clinical and transitional investigations.
Table 4. National Clinical Trials Network Group Review

and Funding
Modified U10 program
Increased per case reimbursement for high accruing sites
National Cancer Institute (NCI) external peer review of groups in the same
review cycle
New review criteria to include evaluation of collaboration within the network and
other NCI-funded programs (e.g., Specialized Programs of Research Excellence),
operational efficiency, and overall scientific quality.
Table 5. Community Clinical Trial Site Desirables

I. Clinical Trial Access
Phase II and III treatment trials
Cancer control and prevention trials
Comparative effectiveness trials
Innovative scientific questions
Common diseases and expanded performance status eligibility
II. Group Participation
Group identity
Interaction with academic colleagues
Leadership opportunities in both scientific and operation activities
Publication authorships
Professional recognition
III. Operational and Regulatory Standardization
Common protocol templates and data submission
Consolidated audit system
Regulatory burden easement
IV. Patient Recruitment Enhancement
Electronic health record cues
Clinical trial marketing to the public
Underserved community accrual
V. Adequate Funding
Reimbursement for actual cost
Table 3. Adult Cooperative Group Consolidation

Table 6. Community Clinical Trial Site Challenges
American College of Surgeons Oncology Group, Cancer and Leukemia Group B,
North Central Cancer Treatment Group (Alliance for Clinical Trials in Oncology)
Gynecologic Oncology Group, National Surgical Adjuvant Breast and Bowel
Project, Radiation Therapy Oncology Group (Neuroinformatics Research Group
[NRG] Oncology)
American College of Radiologys Imaging Network, Eastern Cooperative
Oncology Group
Southwest Oncology Group
main areas of support from NCTN and NCI: (1) clinical trial
access, (2) group participation, (3) operation and regulatory
standardization, (4) patient recruitment enhancement, and
(5) adequate funding and support (Table 5). However, the
new proposed network poses both old and new challenges,
including NCTN membership issues, site trial portfolio
management, volunteer investigator time, and funding in a
time of stagnant and diminishing government resources
(Table 6).
The new NCI NCTN offers the promise of a more efficient
clinical trials system that will translate the rapidly expand-
I. Network Membership
High volume sites versus broad membership inclusion
Single versus multiple group membership
Trial accrual credit assignment
II. Site Trial Portfolio Management
Access to all four groups trials
Orphan or less common disease trial resource allocation
II. Volunteer Investigator Time
Engagement
Promotion
Value
III. Reimbursement
Tiered per case reimbursement
Anticipated static or decreasing National Cancer Institute budget
ing scientific knowledge of cancer to practice and provide

our patients and their families with innovative cancer management strategies in a more timely fashion. Surely our
community clinical investigators will embrace NCTN and
contribute to its success.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Stephen S. Grubbs*
147
STEPHEN S. GRUBBS
REFERENCES
1. Nass J, Moses H, Mendelsohn J (eds). A National Cancer Clinical Trials
System for the 21st Century: Reinvigorating the NCI Cooperative Group
Program. Washington, DC: National Academies Press, 2010.
2. Abrams J. Transforming the NCI Trials System. 14th Meeting of the
148
Clinical Trials and Translational Research Advisory Committee. July 13, 2011.
3. Mack A, Nass S. Implementing a National Cancer Clinical Trials System
for the 21st Century: Workshop Summary. Washington, DC: National Academies Press, 2011.
Successful Integration of Cooperative

Groups: The Origin of the Childrens
Oncology Group
By Gregory H. Reaman, MD
Overview: In March 2000, the four legacy pediatric cooperative groups officially merged to become the Childrens Oncology Group (COG). This was accomplished by the ratification of
a new constitution by the respective executive committees
and voting membership of the four legacy groups. The actual
merger was preceded by a 12 to 18 month period of planning,
negotiation, and transition, overseen by a Transition Committee of select executive leadership under the direction of the
four current chairs of the existing pediatric groups. Despite
the constant threat of budget reductions and questions related to the judicious use of National Cancer Institute (NCI)
funds to support four pediatric groups when children constitute only 3% of the US cancer problem, the decision to unify
was initiated and driven internally. The merger was envisioned
as an opportunity to create efficiency by reducing duplicative
systems and processes, which was becoming increasingly
apparent as more planned clinical trials required intergroup
collaboration. It was also recognized that such intergroup
efforts would become more of a reality as clinical trial para-
ECENT DECISIONS to restructure the NCIs clinical

trial enterprise in accordance with recommendations
from the Institute of Medicine and external advisory groups
have intended to reduce the number of cooperative groups
through a process of integration. This integration of legacy
groups and reorganization of the nations cancer clinical
trials program has resulted in considerable angst for both
well-established clinical investigators and those beginning
their careers. All are now transitioning to a new reality of
team science and collaboration in clinical cancer research.
The well-intentioned motivation to move a somewhat fragmented, competitive, duplicative system with parallel and
under-resourced infrastructures to a system where scientific
discovery can be rapidly translated to design, implementation, and conduct of high-impact studies and dissemination
of results to improve care standards cannot be questioned.
Nonetheless, reasonable concerns remain: the proposed
model of centralized infrastructure supportincluding critical information systems, uncertainty about future budget
allocations and funding instruments, and the lack of concrete metrics by which to judge the performance and success
of this reorganization continue during a period of intense
transition. The anxiety of the integrated group leadership
and their membership is understandable and familiar.
Although the current changes are decidedly more far
reaching in scope, this is not the first integration or merger
within NCIs Cooperative Group Program. In 2000, the
Childrens Cancer Group, Pediatric Oncology Group (POG),
National Wilms Tumor Study Group, and Intergroup Rhabdomyosarcoma Study Group merged to become COG. More
than two decades before this, the pediatric divisions of the
Cancer and Leukemia Group B and the Southwest Oncology
Group came together to form POG.1,2 With the consolidation of the pediatric cooperative groups, well-recognized
names and acronyms disappeared, but sense of mission and
a strong foundation arising from legacy group accomplishments provided a framework for success. Naysayers may
digms were built on risk-adjusted approaches. Clinically,

biologically, and molecularly defined homogeneous subgroups
of patients were of insufficient sample size within each group
to design and conduct studies within a reasonable time frame.
In essence, this merger was motivated by an overwhelming
sense of necessity to preserve our mission of defining and
delivering compassionate and state-of-the-art care through
scientific discovery. The merger process itself was challenging, time consuming, not supported by any supplemental
funding, and at times painful. What has emerged as a result
is the largest pediatric cancer research organization in the
world. Accomplishments in epidemiology, biology, translational science, and improved clinical outcomes for some
pediatric cancers would have never been achieved without the
merger. The very fact that outcome improvements were not
realized in every type of pediatric cancer is testimony to the
commitment of the COG membership to continue to look and
move forward.
comment that pediatric oncology is now different. In fact,

pediatric oncology is different; the overall 5-year event-free
survival rates for children with cancer exceed 80%. The
clinical management of childhood cancer is integrally linked
to clinical and translational research. Improvement in survival outcomes and cure rates are, in large part, a direct
result of the fact that the overwhelming majority of children
with cancer are enrolled on clinical trials.3
The decision to unify the pediatric groups was conceived
and driven internally by a perceived affront to their missions
and a very real threat to their scientific agenda.2 This
resulted from the increasing focus on risk-adjusted therapy
approaches for childhood cancer and the increasing number
of subgroups defined by clinical and biologic prognostic
factors. Designing phase III clinical trials in homogenous
patient groups within each pediatric cancer diagnosis that
could be completed within reasonable timelines was severely
hampered by sample size constraints initially mandating
intergroup collaborations. Now such trials increasingly require consideration of international group collaborations.
Also, any future plans to explore the feasibility of personalized therapy approaches with molecularly targeted
agents in pediatric cancers would mandate a coordinated
and consolidated effort. Infrastructure duplication and lack
of process and procedure harmonization even within the
pediatric groupswere exaggerated by continual budget
shortfalls. Despite the expectation that consolidation of the
pediatric groups would result in improved efficiencies and
From the Center for Drug Evaluation and Research, US Food and Drug Administration,
Silver Spring, MD, and the Division of Oncology, Childrens National Medical Center,
Washington, DC 20010.
Address reprint requests to Gregory H. Reaman, MD, Childrens National Medical
Center, Washington, DC 20010; email: greaman@childrensnational.org.
1092-9118/10/1-10
149
GREGORY H. REAMAN
economies of scale, no predetermined metrics were developed and no resources were made available to collect and
analyze data to evaluate success. In fact, annual budget
reductions after the merger precluded any comprehensive,
detailed self-assessment.
There were considerable immediate challenges in the
transition process of unifying culturally and organizationally distinct, procedurally and operationally disparate, competing clinical trial networks. COG leadership focused
initially on individual, discipline-specific, and institutional
membership requirements; financial planning and new models for funds distribution (given the planned discontinuation
of institutional U-10 awards); consolidation of administrative functions; and coordination of data management and
statistical operations. The last of these was particularly
problematic since three separate sites served as statistics
and data management centers using different and noncompatible, internally generated, electronic, Web-based remote
data entry and registration systems. The importance of
providing adequate and centralized IT support and an
enterprise-wide Clinical Data Management System cannot
be overemphasized.
It was a priority to maintain active legacy study conduct
(80 phase III, 22 phase II, and 7 phase I) open studies using
different data management platforms while planning for the
importation of all legacy as well as new data to a single
common platform. Of equal and given resource constraints competing priority was the development and implementation of new biologically based and hypothesisdriven clinical trials.
Despite the lack of prespecified metrics, many successes
were realized as a direct result of the consolidation. Three
years after the merger, annual enrollment on therapeutic
studies increased by more than 25% and enrollment on
KEY POINTS
The merger of the legacy pediatric cooperative groups

was self-motivated and self-directed.
No preestablished metrics of success were developed
other than to maintain focus and achieve mission
success.
Supplemental resources during a challenging transition period were not made available.
The lack of supplemental resources precluded the
Childrens Oncology Groups ability to perform a
detailed assessment of evaluating efficiency, and progressive budget reductions made evaluation of any
cost-effectiveness impossible.
Significant accomplishments in pediatric cancer were
made possible only as a direct result of its planned
consolidation and successful integration.
nontherapeutic (epidemiology and biology) studies more

than doubled compared with combined enrollment statistics
of the legacy groups.4 COG became the largest childhood
cancer research organization in the world and expanded its
international collaborations. New clinical and biologically
based classification systems were developed for both acute
lymphoblastic and myeloid leukemia and neuroblastoma.5,6,7 Cytogenetic and molecular genetic predictors of
outcome were defined for these same diseases as well as for
Wilms tumor and medulloblastoma. The prognostic significance of minimal residual disease in a number of tumor
systems (acute leukemia, lymphoma, neuroblastoma) was
established and incorporated into the expansion of riskadjusted treatment approaches to these diseases.6,7 Expanding active programs in biomarker development for patient
enrichment and risk classification was made possible by
the retrospective analysis of independent data sets from
legacy groups as test and validation cohorts. Early in its
history, COG opened a frontline, randomized clinical trial
for osteosarcoma, designed and conducted in collaboration
with multiple European cooperative groups, which led to the
development of the Cancer Therapy Evaluation Program
guidelines for performance of international clinical trials.
By consolidating legacy bio-specimen banks, COGs Biopathology Center has become the national and international
resource for clinically well-annotated specimens of virtually
every type of pediatric cancer. Important genomic investigations in collaboration with the NCIs Cancer Genome
Atlas and the TARGET initiative have been made possible
through this resource. These investigations have identified
novel genetic lesions in acute leukemia and neuroblastoma,
which are being evaluated as predictive biomarkers and
for their potential therapeutic exploitation.7-11 Since the
consolidation, statistically significant improvements in
5-year event-free survival rates have been achieved in all
prognostic subgroups of acute lymphoblastic leukemia,
acute myeloid leukemia, high-risk neuroblastoma, and medulloblastoma.7,8,12,13 COG has developed Long-Term
Follow-Up Guidelines for childhood cancer survivors, which
are regularly updated in accordance with increasing evidence and utilized worldwide.7 Relevant portions of these
exposure-dependent risk assessments and recommended
surveillance parameters have been extrapolated for adult
cancer survivors.
Despite the initial challenges and anxiety experienced
during the early days of the consolidation, and despite the
criticism in response to the reduction in healthy competition, COG has emerged as more successful and impactful
than the sum of its legacy groups. With thoughtful transition, focus on mission, assurance of adequate centralized
resources and utilization of well-constructed, predetermined
metrics, the current cooperative group reorganization
should be viewed as a real opportunity to further advance
clinical and translational research.
Author
Gregory H. Reaman*
150
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
COG: GIVING NEW MEANING TO COOPERATIVE
REFERENCES
1. Reaman GH. Pediatric oncology; current views and outcomes. Pediatr
Clin Nort Am. 2002;49:1305-1318.
2. Reaman GH. Clinical advances in pediatric hematology and oncology:
Cooperative group research. Clin Adv Hematol Oncol. 2005;3:133-135.
3. Anderson BD, Smith MA, Reaman GH, et al. Views of American
oncologists about the purposes of clinical trials. J Natl Cancer Inst. 2003;95:
630-631.
4. Steele JR, Wellemeyer AS, Hansen MJ, et al. Childhood cancer research
network; A North American pediatric cancer registry. Cancer Epidemiol
5. Schultz KR, Pullen DJ, Sather HN, et al. Risk and response based
classification of childhood B precursor acute lymphoblastic leukemia: A
combined analysis of prognostic markers from the Pediatric Oncology Group
(POG) and the Childrens Cancer Group (CCG). Blood. 2007;109:926-935.
6. Borowitz MJ, Devidas M, Hunger S, et al. Clinical significance of
minimal residual disease in childhood acute lymphoblastic leukemia and its
relationship to other prognostic factors: A Childrens Oncology Group study.
Blood. 2008;111:5477-5485.
7. OLeary M, Krailo M, Anderson JR, et al. Progress in childhood cancer:
50 years of research collaboration: A report from the Childrens Oncology

Group. Semin Oncol. 2008;35:484-493.
8. Hunger SP, Devidas M, Camitta B, et al. Improved survival for children
with acute lymphoblastic leukemia (ALL) from 1990 2005: A report from the
Childrens Oncology Group. J Clin Oncol. In press.
9. Hunger SP, Raetz EA, Loh ML, et al. Improving outcomes for high-risk
ALL: Translating new discoveries into clinical care. Pediatr Blood Cancer.
2011;56:984-993.
10. Mullighan CC, Su X, Zhang J, et al. Deletion of IZKF1 and prognosis in
acute lymphoblastic leukemia. N Engl J Med. 2009;360:470-480.
11. Mullighan CC, Zhang J, Harvey RC, et al. JAK mutations in high risk
childhood acute lymphoblastic leukemia. Proc Natl Acad Sci U S A. 2009;106:
914-918.
12. OLeary MC, Reaman GH. Principles of pediatric oncology. Hong WK,
Bast RC, Hait W, et al (eds). Cancer Medicine 8th Edition. Shelton, CT: BC
Decker, 2010, pp. 1723-1741.
13. Smith MA, Seibel NL, Altekruse SF, et al. Outcomes for children and
adolescents with cancer: Challenges for the 21st century. J Clin Oncol.
2010;28:2625-2634.
151
OVERCOMING DISPARITIES IN CLINICAL TRIAL

ACCRUAL AMONG AFRICAN AMERICANS
CHAIR
Sandra M. Swain, MD
Washington, DC
SPEAKERS
Jean G. Ford, MD
Johns Hopkins School of Medicine
Baltimore, MD
Deliya R. Banda, PhD
Washington, DC
Worta McCaskill-Stevens, MD
Bethesda, MD
A Critical Review of the Enrollment of Black

Patients in Cancer Clinical Trials
By Deliya R. Banda, PhD, Diane St. Germain, RN, MS, Worta McCaskill-Stevens, MD,
Jean G. Ford, MD, and Sandra M. Swain, MD
Overview: Although clinical trials represent a vital opportunity for improvements in cancer treatment, data show that a
small proportion of patients with newly diagnosed cancer
participate in clinical research. Black patients continue to
have a worse prognosis for most cancers compared with other
patients of other races/ethnicities. Racial/ethnic- and agerelated disparities in clinical trial accrual are also well documented. The recruitment and retention of minorities in these
trials present an even greater challenge despite regulatory
efforts and initiatives to increase representation. Treatment
data from homogenous populations prevent us from under-
CCRUAL TO clinical trials remains a longstanding

challenge, with 5% to 10% of patients with cancer from
the general population participating and even lower rates of
black patients participating.1 For the general population,
the reasons are numerous, including lack of trial availability, lack of physician engagement and awareness of available trials, ineligibility, lack of insurance coverage, time
commitment, and lack of knowledge and misperceptions
regarding the conduct of clinical trials.2-5 In addition to
these factors, a multitude of additional subgroup-specific
reasons have been reported, including a mistrust of the
research and medical system, awareness of historical clinical research abuse, a lack of access to state-of-the-art care, a
higher incidence of comorbidities (resulting in higher rates
of ineligibility), religious and cultural beliefs that influence
attitudes toward clinical trials, economic issues, and logistical concerns, such as child care and transportation.6-11
Cancer rates and incidence are disproportionately higher
for some cancers in black patients compared with white
patients.12 Inclusion of black patients in clinical trials is
therefore vital to gain an understanding of cancer biology
and response to treatment in this population and to provide
access to state-of-the-art care. Lack of access to state-of-theart care contributes to advanced disease at diagnosis, treatment morbidity, and decreased survival.13,14 It may also
increase the rate of comorbidities that deem many of this
population ineligible for clinical trials. Among 235 black
patients screened for protocol eligibility, only 8.5% were
eligible. Patients were deemed ineligible because of comorbidities (with respiratory failure, HIV positivity, and anemia
accounting for most), advanced disease stage, poor performance status, premature death, and short life expectancy.9
It is imperative that the rate of accrual of underserved
populations increase to maximize the generalizability of
results from clinical trials. Several efforts have been made
to increase accrual of underrepresented populations to
clinical trials. The National Institutes of Health Revitalization Act of 1993 mandates the inclusion of women and
minorities in federally sponsored cancer clinical trials. The
aim is to conduct subset analyses to determine whether
there are differences in effect based on race and sex. The
National Cancer Institute (NCI) established the MinorityBased Community Clinical Oncology Program in 1990,
standing therapeutic response and the true safety profile of

novel therapies. Patient-, physician-, and system-level factors
that affect trial participation have been extensively studied.
However, years of accrual data remain largely unchanged,
suggesting the challenge lies in effectively addressing these
factors. Furthermore, data showing that black patients tend to
have more advanced stage cancers at the time of diagnosis in
fact beg their overrepresentation on clinical trials. An inability
to successfully enroll diverse populations in clinical trials only
exacerbates racial/ethnic differences in cancer treatment and
survivorship.
which supports institutions with 40% minority population in

their catchment area to accrue racial minorities to treatment, cancer control, and prevention trials. This program
has had successful accrual rates of 60% (Table 1). The
National Medical Association (NMA) developed I.M.P.A.C.T.
(Increase Minority Participation and Awareness of Clinical
Trials), an initiative to increase awareness, knowledge, and
participation of black physicians and patients in biomedical
research and clinical trials. There has been an increase in
minority accrual to clinical trials sponsored by the NCI
(Fig. 1). This may be attributable to enhanced efforts or an
increase in minority populations.
The University of CaliforniaDavis Cancer Center has
recently embarked on a national effort to boost clinical trial
recruitment of minorities as part of a major grant from the
National Center on Minority Health and Health Disparities.15 This project, EMPaCT (Enabling Minority Participation in Clinical Trials), will engage five other centers and
assess existing efforts to accrue minorities into trials. The
goal is to develop consensus on evidence-based accrual
models that can be adopted by cancer centers across the
country.
There has been a heightened understanding of the issues
as evidenced by the plethora of published studies documenting the barriers to clinical trial enrollment in underserved
populations.3,4,13,16-18 However, methods to increase minority accrual elude clinicians largely because of the lack of
evidence-based strategies and practices in the literature.
One of the most extensive literature reviews identified only
14 articles that examined strategies to accrue underrepresented populations to cancer clinical trials. Notably, the
efficacy or effectiveness of the strategies was evaluated in
five of the 14 studies.19 The first was a randomized two-arm
study comparing the use of a media campaign involving
newspapers and fliers with a clinic registry strategy where
From the Washington Cancer Institute at Medstar Washington Hospital Center, Washington, DC; Medstar Health Research Institute, Hyattsville, MD; Division of Cancer
Prevention, National Cancer Institute, Bethesda, MD; Johns Hopkins Center to Reduce
Cancer Disparities, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD.
Address reprint requests to Deliya R. Banda, PhD, 110 Irving St. NW, Suite CG-111,
Washington, DC 20010; email: Deliya.R.Banda@MedStar.net.
1092-9118/10/1-10
153
BANDA ET AL
Table 1. Minority-Based Community Clinical Oncology Program (MB-CCOP) Accrual, 2000 2009
Fiscal Year
No. of
Funded MB-CCOPs
No. of
Treatment Accruals
No. of Prevention
and Control Accruals
No. of
Overall Accruals
No. of
Minority Patients
Overall Minority
Participation, %
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Total (20002009)
8
10
11
11
13
13
13
14
13
14
14
425
642
567
521
673
709
684
805
895
851
851
358
541
682
930
467
428
393
776
733
461
461
783
1183
1249
1451
1140
1137
1077
1581
1628
1312
1312
427
672
949
1249
718
569
612
962
1051
830
830
55
57
76
86
63
50
57
61
65
63
63
Courtesy of Worta McCaskill-Stevens, MD, MS, program director, MB-CCOP, National Institutes of Health.
women were invited to participate in person. The media

campaign proved more effective as evidenced by more
women consenting to participate and fewer no-shows in this
group compared with the clinic registry arm.20 In the second
study, a randomized study of older black men in a screening
trial, three progressively intensive recruitment methods
were compared. Intervention arms received different combinations of standard compared with enhanced mailings; baseline information gathered via telephone, via mail, or during
a church session; reminders by telephone or mail; and
consent forms mailed or administered during a church
session. The most intensive of the intervention arms yielded
the higher enrollment compared with the control and other
intervention arms.21 In the third study, enrollment in worksite cancer prevention programs was assessed comparing
passive (telephone contact using telephone numbers provided by the company) with active recruitment strategies (a
signup sheet). Enrollment and retention were lower among
employees recruited by passive methods than by active
methods. Passive methods, however, resulted in a representative employee sample.22 In the fourth study, the use of
minority outreach recruiters, a specialized recruitment
manual, and engagement of consultants for minority recruit-
KEY POINTS
154
Accrual to cancer clinical trials remains low and

largely unchanged despite initiatives and interventions to increase rates beyond the estimated 5% to
10%.
In the face of poorer prognosis, even lower rates of
accrual of minorities in cancer trials is compounded
by cultural-specific attitudes and sociodemographic
factors that must be addressed in addition to the
factors identified for other patients with cancer.
An intervention approach that directly addresses
attitudes through use of a culturally appropriate
video has been shown to affect the intention of black
patients with cancer to enroll in therapeutic clinical
trials.
Trial design factors, including eligibility criteria,
data collection, and quality, are among the systemwide factors that also must be addressed to improve
accrual.
ment failed to yield increases in recruitment of minority

participants in a prostate cancer prevention trial.23 In the
final study, from the Southeast Cancer Control Consortium,
no improvement was found in enrollment of patients with
cancer into clinical trials after their intervention. This
intervention involved use of nurse facilitators, quarterly
newspapers, and health educators.24
The problem is clearly complex, with multiple contributing factors. Developing strategies that target factors that
yield the greatest effect is thus a challenge. Among black
patients, this is further complicated by the need to ensure
such strategies are culturally sensitive. This review describes barriers to minority clinical trial accrual, a novel
approach to change black patients attitudes toward clinical
trial participation and strategies to enhance accrual.
Factors That Influence Black Patients Decisions to
Participate in Cancer Clinical Trials
A comprehensive review identified 150 distinct barriers to

accrual of underrepresented populations to cancer-related
clinical trials. The most frequently reported of these are
presented in Table 2.13 Often, barriers are categorized
according to their source (e.g., patient, physician, or institution or environment). A conceptual framework was developed that proposes that these barriers may influence accrual
through their effects on awareness of trials, the opportunity
to participate, and the decision to accept or refuse participation. The most frequently reported barrier related to
awareness of trials was lack of clinical trials education.
Education is crucial and can address other barriers, such as
mistrust of research and the medical system, perceived
harm, and fear. The most frequently reported barrier related
to opportunity to participate was physician attitudes and
patient eligibility. The most frequently reported barrier to
acceptance of enrollment was mistrust of research and the
medical system and costs of participation.13
Black Patients Attitudes Toward Clinical
Trial Participation
Many of the attitudes black patients have toward clinical

trial participation stem from historical abuses, such as the
U.S. Public Health Services Syphillis Study at Tuskegee,
and related concerns of ethical misconduct and mistrust.6,25-27
In focus group interviews with 33 black patients and a black
moderator, most participants viewed clinical research negatively and expressed concern they would be treated as a
guinea pig. They also believed black patients would not
BLACK PATIENTS AND CANCER CLINICAL TRIALS
Fig. 1. Minority enrollment in National Cancer Institute clinical

trials, 2000 2010. Courtesy of Worta McCaskill-Stevens, MD,
MS, program director, Minority Based Community Clinical Oncology Program, National Institutes of Health.
Abbreviations: CTEP, Cancer Therapy Evaluation Program; DCP,
Division of Cancer Prevention; RRP, Radiation Research Program.
benefit from clinical research because of racism and the

inability to pay for medical care.7
Participants were asked specifically about the U.S. Public
Health Services study and were found to have misinformation and misperceptions about the study. Many believed it
was a conspiracy and, as a result, expected that investigators would be dishonest and not provide full disclosure of the
risks related to research. They also believed in other conspiracies, such as the creation of HIV, military experiments with Agent Orange, and Central Intelligence Agency
distribution of crack cocaine in black communities.7
In another set of focus groups with black women, the
theme of trust also emerged.28 Women were suspicious of
how clinical research was funded and expressed concerns
regarding the termination of programs that specifically help
the black community.28 Focus group participants were presented with a hypothetical phase III randomized clinical
trial in which adjuvant oral and intravenous chemotherapy
were compared. Some participants did not understand the
concept of randomization and lost interest in participating
because of a lack of choice of treatment arm. Other participants, especially older women, were in favor of natural
treatments and stressed the importance of prayer or God as
healer. Most women were willing or at least open to considering a clinical trial using an oral chemotherapy agent.
Physician Attitudes Toward Enrolling Black Patients
in Clinical Trials
Physicians play an important role in clinical trial accrual.

Their perceptions and support of clinical research influence
the discussions they have with patients and consequently
the patients decision to participate in clinical trials. A
survey of 166 black physicians revealed several challenges of
Table 2. Frequently Reported Patient Barriers to Minority
Accrual to Clinical Trials 13
Mistrust of research and medical system
Perceived harm
Cost
Patient demographics
Transportation
Lack of education regarding clinical trials
Time commitment
Fear
Family issues
participating in clinical research,29 including additional

paperwork or telephone calls, masked drug assignment,
excess patient care costs, losing the patient from their
medical practice, and the potential effect of clinical research
on the managed care status of their medical practice. Although most physicians surveyed (97%) believed clinical
trials were important, only 63% referred patients and 64%
encouraged patients to participate in clinical trials. This
validated the focus group reports of mistrust of the medical
system, lack of awareness of clinical trials, communication
with study staff, and economic issues as factors that play a
role in minority participation in clinical trials.7,28,29
The Eastern Oncology Cooperative Group (ECOG) conducted a pilot project with the NMA to determine whether
an outreach intervention program would increase minority
accrual.30 The program consisted of five physician workshops focused on the identification of physician and patient
barriers to minority accrual, potential solutions to address
these barriers, and the willingness of the participating
physicians to refer their patients to cancer clinical trials.
The two most frequent physician barriers for the NMA
physicians were lack of awareness and information about
clinical trials and lack of trust of the medical center sponsoring the trial. Many had referred patients to a medical
center but received little to no information about them
thereafter; hence, this generated fear of losing their patients. Moreover, this lack of communication was viewed
as a lack of respect for minority physicians. The ECOGaffiliated physicians cited these barriers less frequently.
Regarding patient barriers, both physician groups cited
patient or family suspicion and fear. All agreed to the need
for patient education materials specifically designed for
minority patients. ECOG subsequently developed a culturally sensitive patient educational brochure. In addition, a
physician newsletter and laminated pocket cards were developed to provide clinical trial information, information
regarding the importance of minority participation in cancer
clinical trials, how to enroll patients in cancer clinical trials,
and other information that would enhance communication
between the NMA and ECOG physicians.
Additional Strategies to Enhance Accrual of
Black Patients
In addition to addressing patient and physician barriers,

it is crucial to examine clinical trial design and patient
155
BANDA ET AL
eligibility, particularly as cancer clinical trials increase in

complexity. Comorbidities are a substantial barrier to black
patient participation in clinical trials. It remains unclear
whether inferior treatment outcomes for minority patients
are caused by the comorbidity or drug and/or dose alterations because of the comorbidity. Lee et al33 provided a
review of chemotherapy trials to determine the effect of
comorbidities on treatment outcomes and survival. Most
studies included in the review were based on retrospective
cancer registry and administrative data. Patients with comorbidities were less likely to receive chemotherapy regardless of cancer type and stage. Most trials reported decreased
survival for patients who had comorbidities compared with
those who did not. However, both chemotherapy use and
survival was not evaluated in any of the studies; hence, a
correlation between the two cannot be made. Major limitations of the review were the heterogeneity of the studies and
the poor quality of the data.
Collection of comorbidity data lacks uniformity and is
often insufficiently detailed.33 Furthermore, there is a lack
of assessment tools, heavy reliance on performance status,
and lack of management guidelines for specific comorbidities. There may be comorbidities that are modifiable before
enrollment and patients for whom modifications of eligibility
criteria is appropriate to increase enrollment. It may be that
patients with certain comorbidities could be included in cancer
clinical trials if the patient is closely monitored and with
strict management guidelines in place. However, the appropriate collection of data is essential to support these efforts.
In addition to comorbidity data, there is an overriding
need to uniformly collect data that will encourage crossstudy analyses and be used to develop a databank for future
queries. In a conversation with K. Flaherty, he indicated
that recommendations include the collection and inclusion
of race/ethnicity, age, educational level, income, zip code,
marital status, smoking, obesity, spoken language, and
insurance (October 2011). These data should be selfreported and collected in a standardized manner. However,
there are several challenges, including who collects the data,
the format used, commitment level from institutional administrators, resources and support services, staff training,
implementation planning, and data quality assurance.34
Consideration must be given to the feasibility of enrolling
minority and other underserved populations to a clinical
trial during its design. Factors that must be considered are
identifying unnecessarily restrictive eligibility criteria, identifying important end points that would address research
questions geared toward a minority population, and determining additional data to collect that may inform future
clinical care of minority populations. Additional considerations include participant burden, such as frequent and
inconvenient clinic visits and assessments that prohibit
enrollment by those patients with transportation issues,
work, or family commitments. Finally, engaging minority
physicians and community representatives during design
development is also key to the viability of minority accrual.
A Novel Approach: Addressing Culturally Specific
Attitudes Toward Clinical Trials
Lack of awareness or understanding of clinical trials is

arguably one of the easier barriers to address. In a NCIs
156
2007 Health Information National Trends Survey, black and

Hispanic respondents placed the greatest trust in information from radio, television, newspapers, and religious organizations compared with white respondents, who placed
greater trust in physicians.26 Hence, the use of media that
is culturally and ethnically sensitive to enhance awareness and promote clinical trials is a viable approach. In their
review, Lai et al19 reported on three studies that effectively
used media campaigns and church-based project sessions.
For black patients with cancer, the literature identifies a
range of attitudinal barriers to trial participation.3,6,25,31
With such well-documented attitudinal barriers, it seems
reasonable to assume the importance of integrating and
addressing such sociocultural concerns in any effective approach to increasing accrual.
A recent pilot study aimed to address the specific attitudes
of black patients toward clinical trials.32 The goal was to
affect specific attitudes and increase the intention of patients to enroll in a therapeutic trial. The approach involved
developing and testing a 15-minute, culturally targeted,
narrative video to address six documented attitudinal barriers to trial participation by black patients. These six
attitudes were the concern about the ethical misconduct of
investigators, poor treatment for being a minority or economically disadvantaged, fear and distrust of the medical
establishment, worry about loss of autonomy, concern about
privacy, and a general lack of knowledge and awareness
about clinical trials.3,6,25,31
One hundred eight black patients with cancer in active
treatment were enrolled in the pilot intervention at a large,
urban cancer institute. Patients had never previously participated in any type of research study or clinical trial.
Attitudes and self-reported intention to enroll in a clinical
trial were assessed before and immediately after viewing the
video. Results showed changes in all six attitudinal barriers
from before the video to after the video. Patients likelihood
of enrolling in a clinical trial increased after watching the
video, with 34% of the patients showing positive changes in
intention.32
The results of this study suggest that effectively addressing myths and strongly held attitudes toward trials, particularly among black patients, is a key patient-level approach
to improving accrual. Despite the modest sample size, patients with more than 12 different primary tumor types were
included, suggesting the ability to generalize the effectiveness of this video and its potential utility to increase willingness to participate across all cancer types and varied
therapeutic trials. This intervention presents a practical
approach for delivering clinical trial information in a culturally sensitive and effective format.
Conclusion
The extra time needed to gain individual and community

acceptance of clinical trials among minorities is a known
prerequisite to affect their participation in clinical trials.34
It is crucial to devote resources and continue research efforts
to accrue minorities to clinical trials, particularly in the face
of the changing demographics in the United States, a changing health care system, and a shifting economic climate.
With an anticipated increase in cancer incidence for minor-
BLACK PATIENTS AND CANCER CLINICAL TRIALS
ity populations far greater than the corresponding increase

for white patients,35 it is imperative that the inherent selection
bias and objectivity of clinical trial design be addressed.9
Acknowledgment
Supported by National Center on Minority Health and Health
Disparities: 1RC1MD004185-01.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Deliya R. Banda*
Diane St. Germain*
Worta McCaskill-Stevens*
Jean G. Ford*
Sandra M. Swain
BiPar Sciences
(U); Eisai (U);
Nektar (U);
Novartis (U);
Roche/Genentech
(U); Sanofi (U)
Agendia; BiPar
Sciences;
Bristol-Myers
Squibb; Novartis;
Pfizer; Roche/
Genentech;
Sanofi
Sanofi
REFERENCES
1. Baseline Study of Patient Accrual onto Publicly Sponsored US Cancer
Clinical Trials: An Analysis Conducted for the Global Access Project of the
National Patient Advocate Foundation. Philadelphia, PA: Coalition of Cancer
Cooperative Groups; 2006.
2. Comis RL, Miller JD, Aldige CR, Krebs L, Stoval E. Public attitudes
toward participation in cancer clinical trials. J Clin Oncol. 2003;21:830-835.
3. Advani AS, Atkeson B, Brown CL, et al. Barriers to the participation of
African-American patients with cancer in clinical trials: a pilot study. Cancer.
2003;97:1499-1506.
4. Lara PN Jr, Paterniti DA, Chiechi C, et al. Evaluation of factors affecting
awareness of and willingness to participate in cancer clinical trials. J Clin
Oncol. 2005;23:9282-9289.
5. Simon MS, Du W, Flaherty L, et al. Factors associated with breast
cancer clinical trials participation and enrollment at a large academic medical
center. J Clin Oncol. 2004;22:2046-2052.
6. Corbie-Smith G. The continuing legacy of the Tuskegee Syphilis Study:
considerations for clinical investigation. Am J Med Sci. 1999;317:5-8.
7. Corbie-Smith G, Thomas SB, Williams MV, Moody-Ayers S. Attitudes
and beliefs of African Americans toward participation in medical research.
J Gen Intern Med. 1999;14:537-546.
8. Swanson GM, Ward AJ. Recruiting minorities into clinical trials: toward
a participant-friendly system. J Natl Cancer Inst. 1995;87:1747-1759.
9. Adams-Campbell LL, Ahaghotu C, Gaskins M, et al. Enrollment of
African Americans onto clinical treatment trials: study design barriers. J Clin
Oncol. 2004;22:730-734.
10. Shavers-Hornaday VL, Lynch CF, Burmeister LF, Torner JC. Why are
African Americans under-represented in medical research studies? Impediments to participation. Ethn Health. 1997;2:31-45.
11. Baquet CR, Commiskey P, Daniel Mullins C, Mishra SI. Recruitment
and participation in clinical trials: socio-demographic, rural/urban, and
health care access predictors. Cancer Detect Prev. 2006;30(1):24-33.
12. American Cancer Society. Cancer Facts & Figures for African Americans 2011-2012. Chicago, IL: American Cancer Society; 2011.
13. Ford JG, Howerton MW, Lai GY, et al. Barriers to recruiting underrepresented populations to cancer clinical trials: a systematic review. Cancer.
2008;112:228-242.
14. Shavers VL, Brown ML. Racial and ethnic disparities in the receipt of
cancer treatment. J Natl Cancer Inst. 2002;94:334-357.
15. Broadening Recruitment for Minorities, the Elderly. Cancer Discovery.
2011;1:461.
16. Branson RD, Davis K Jr, Butler KL. African Americans participation
in clinical research: importance, barriers, and solutions. Am J Surg. 2007;
193:32-39.
17. Townsley CA, Selby R, Siu LL. Systematic review of barriers to the
recruitment of older patients with cancer onto clinical trials. J Clin Oncol.
2005;23:3112-3124.
18. Mouton CP, Harris S, Rovi S, Solorzano P, Johnson MS. Barriers to
black womens participation in cancer clinical trials. J Natl Med Assoc.
1997;89:721-727.
19. Lai GY, Gary TL, Tilburt J, et al. Effectiveness of strategies to recruit
underrepresented populations into cancer clinical trials. Clin Trials. 2006;3:
133-141.
20. Brewster WR, Anton-Culver H, Ziogas A, et al. Recruitment strategies

for cervical cancer prevention study. Gynecol Oncol. 2002;85:250-254.
21. Ford ME, Havstad SL, Davis SD. A randomized trial of recruitment
methods for older African American men in the Prostate, Lung, Colorectal
and Ovarian (PLCO) Cancer Screening Trial. Clin Trials. 2004;1:343-351.
22. Linnan LA, Emmons KM, Klar N, Fava J, LaForge R, Abrams D.
Challenges to improving the impact of worksite cancer prevention programs:
comparing reach, enrollment, and attrition using active versus passive
recruitment strategies. Ann Behav Med. 2002;24:157-166.
23. Moinpour CM, Atkinson JO, Thomas SM, et al. Minority recruitment in
the prostate cancer prevention trial. Ann Epidemiol. 2000;10(8, suppl 1):S85S91.
24. Paskett ED, Cooper MR, DeGraffinreid CR, Spurr CL, Dolecek TA.
Community Clinical Oncology Program as a recruitment vehicle for cancer
control research. The Southeast Cancer Control Consortium experience. N C
Med J. 1995;56:283-286.
25. Scharff DP, Mathews KJ, Jackson P, Hoffsuemmer J, Martin E,
Edwards D. More than Tuskegee: understanding mistrust about research
participation. J Health Care Poor Underserved. 2010;21:879-897.
26. Katz RV, Kegeles SS, Kressin NR, et al. Awareness of the Tuskegee
Syphilis Study and the US presidential apology and their influence on
minority participation in biomedical research. Am J Public Health. 2008;98:
1137-1142.
27. Katz RV, Green BL, Kressin NR, et al. The legacy of the Tuskegee
Syphilis Study: assessing its impact on willingness to participate in biomedical studies. J Health Care Poor Underserved. 2008;19:1168-1180.
28. Linden HM, Reisch LM, Hart A Jr, et al. Attitudes toward participation
in breast cancer randomized clinical trials in the African American community: a focus group study. Cancer Nursing. 2007;30:261-269.
29. Shavers VL, Lynch CF, Burmeister LF. Factors that influence AfricanAmericans willingness to participate in medical research studies. Cancer.
2001;91(1 suppl):233-236.
30. McCaskill-Stevens W, Pinto H, Marcus AC, et al. Recruiting minority
cancer patients into cancer clinical trials: a pilot project involving the Eastern
Cooperative Oncology Group and the National Medical Association. J Clin
Oncol. 1999;17:1029.
31. Chandra A, Paul DP 3rd. African American participation in clinical
trials: recruitment difficulties and potential remedies. Hosp Top. 2003;81:3338.
32. Banda D, Wang H, Libin AV, Swain SM. A clinical trial with culturally
tailored video to increase participation of African Americans in cancer clinical
trials. Oncologist. 2012;(In Press).
33. Lee L, Cheung WY, Atkinson E, Krzyzanowska MK. Impact of comorbidity on chemotherapy use and outcomes in solid tumors: a systematic
review. J Clin Oncol. 2011;29:106-117.
34. Probstfield JL, Frye RL. Strategies for recruitment and retention of
participants in clinical trials. JAMA. 2011;306:1798-1799.
35. Smith BD, Smith GL, Hurria A, Hortobagyi GN, Buchholz TA. Future
of cancer incidence in the United States: burdens upon an aging, changing
nation. J Clin Oncol. 2009;27:2758-2765.
157
ANTIBODY CONJUGATES: TARGETED THERAPY

MEETS CHEMOTHERAPY IN ONE DRUG
CHAIR
Howard A. Burris III, MD
The Sarah Cannon Research Institute
Nashville, TN
SPEAKERS
John P. Leonard, MD
New York, NY
James D. Marks, MD, PhD
San Francisco, CA
Trastuzumab Emtansine (T-DM1): Hitching a

Ride on a Therapeutic Antibody
By Howard A. Burris III, MD
Overview:
The treatment of cancers with chemotherapy is
frequently limited by side effects. The effectiveness may be
improved by the use of monoclonal antibodies to deliver
cytotoxic agents to cancer cells while limiting exposure to
normal tissues. The use of antibody-drug conjugates (ADCs) is
one such strategy: a drug connected by a linker to an antibody
specific for a tumor antigen is the basic makeup of an ADC.
Overexpression and amplification of HER2 is associated with
clinically aggressive breast cancers, and the use of trastuzumab to target HER2 has been highly effective. That said,
most patients with HER2-positive metastatic breast cancer
will eventually have disease progression during targeted therapy. Trastuzumab emtansine (TDM1) is a novel ADC that
combines the humanized antibody trastuzumab and the potent
antimicrotubule agent T-DM1 (derivative of maytansine) using
a unique and highly stable linker. The potential of maytansine
was found in the 1970s with clinical responses noted against
breast cancer; however, substantial toxicity prohibited further
development. DM1 possesses in vitro cytotoxicity 10 to 200
times greater than that of taxanes and vinca alkaloids. A phase
I trial of T-DM1 for patients with heavily pretreated HER2positive breast cancer determined a recommended dose of 3.6
HE TREATMENT of many cancers is often based on the

use of standard chemotherapy agents and regimens.
The effectiveness of this approach is frequently limited by
substantial systemic toxicities associated with these agents.
The therapeutic index can be markedly improved with the
use of monoclonal antibodies to deliver potent cytotoxic
agents to cancer cells, thus minimizing the exposure of the
agents to normal tissues. The use of antibody-drug conjugates (ADCs) is one such strategy; a cytotoxic drug connected by a chemical linker to a monoclonal antibody specific
for a tumor antigen is the basic makeup of an ADC. More
stable linkers and different cytotoxic agents have enabled a
new generation of ADCs to enter the clinic. First and
foremost, the design of an ADC centers on the selection of an
antigen that is tumor-specific and accessible to antibody
binding at the tumor cell.
Overexpression and amplification of HER2 is associated
with clinically aggressive breast cancers that have historically had an overall poor prognosis and therapeutic resistance to traditional drugs. The use of trastuzumab to target
the extracellular domain of HER2 has been highly effective
in the treatment of this type of breast cancer. Multiple
mechanisms for the efficacy of trastuzumab have been
proposed, including inhibition of the PI3K signal transduction pathway, antibody-dependent cell-mediated cytotoxicity (ADCC), and induction of apoptosis. When combined with
chemotherapy, trastuzumab improves the time to disease
progression and overall survival for patients with HER2positive metastatic breast cancer. Furthermore, mature
data from four large phase III trials in which trastuzumab
was evaluated in the adjuvant setting, have demonstrated
marked improvements in both disease-free and overall survival. That said, most patients with HER2-positive metastatic breast cancer will eventually have disease progression
during targeted therapy while the cancer continues to both
mg per kilogram delivered every 3 weeks. Responses were

seen in multiple patients. T-DM1 was then studied in phase II
trials of patients with HER2-positive metastatic breast cancer.
In a studies of 112 and 110 patients in whom disease had
progressed during HER2-directed therapy, T-DM1 was associated with objective response rates of 26% and 34%, respectively. The agent was well tolerated in both trials, with most
toxicities being grade 1 and 2, and no bleeding episodes or
cardiac events occurring. Additional phase II and III trials are
now evaluating T-DM1 in the first-line setting. In one such
trial, T-DM1 was compared with standard dosing of trastuzumab every 3 weeks plus docetaxel every 3 weeks. Objective
response rates were comparable and grade 3 or4 adverse
events were substantially reduced in the T-DM1 arm. The
anticipated selective activity and reduction in side effects
were thus noted. Randomized multicenter phase III trials are
ongoing, including the EMILIA trial, an open-label trial of
T-DM1 compared with the U.S. Food & Drug Administrationapproved regimen of capecitabine plus lapatinib. The results
of studies completed to date suggest T-DM1 is active in
patients who have cancer resistant to trastuzumab-based
combinations.
express HER2 and demonstrate sensitivity to antimicrotubule agents.

T-DM1 is a novel ADC that combines the humanized
antibody trastuzumab and the potent antimicrotubule agent
DM1 (derivative of maytansine) using a unique and highly
stable linker.1 T-DM1, with its ability to bind HER2 with the
same affinity as trastuzumab, maintains the activity of
trastuzumab in addition to providing intracellular delivery
of the antimicrotubule agent DM1. It is hypothesized that
when T-DM1 binds to HER2 receptors, a portion of them
undergo receptor internalization, followed by lysosomal degradation. Activated DM1 is then released from lysosome into
the cellular cytoplasm after antibody degradation, inhibiting
microtubule assembly and causing cell death. Potent cytotoxic agents are needed to maximize the role of drug conjugates. In addition, the drug must be inactive and nontoxic in
the conjugated form to avoid systemic toxicities. Few agents
are able to fulfill these characteristics, including the inhibitors of tubulin polymerization (the maytansinoids and the
auristatins).
The potential of maytansine as an anticancer agent was
originally discovered in the 1970s with clinical responses
noted against breast cancer. However, substantial and random toxicities of neuropathy and myelosuppression were
prohibitive of further clinical development. Recently, an
attempt at improving the therapeutic index through conjugation with trastuzumab was undertaken, leading to the
From the Sarah Cannon Research Institute, Nashville, TN.

Address reprint requests to Howard A. Burris, MD, Tennessee Oncology/Sarah Cannon
Research Institute, 250 25th Avenue N., Suite 110, Nashville, TN 37203; email:
Howard.burris@scresearch.net
1092-9118/10/1-10
159
HOWARD A. BURRIS III

Table 1. Summary of Clinical Efficacy Data from Trastuzumab Emtasine (T-DM1) Clinical Trials
Trial and Reference
TDM3569g2
TDM4258g3
TDM4374g4
TDM4373g5
TDM4450g6
Number of
Patients
Treatment Regimens
T-DM1 0.3 tp 4.8 mg/kg q3w for previously treated HER2

MBC after previous chemotherapy and disease progression
on trastuzumab
T-DM1 3.6 mg/kg q3w for HER2-positive MBC after previous
chemotherapy and disease progression on HER2-targeted
therapy
T-DM1 3.6 mg/kg q3w for HER2-positive MBC after previous
exposure to an anthracycline, a taxane, capecitabine and 2
HER2-directed therapies in the metastatic setting
T-DM1 3.6 mg/kg q3w pertuzumab 840 mg loading dose
then 420 mg q3w, for HER2 MBC first-line treatment or
after previous chemotherapy and HER2-directed therapy
T-DM1 3.6 mg/kg q3w
Or
Trastuzumab 8 mg/kg loading dose then 6 mg/kg q3w
docetaxel 75 mg/m2 or 100 mg/m2 q3w
ORR, Number of
Patients (%)
24
6/24 (25.0)
112
29 (26)
110
38 (34.5)
67
67
70
First line: 9/22 (41),

Relapsed: 19/45 (42)
32 (47.8)
29 (41.4)
Abbreviations: MBC, metastatic breast cancer; MTD, maximum tolerated dose; NR, not reported; ORR, objective response rate; q3w, every 3 weeks.
development of DM1 (derivative of maytansine 1). DM1

possesses in vitro cytotoxicity that is 10 to 200 times greater
than that of other tubulin inhibitors, such as taxanes and
vinca alkaloids. A suitable linker is critical to this process,
and it must have a higher degree of stability in the circulation and allow efficient release of the potent cytotoxic agent
once inside the tumor cell. The cytotoxic drug DM1 is
conjugated to lysine residues of trastuzumab using a unique
hetero-bifunctional
reagent,
N-succinimidyl
4-(Nmaleimidomethyl) cyclohexane-1-carboxylate (SMCC), in a
two-step process. Trastuzumab is initially reacted with
SMCC to form trastuzuamb-MCC. Next, T-MCC is then
conjugated to DM1 to make T-DM1. The thioether linker
was developed to provide a bond between trastuzumab and
KEY POINTS
160
The therapeutic index for treating patients can be

improved by the use of monoclonal antibodies to
deliver potent cytotoxic agents to cancer cells while
minimizing exposure of the agents to normal tissues.
The use of antibody-drug conjugates (ADCs) is one
such strategy; a cytotoxic drug connected by a chemical linker to a monoclonal antibody specific for a
tumor antigen is the basic makeup of an ADC.
Trastuzumab emtansine (TDM1) is a novel ADC
that combines the humanized antibody trastuzumab
and the potent antimicrotubule agent T-DM1 (derivative of maytansine) with use of a unique and highly
stable linker.
In phase II studies of patients in whom metastatic
breast cancer previously progressed during multiple
HER2-directed therapies, T-DM1 was associated
with response rates of 26% to 34% and was reasonably well tolerated.
The results of phase II studies suggest T-DM1 can
improve outcomes for patients with cancers that are
resistant to trastuzumab-based combinations, and
phase III trials are underway.
DM1 that is more stable than hydrazone or disulfide linkers.

The therapeutic index of DM1 is thus enhanced by minimizing systemic exposure to free DM1 and improving exposure
to T-DM1. Before development in the clinic, the conjugate
was extensively studied in preclinical models. The effectiveness of T-DM1 was established in three murine models of
HER2-expressing human breast carcinoma. In contrast,
little activity was seen in the normal human cells or breast
cancer cells not overexpressing HER2, demonstrating the
specificity of the ADC.1
T-DM1 was the first HER2-targeted ADC with this unique
SMCC linker to be studied in patients. A phase I trial in
patients with HER2-positive breast cancer that had progressed during prior trastuzumab-based therapy determined a maximum tolerated dose of 3.6 mg per kilogram,
delivered every 3 weeks.2 The dose-limiting toxicity was
grade 4 thrombocytopenia that was rapidly reversible and
not associated with clinically meaningful bleeding events.
No cardiac events or left ventricular ejection fraction declines were noted. In addition, no alopecia greater than
grade 1 was noted, further evidence for the lack of systemic
toxicity.
Six of the 24 patients had an objective partial response.
All patients had previously been treated with trastuzumab,
with a median exposure of approximately 2 years, as well as
microtubulin inhibiting agents. Of the six responses, four
occurred in the nine patients treated at the maximum
tolerated dose. The trial pharmacokinetics demonstrated
peak free (unconjugated) DM1 plasma concentrations immediately after dosing which were low on all time points,
suggesting that any systemic toxicity was unrelated to
circulating unconjugated DM1. Weekly dosing was also
explored and was both active and well tolerated, but it
showed no particular advantage from either a dose intensity
or density standpoint.
After these results, T-DM1 was studied in phase II trials
of patients with HER2-positive metastatic breast cancer at
the recommended dose of 3.6 mg per kilogram every 3 weeks.
In a study of 112 patients who had disease progression
during HER2-directed therapy, T-DM1 was associated with
an objective response rate of 26% based on independent
review, and progression-free survival of 4.6 months.3 The
T-DM1 AND THERAPEUTIC ANTIBODIES
agent was well tolerated, with most toxicities being grade 1

and 2, and no bleeding episodes or cardiac events.
A second trial was conducted in 110 patients with HER2positive metastatic breast cancer who had received prior
treatment with an anthracycline, a taxane, capecitabine,
trastuzumab, and lapatinib (and had had disease progression during treatment with the most recent regimen).4
T-DM1 demonstrated an objective response rate of 34.5%
and a median progression-free survival of 6.9 months,
based on independent reviews. The agent was again well
tolerated in this heavily pretreated population, and no
cardiac toxicity signals were noted. During these phase II
trials, central review for HER2-positivity was required.
Response rates were higher amongpatients with centrally
verified HER2-positive tumors, whereas few responses were
noted among patients with tumors that tested negatively on
central review, confirming the relationship with drug activity.
Additional trials are now being done to evaluat T-DM1
asfirst-line treatment. One such trial involved 137 patients
who received either T-DM1 or standard dosing of trastuzumab every 3 weeks plus docetaxel 75 or 100 mg/m2 every
3 weeks.5 Objective response rates were comparable, at 48%
and 41%, respectively. Of note, grade 3 or 4 adverse events
were substantially reduced in the T-DM1 arm (37% compared with 75%). The selective activity and proposed reduction in side effects were demonstrated in this randomized
phase II trial.
Two randomized multicenter phase III trials are ongoing
to evaluate the role of T-DM1 in earlier lines of therapy.
MARIANNE is designed to compare the efficacy and safety
of single-agent T-DM1, alone or in combination with pertu-
zumab, with the standard trastuzumab plus a taxane (paclitaxel or docetaxel). The exposure to pertuzumab is blinded,
and the standard arm is open-label.6
Another study, EMILIA, is an open-label trial in which
T-DM1 is being compared with the U.S. Food and Drug
Administration (FDA)-approved regimen of capecitabine
plus lapatinib in patients previously treated with both a
taxane and trastuzumab. Interestingly, in another trial,
T-DM1 is being compared with physicians choice of treatment in patients who had disease progression after multiple
prior regimens. Trastuzumab alone or with chemotherapy is
allowed in the physicians-choice arm.
Conclusion
The discovery of HER2 gene amplification and the subsequent development of trastuzumab has markedly improved
the prognosis for patients with HER2-positive breast cancer.
Unfortunately, not all patients have a response to trastuzumab, and disease will progress in most patients with
metastatic HER2-positive disease. T-DM1 meets the criteria
for a successful ADC by combining the targeted effect of
trastuzumab with the cytotoxic potency of DM1 using a
stable linker and minimizing systemic toxicity. In addition,
other tumor histologies, such as gastrointestinal cancers
that are HER2-positive may be sensitive to this agent. The
results of phase II studies suggest T-DM1 can improve
outcomes for patients with cancers that are resistant to
trastuzumab-based combinations. An aggressive portfolio of
phase II and III clinical trials will help determine the role
of T-DM1 in earlier lines of therapy or with combinations of
other targeted agents.
Author
Howard A. Burris III*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Burris HA. Trastuzumab emtansine: A novel antibody-drug conjugate
for HER2-positive breast cancer. Expert Opin Bio Ther. 2011;11:807-819.
2. Krop IE, Beeram M, Modi S, et al. Phase I study of trastuzumabDM1, a HER2 antibody-drug conjugate, given every 3 weeks to patients
with HER2-positive metastatic breast cancer. J Clin Oncol. 2010;28:26982704.
3. Burris HA 3rd, Rugo HS, Vukelja SJ, et al. Phase II study of the antibody
drug conjugate trastuzumab-DM1 for the treatment of human epidermal
growth factor receptor 2 (HER2)-positive breast cancer after prior HER2directed therapy. J Clin Oncol. 2011;29:398-405.
4. Krop I, Lorusso P, Miller KD, et al. A phase II study of trastuzumabDM1 (T-DM1), a novel HER2 antibody-drug conjugate, in patients with
HER2-positive metastatic breast cancer who were previously treated with an
anthracycline, a taxane, capecitabine, lapatinib, and trastuzumab. Paper

presented at: the European Society for Medical Oncology. October 10, 2010,
Milan, Italy, abstract 277O.
5. Perez EA, Dirix L, Kocsis J, et al. Efficacy and safety of trastuzumabDM1 versus trastuzumab plus docetaxel in HER2-positive metastatic breast
cancer patients with no prior chemotherapy for metastatic disease: Preliminary results of a randomized, multicenter open-label phase 2 (TDM4450g).
Ann Oncol. 2010; 21 Suppl 8:LBA3.
6. A study of trastuzumab-MCC-DM1 (T-DM1) in combination with pertuzumab administered to patients with HER2-positive locally advanced or
metastatic breast cancer who have previously received trastuzumab. http://
www.clinicaltrials.gov/ct2/show/NCT00875979?termt-dm14373&rank1.
Accessed August 3, 2010.
161
Targeting CD30 in Hodgkin Lymphoma:

Antibody-Drug Conjugates Make a Difference
By Catherine S. M. Diefenbach, MD, and John P. Leonard, MD
Overview: CD30 expression is characteristic of the malignant

Reed-Sternberg cell in Hodgkin lymphoma (HL) and several
other lymphoid malignancies, such as anaplastic large-cell
lymphoma (ALCL). Although unconjugated anti-CD30 antibodies have had minimal therapeutic activity in patients with HL
as single agents, the CD30-directed antibody-drug conjugate
(ADC) brentuximab vedotin has demonstrated activity that has
resulted in its recent regulatory approval for the treatment of
patients with relapsed HL and ALCL. Approximately 75% of
patients with recurrent HL achieve objective responses, with
ODGKIN LYMPHOMA (HL) is an uncommon B-cell

lymphoid neoplasm, which accounts for approximately
10% of all lymphomas and 0.6% of all cancers diagnosed in
the developed world each year. In 2009, approximately 8,500
cases of HL were diagnosed in the United States.1 The
median age at diagnosis is 38 years, and at least 40% of
patients are younger than age 35 at the time of diagnosis.2
Over the past 30 years, advances in the management of
HL have led to generally successful clinical outcomes, with
roughly 80% of patients cured with chemotherapy or
combined-modality therapy. Despite this high cure rate,
approximately 5% to 10% of patients have primary refractory disease, and 20% to 30% of patients will experience
relapse after attaining initial complete remission. Radiation
therapy has an important role in treatment, although primarily in limited-stage and bulky disease. Second-line chemotherapy and autologous stem cell transplantation (ASCT)
approaches are curative for only approximately 50% of
patients with relapsed or refractory disease. Patients whose
disease recurs after second-line therapy or who have primary refractory disease have generally poor outcomes with
conventional approaches. For patients whose disease recurs
after ASCT, the median time-to-progression with subsequent therapy is reported to be 3.8 months, and median
survival is 26 months.3,4 Allogeneic stem cell transplantation (alloSCT) can induce durable remissions in some patients with relapsed and primary refractory HL; however,
the use of this modality is limited in part by the challenges
of achieving adequate disease control in this patient population before transplantation. Despite the successes associated with the modern management of HL, an estimated
1,300 deaths from HL still occur each year in the United
States,5 many of which are young adults. Novel therapies for
these patients are clearly needed.
Regarding disease pathology, HL is a lymphoid neoplasm,
likely derived from preapoptotic germinal-center B cells,
characterized by the presence of large binucleated or multinucleated cells with prominent nucleoli termed Hodgkin and
Reed-Sternberg (HRS) cells. Immunostains are characteristically positive for CD15 and CD30, and this pattern confirms diagnosis. The malignant HRS cells, typically
comprise a small fraction of the total cellular population
(0.1% to 10%), and reside in an environment of reactive
inflammatory cells that produce soluble and membranebound factors that promote tumor cell growth, evasion of
self-immunity, and survival,6,7 and typically comprise a
162
the principal toxicity being peripheral neuropathy. Ongoing

studies are evaluating treatment with this agent as part of
first-line therapy, for patients with relapsed disease, and for
patients with resistant disease and limited other options.
Brentuximab vedotin demonstrates the therapeutic value of
antibody-drug conjugation and serves as a model of how a
novel, targeted approach can be employed to potentially
further improve outcomes in settings where curative chemotherapeutic regimens are already available.
small fraction of the total cellular population (0.1% to 10%).

HRS cells may constitutively express nuclear factor kappa B
(NF-B) and other antiapoptotic proteins that inhibit both
the intrinsic and extrinsic pathways of apoptosis, such as
c-FLICE and X-linked inhibitor of apoptosis protein.8 A
therapy targeted to the malignant cells in HL must therefore effectively disrupt a small but key component of the
actual tumor mass, the malignant HRS cells.
CD30: A Novel Therapeutic Target
CD30 is a 120-KDa type I transmembrane glycoprotein

belonging to the tumor necrosis factor (TNF) superfamily.
The cytoplasmic tail of CD30 contains TNF receptorassociated (i.e., TRAF) binding sequences that can mediate the
activation of NF-B. Overexpression of CD30 may cause
NF-B activation, independent of its ligand CD30L. CD30
may also have a role in maintaining CD4 memory, but the
precise function of CD30 signaling in healthy individuals
has not been well clarified. In normal lymphocytes, CD30 is
expressed by only a small percentage of activated B cells and
on nave T and B cells where CD30 expression is enhanced
by the presence of T helper 2 cytokines, costimulatory
signals from interleukin-4, and CD28 during T-cell activation. CD30 signaling is associated with pleotropic downstream effects including cellular survival, differentiation,
and lymphocyte activation. Although CD30 signaling may
induce apoptosis in some cell types, HL cells are protected
from CD30-mediated apopotosis by concomitant constitutive
activation of NF-B.
A preferred antigen target for tumor-directed antibody
therapy should both be differentially expressed and have
strong immunogenicity (if immune mechanisms are the
dominant mechanism of action). The high expression of
CD30 on the malignant HRS cells combined with its limited
expression on normal activated T and B cells makes CD30
an appealing target for directed therapy. In addition, CD30
From the New York University Cancer Institute, NYU Langone Medical Center, New
York, NY; Weill Cornell Cancer Center, Weill Cornell Medical College and New York
Presbyterian Hospital, New York, NY.
Address reprint requests to John P. Leonard MD, Center for Lymphoma and Myeloma,
Division of Hematology and Medical Oncology, Weill Cornell Cancer Center, Weill Cornell
Medical College and New York Presbyterian Hospital, 1305 York Avenue, New York, NY;
email: jpleonar@med.cornell.edu.
1092-9118/10/1-10
TARGETING CD30 IN HODGKIN LYMPHOMA

Table 1. Summary of Clinical Response Data to CD30-Targeted Antibody Therapy
Drug Investigated
(reference)
SGN-30 (13)
MDX-060 (14)
Brentuximabvedotin
(19)
Brentuximabvedotin
(20)
Brentuximabvedotin
(21,22)
Brentuximab vedotin
(23)
ORR (%)
HL
ALCL
Total
Treatment Dosage
(mg/kg body weight)
HL
ALCL
Phase II
Phase I/II
Initial phase I
38
63
42
41
7
2
79
72*
45
612 weekly
0.115 weekly
0.13.6 every 3 wk
0
6 (CR 3, PR 3)
36 (CR 21)
17 (CR 5, PR 12)
29 (CR 29)
100 (CR 100)
Phase I
44
38
0.41.4 weekly
102
102
1.8 every 3 wk
58
58
1.8 every 3 wk
Clinical trial
Pivotal phase II, in

patients with HL
Phase II, in patients
with sALCL
75 (CR 34)
Overall
9
8*
38 (CR 24; ORR 50 for
patients treated at the
MTD)
59 (CR 34)
86 (CR 53)
Abbreviations: ALCL: anaplastic large-cell lymphoma; CR: complete remission; HL: Hodgkin lymphoma; MTD: maximum-tolerated dose; ORR: overall response rate;
PR: partial remission.
* Including two patients with CD30 T cell lymphoma. Including one patient with CD30 angioimmunoblastic T cell lymphoma. Including one patient with peripheral
T cell lymphoma.
protein may be shed into a soluble form in the serum

(sCD30). Elevated levels of sCD30 have been detected in the
serum of patients with CD30-positive malignancies, and
there is some suggestion that this might correlate with
tumor burden, clinical outcome, or both.9 CD30 expression is
present in many conditions of dysregulated immunity, including cutaneous T-cell lymphoma, diffuse large B-cell
lymphomas, follicular lymphoma, post-transplantation lymphoproliferative disorder, and virally infected T cells (e.g.,
HIV).10-12
Unconjugated CD30-Directed Monoclonal Antibodies
Several groups have evaluated unconjugated monoclonal

antibodies against CD30 as therapeutic agent. Recent advances in antibody engineering have optimized two unconjugated antibodies: SGN-30 (cAC10), a chimerized
immunoglobulin G1 (IgG1) monoclonal antibody, and MDX060, a fully human monoclonal antibody that has shown
enhanced antibody-dependent cell-mediated cytotoxicity
(ADCC) compared with SGN-30. SGN-30 was evaluated in a
phase II trial of patients with HL and systemic CD30positive anaplastic large-cell lymphoma (sALCL) (Table
1).13 The treatment was well tolerated, however objective
responses were not observed in patients with HL. Yet, seven
of the 41 patients with sALCL (17%) attained therapeutic
KEY POINTS
CD30 expression is characteristic of Reed-Sternberg

cells in Hodgkin lymphoma.
Efforts to target CD30 with unconjugated or naked
monoclonal antibodies have had limited therapeutic
success.
Brentuximab vedotin is a CD30-directed antibody
conjugated to a synthetic taxane (MMAE) with substantial therapeutic activity in recurrent Hodgkin
lymphoma.
The toxicity profile of brentuximab vedotin suggests
that it may be combined successfully with several
standard chemotherapeutic agents.
Ongoing efforts are directed toward integrating brentuximab vedotin into Hodgkin lymphoma therapeutic
regimens to maximally improve outcomes.
responses: two (5%) attained complete response (CR) and

five (12%) attained partial remission (PR), which provides
some proof of concept. A phase I/II trial of MDX-060 was
performed in 72 patients with relapsed or refractory CD30positive lymphomas.14 The overall response rate (ORR) was
6% in the patients with HL and 29% in the patients with
ALCL. With preclinical data suggesting synergy between
cytotoxic chemotherapy and SGN-30, this agent was combined with GVD (gemcitabine, vinorelbine, and liposomal
doxorubicin) chemotherapy by the Cancer and Leukemia
Group B (CALGB) in a phase II trial for patients with
relapsed HL.15 This trial was prematurely closed because of
grades 3 to 5 pneumonitis that occurred in five of the 30
patients enrolled. Interestingly, all five patients with pneumonitis were noted to have a valine/phenylalanine (V/F)
polymorphism in the FcgammaRIIIa gene, suggesting a
possible immune component. These pulmonary issues, including the potential for worsened risk of lung toxicity in
combination with bleomycin (which can cause lung injury
itself) are of importance as the development of CD30directed agents moves forward. Given the relatively low
response rate seen with both SGN-30 and MDX-060 (i.e.,
unconjugated anti-CD30 antibodies), their continued development has been limited, and the assessment of novel,
conjugated agents has taken higher priority.
Brentuximab-Vedotin: A Novel CD30-Targeted
Antibody-Drug Conjugate
Antibody-drug conjugates (ADC) can provide specific,

preferential targeting of potent chemotherapeutic agents or
toxins to differentially expressed antibody targets. Regarding binding to cell-surface targets, ADCs are generally
internalized primarily via clathrin-mediated endocytosis
and subsequent lysosomal proteolysis. This feature allows
the therapeutic agent to be delivered into the target cell,
ideally without affecting bystander and normal cells because
they lack the antigen providing specificity. By allowing more
of the drug or toxin to reach the target cell, therapeutic
benefit should result, and because normal cells would be
spared, toxicity should be minimized. In addition, if the
antibody itself and the drug conjugated to it both have
antitumor activity, multiple mechanisms of therapeutic action may be in effect.
Brentuximab vedotin is a CD30-targeted ADC composed
of the chimeric antibody cAC10 (SGN-30) modified by the
addition of a valine-citrulline peptide linker to monomethyl
163
DIEFENBACH AND LEONARD
auristatin E (MMAE), a synthetic analog of the naturally

occurring antimitotic agent dolastatin 10. Brentuximab vedotin appears to have a multistep mechanism of action that
is initiated by binding to CD30 on the cell surface, followed
by clathrin-mediated endocytosis. MMAE is released in the
lysosome from its conjugate through proteolytic degradation
of the peptide linker.16 The binding of released MMAE to
tubulin disrupts the microtubule network, leading to G2/M
phase cell-cycle arrest and apoptosis.17 Overall efficacy of
the ADC may be enhanced by the fact that a small fraction
of drug diffuses out of the targeted tumor cell and exerts
cytotoxic effects on the surrounding tumor microenvironment.
Preclinical in vitro and in vivo studies demonstrated that
brentuximab vedotin was highly selective for its target
CD30, was highly potent, and had low toxicity.18 Brentuximab vedotin displayed a half maximal inhibitory concentration (IC50) of less than 10 ng/mL against CD30-positive
tumor cell lines, but was 300-fold less active against CD30negative cells. In severe combined immunodeficiency (SCID)
mouse xenograft models of HL, brentuximab vedotin was
efficacious at doses as low as 1 mg/kg. SCID mice treated
with doses as high as 30 mg/kg showed no evidence of
toxicity. Enhanced in vitro antitumor activity against HL
was reported for the combination of brentuximab vedotin
with cytotoxic chemotherapy agents such as bleomycin,
dacarbazine, vinblastine, doxorubicin, or gemcitabine. Collectively, these experiences suggested that brentuximab
vedotin has the potential to be a highly effective agent
against CD30-expressing malignancies.
Clinical Data
In the initial phase I trial of brentuximab vedotin in

patients with relapsed or refractory CD30-positive malignancies, 45 patients were enrolled with histologies including
HL (42 patients), sALCL (two patients), and CD30-positive
angioimmunoblastic T-cell lymphoma (one patient).19 The
primary objective of the study was to determine the maximum tolerated dose (MTD), and the secondary objectives
were to evaluate overall response rate (ORR) and pharmacokinetics. Patients had a median age of 36 (range, 20 to 87),
had Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1 (93%), and were heavily pretreated,
with a median of three prior chemotherapy regimens (range,
one to seven). Seventy-three percent had recurrent disease
after prior ASCT. Brentuximab vedotin was administered as
outpatient intravenous infusions at doses ranging from 0.1
mg/kg to 3.6 mg/kg, every 21 days.
Brentuximab vedotin was well tolerated at doses less than
1.8 mg/kg, with the most common adverse events being
fatigue (16 patients [36%]), pyrexia (15 patients [33%]), and
diarrhea, nausea, neutropenia, or peripheral neuropathy (10
patients [l22%] for each). No grade 3 or 4 adverse events
occurred at the 1.2-mg/kg dose level (one level below the
MTD). Grade 3 neutropenia, limb pain, and back pain each
occurred in one patient of the 12 receiving the 1.8-mg/kg
dose. At doses higher than 1.8 mg/kg, grade 3 neutropenia
and pyrexia were seen in two of the 12 patients (17%)
receiving the 2.7-mg/kg dose. One patient receiving the
3.6-mg/kg dose developed febrile neutropenia and died of
infectious complications. Besides the low level of hematologic toxicity, the most important side effect was peripheral
neuropathy, which occurred in 16 patients (36%), 13 of
164
whom were receiving the 1.8-mg/kg or 2.7-mg/kg doses. This

observation was typically characterized by grade 1 or 2
parasthesias in the hands or feet and prompted the discontinuation of treatment in three patients. Importantly, subsequent resolution of peripheral neuropathy was noted in 10
of the 16 patients (63%) at the last safety assessment for the
study.
Of greatest note in this initial trial was that 17 patients
demonstrated objective responses, including 11 CRs. Six of
the 12 patients (50%) treated at the MTD of 1.8 mg/kg had
an objective response with a median response duration of 9.7
months (ranging from 0.6 month to more than 19.5 months).
Some reduction of tumor size was noted in 36 of the 42
evaluable patients (86%), and 13 of the 16 patients (81%)
with disease-related symptoms at baseline (e.g., pruritis and
night sweats) had their symptoms resolve upon receiving
therapy, irrespective of their antitumor response status.19
To explore a weekly dosing schedule (compared with the
3-week schedule in the prior study), a second phase I trial
was conducted in a similar patient population.20 A total of 44
patients38 with HL, five with sALCL, and one with
peripheral T-cell lymphomawere enrolled. Patients had a
median age of 33 (range, 12 to 82) with a median of three
prior therapies (range, one to eight). Sixty-eight percent of
patients had received prior ASCT. Patients received weekly
brentuximab vedotin for 3 of 4 weeks, with doses ranging 0.4
mg/kg to 1.4 mg/kg. The most common treatment-related
adverse events again included peripheral neuropathy, nausea, fatigue, neutropenia, diarrhea, and dizziness. The MTD
for weekly dosing was 1.2 mg/kg. The ORR evaluated across
all dose levels was 59%, with 34% achieving CR. The median
duration of response had not been reached at the follow-up
at 45 weeks. Data from the phase 1 trials demonstrated
that across the dosing range of 1.2 to 2.7mg/kg ADC exposures were dose proportional, and steady state levels were
achieved within 21 days on an every 3 week dosing schedule.
On this schedule there was minimal to no accumulation of
ADC observed with multiple treatments.
Given the high response rates in both of these phase I
trials, a pivotal phase II trial was conducted in patients with
relapsed or refractory HL, and outcomes were reported at
the 2010 American Society of Hematology (ASH) annual
meeting21 with updates presented at the 2011 American
Society of Clinical Oncology (ASCO) annual meeting.22 In
the study, 102 patients with relapsed or refractory HL were
treated every 3 weeks with brentuximab vedotin at a dose of
1.8 mg/kg. Demographics of the study population included a
median age of 31 and a median of four prior therapies. All
participants had undergone ASCT. The ORR was 75% with
a CR rate of 34%, and reductions in tumor volume were
reported in up to 95% of patients. The median duration of
response for patients achieving CR had not been reached at
the time of the presentation.
The data from a second phase II clinical trial of patients
with sALCL were also reported at ASH 2011.24 A total of 58
patients with sALCL were treated with the same dose of
brentuximab vedotin (1.8 mg/kg) on the 3-week schedule.
The spectrum of toxicity was similar to that observed in
patients with HL. The ORR was 86% with a CR rate of 53%,
and response duration also had not been reached at the time
of the presentation. The study was updated at ASH 2011.24
At the time of the updated analysis, all but two patients had
discontinued therapy, and the median number of treatment
TARGETING CD30 IN HODGKIN LYMPHOMA
cycles with brentuximab vedotin was seven (range, one to

16). The median duration of objective response was 13.0
months (range 0.1 month to more than 19.1 months), the
median PFS was 14.6 months, and the median OS had not
yet been reached. Peripheral sensory neuropathy was observed in 41% of patients, which in 10 patients (17%) was
grade 3. Among the patients with neuropathy of any grade,
26 (79%) had achieved either resolution or improvement in
symptoms with a median time-to-improvement of 13.3
weeks (range 0.3 week to 48.7 weeks).
These results led to the unanimous recommendation of
the Oncology Drug Advisory Committee of the United States
Food and Drug Administration for accelerated approval of
brentuximab vedotin (Adcetris; Seattle Genetics, Inc, Bothell, WA) for the treatment of relapsed or refractory HL in
patients with progressive disease after either ASCT or two
chemotherapy regimens in patients ineligible for transplantation. Brentuximab vedotin was also approved for patients
with sALCL who had at least one prior treatment fail.
Brentuximab vedotin is the first new therapy approved for
the treatment of HL in more than 30 years.
Ongoing Trials
Although brentuximab vedotin therapy demonstrated

substantial single-agent activity in patients with relapsed or
refractory HL, one would anticipate that even greater benefits could be achieved if it is integrated with or substituted
for components of standard treatment. In addition, aspects
of the optimal use of single-agent therapy need clarification.
A variety of planned and ongoing studies are attempting to
answer several key questions:
Can patients who achieved CR with brentuximab vedotin
who then stopped therapy in remission benefit from retreatment at a time of subsequent relapse? Small case series
suggest that indeed patients may display sensitivity to
retreatment with brentuximab vedotin at the time of relapse.25 Bartlett and colleagues reported a study of seven
patients retreated after relapse (one patient twice), for a
total of eight retreatment experiences. Responses included
two patients with CR, four with PR, and two with stable
disease.26
Can brentuximab vedotin be safely combined with and
improve the results of first-line treatment regimens? An
ongoing phase I trial of brentuximab vedotin in combination with either ABVD (doxorubicin, bleomycin, vinblastine,
and dacarbazine) or AVD (doxorubicin, vinblastine, and
dacarbazine) chemotherapy in untreated patients with
HL is evaluating this issue (ClinicalTrials.gov identifier:
NCT01060904). Interim data for the first 31 patients treated
were presented at ASH 2011.27 Patient demographics included an international prognostic index greater than 4 in
29% of patients, stage IV disease in 55%, and a median age
of 35 (range, 19 to 59). Brentuximab vedotin at either 0.6
mg/kg, 0.9 mg/kg, or 1.2 mg/kg was administered with
standard doses of ABVD or at 1.2 mg/kg with AVD, on days
1 and 15 of each 28-day cycle for six cycles. Overall, adverse
events included peripheral neuropathy (48%), fatigue (45%),
and neutropenia (77%). Grades 3 or 4 adverse events included neutropenia (74%), febrile neutropenia (16%), and
anemia (13%). Importantly, in the ABVD cohort of 25
patients, seven patients (28%) developed pulmonary toxicity, dyspnea, and interstitial lung disease, which led to the
discontinuation of bleomycin. Of the 10 patients who were
available for response assessment, all achieved CR. Based

on these data, the ABVD/brentuximab vedotin arm has been
closed to further enrollment and AVD/brentuximab vedotin
is undergoing further evaluation.
Does brentuximab vedotin have a role in targeting minimal residual disease as a maintenance strategy to prevent
relapse in high-risk patients? This question is currently
being investigated in patients who are at high risk after
ASCT in a placebo-controlled, randomized trial of brentuximab vedotin maintenance, referred to as AETHERA (ADC
Empowered Trial for Hodgkin to Evaluate Progression after
ASCT; ClinicalTrials.gov identifier: NCT01100502).
Future Directions
Determining the optimal timing for brentuximab vedotin

therapy and whether brentuximab vedotin will ultimately
be best used alone as a single agent or as a component of a
novel treatment platform remain ongoing issues. Many of
the patients treated in both of the original phase I studies
and the pivotal phase II study were heavily pretreated, often
with relapse after autologous or allogeneic bone marrow
transplantation. The response rate to brentuximab vedotin
in treatment-nave patients is not known, but because this is
a patient population with chemotherapy-sensitive disease,
one might expect that the activity of the drug would be
greater as part of initial or second-line treatment. First-line
ABVD continues to be generally well tolerated, based on 30
years of safety data and an extremely high response rate.
Thus an improvement in outcome for low-risk patients
might not be easily discernible without very large studies.
However, a meaningful relapse rate occurs, and morbidity
and (albeit rare) mortality with this regimen is measurable.
Improvements in both efficacy and tolerability may be
achievable. For high-risk patients, the issue of whether
including brentuximab vedotin would obviate the need for
more intensive chemotherapy, such as augmented BEACOPP, remains quite relevant.
In addition, the role of brentuximab vedotin as part of a
dose-intensive transplantation approach is important. Maximal cytoreduction before ASCT is associated with a more
favorable outcome. The current standard salvage chemotherapy regimens, such as ifosfamide, carboplatin, and etoposide (i.e., ICE), have a low CR rate (26%) despite a high
ORR.28 Improving the CR rate might allow more patients
with relapsed or refractory HL to proceed to and have
successful results from ASCT. A retrospective analysis reported at ASH 2011 suggests that brentuximab vedotin
therapy before reduced-intensity alloSCT for patients with
HL may result in prolonged disease control without an
increase in engraftment delay, post-transplant infectious
complications, acute or chronic graft-versus-host disease, or
nonrelapse-related transplant mortality.29
The retreatment data, albeit on a small number of patients, suggest that some patients with relapsed HL may
maintain sensitivity to brentuximab vedotin. This issue
warrants more detailed characterization in larger numbers
of patients to better identify which patients might be most
likely to benefit from retreatment. A phase II retreatment
trial in patients who have previously achieved objective
responses to brentuximab vedotin is currently ongoing
(ClinicalTrials.gov identifier: NCT00947856).
The challenges for this highly active and novel therapy
165
DIEFENBACH AND LEONARD
also include the assessment of potential long-term toxicities

in patients, which might be more relevant in individuals
receiving brentuximab vedotin as part of curative therapy
at earlier time points. The incidence of bleomycin-induced
pulmonary toxicity in patients receiving ABVD/brentuximab
vedotin (28%) is unusually high for this population. The
standard risk of bleomycin-induced pulmonary toxicity
with six cycles of ABVD is approximately 2% in patients
without other pulmonary risk factors. These data, combined
with the earlier data from Blum and the CALGB indicating
a specific V/F polymorphism in the FcgammaRIIIa gene
underlying susceptibility,15 suggest that more research on
intrinsic risk factors for pulmonary toxicity should be undertaken and that the ideal combination platform for brentuximab vedotin must also avoid concurrent use of other
agents with overlapping toxicities. Given the relatively high

incidence of peripheral sensory neuropathy seen in phase II
trials, attention should also be paid to this side effect during
long-term therapy, although the data thus far suggest that
this symptom improves in most patients with the cessation
of therapy.
As we gain greater scientific understanding of the role of
CD30 signaling, the immunologic microenvironment, and
mechanisms of tumor resistance in CD30-positive lymphomas, one may anticipate that even more rationally designed
treatment regimens will emerge. Despite the relative success achieved for most patients with HL with standard,
largely empirically designed regimens, brentuximab vedotin
use provides an example of the direction of future collaborative and translational efforts.
Author
Catherine S. M.
Diefenbach
John P. Leonard
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Seattle Genetics
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Seattle Genetics
Seattle Genetics
REFERENCES
1. American Cancer Society. Cancer Facts & Figures 2009. Atlanta: American Cancer Society, 2009.
2. SEER Cancer Statistics Review 2001-2008. National Cancer Institute.
www.seer.cancer.gov.
3. Bonfante V, Santoro A, Viviani S, et al. Outcome of patients with
Hodgkins disease failing after primary MOPP-ABVD. J Clin Oncol. 1997;15:
528-534.
4. Longo DL, Duffey PL, Young RC, et al. Conventional-dose salvage
combination chemotherapy in patients relapsing with Hodgkins disease after
combination chemotherapy: the low probability for cure. J Clin Oncol.
1992;10:210-218.
5. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2009. CA Cancer
J Clin. 2009;59:225-249.
6. Aldinucci D, Gloghini A, Pinto A, et al. The classical Hodgkins lymphoma microenvironment and its role in promoting tumour growth and
immune escape. J Pathol. 2010;221:248-263.
7. Steidl C, Connors JM, Gascoyne RD. Molecular pathogenesis of Hodgkins lymphoma: increasing evidence of the importance of the microenvironment. J Clin Oncol. 2011;29:1812-1826.
8. Farrell K, Jarrett RF. The molecular pathogenesis of Hodgkin lymphoma. Histopathology. 2011;58:15-25.
9. Nadali G, Vinante F, Ambrosetti A, et al. Serum levels of soluble CD30
are elevated in the majority of untreated patients with Hodgkins disease and
correlate with clinical features and prognosis. J Clin Oncol. 1994;12:793-797.
10. Younes A, Carbone A. CD30/CD30 ligand and CD40/CD40 ligand in
malignant lymphoid disorders. Int J Biol Markers. 1999;14:135-143.
11. Gardner LJ, Polski JM, Evans HL, et al. CD30 expression in follicular
lymphoma. Arch Pathol Lab Med. 2001;125:1036-1041.
12. Horie R, Watanabe T. CD30: expression and function in health and
disease. Semin Immunol. 1998;10:457-470.
13. Forero-Torres A, Leonard JP, Younes A, et al. A Phase II study of
SGN-30 (anti-CD30 mAb) in Hodgkin lymphoma or systemic anaplastic large
cell lymphoma. Br J Haematol. 2009;146:171-179.
14. Ansell SM, Horwitz SM, Engert A, et al. Phase I/II study of an
anti-CD30 monoclonal antibody (MDX-060) in Hodgkins lymphoma and
anaplastic large-cell lymphoma. J Clin Oncol. 2007;25:2764-2769.
15. Blum KA, Jung SH, Johnson JL, et al. Serious pulmonary toxicity in
patients with Hodgkins lymphoma with SGN-30, gemcitabine, vinorelbine,
and liposomal doxorubicin is associated with an FcRIIIa-158 V/F polymorphism. Ann Oncol. 2010;21:2246-2254.
16. Sutherland MS, Sanderson RJ, Gordon KA, et al. Lysosomal trafficking
and cysteine protease metabolism confer target-specific cytotoxicity by
peptide-linked anti-CD30-auristatin conjugates. J Biol Chem. 2006;281:
10540-10547.
166
17. Francisco JA, Cerveny CG, Meyer DL, et al. cAC10-vcMMAE, an

anti-CD30-monomethyl auristatin E conjugate with potent and selective
antitumor activity. Blood. 2003;102:1458-1465.
18. McEarchern JA, Kennedy D, McCormick R, et al. Activity of SGN-35 in
preclinical models of combination therapy and relapse prevention. Haematol.
2010;95 (suppl; abstr 49)
19. Younes A, Bartlett NL, Leonard JP, et al. Brentuximab vedotin
(SGN-35) for relapsed CD30-positive lymphomas. N Engl J Med. 2010;363:
1812-1821.
20. Fanale MA, Forero-Torres A, Rosenblatt JD, et al. A phase I weekly
dosing study of brentuximab vedotin in patients with relapsed/refractory
CD30-positive hematologic malignancies. Clin Cancer Res. 2012;18:248-255.
21. Chen R, Gopal AK, Smith SE, et al. Results of a pivotal Phase 2 study
of brentuximab vedotin (SGN-35) in patients with relapsed or refractory
Hodgkin lymphoma. Blood . 2010;116:21 (suppl; abstr 283).
22. Chen RW, Gopal AK, Smith SE, et al. Results from a pivotal phase II
study of brentuximab vedotin (SGN-35) in patients with relapsed or refractory Hodgkin lymphoma (HL). J Clin Oncol. 2011;29:15 (suppl; abstr 8031).
23. Pro B, Advani R, Brice P, et al. Durable remissions with brentuximab
vedotin (SGN-35): Updated results of a phase II study in patients with
relapsed or refractory systemic anaplastic large cell lymphoma (sALCL).
J Clin Oncol. 2011;29;15 (suppl; abstr 8032).
24. Advani RH, Shustov AR, Brice P, et al. Brentuximab vedotin (SGN-35)
in patients with relapsed or refractory systemic anaplastic large cell lymphoma: A Phase 2 study update. Blood. 2011;118;21 (suppl; abstr 443).
25. Foyil KV, Kennedy DA, Grove LE, et al. Extended retreatment with
brentuximab vedotin (SGN-35) maintains complete remission in patient with
recurrent systemic anaplastic large-cell lymphoma. Leuk Lymphoma. 2012;
53:506-507
26. Bartlett N, Grove LE, Kennedy DA, et al. Objective responses with
brentuximab vedotin (SGN-35) retreatment in CD30-positive hematologic
malignancies: A case series. J Clin Oncol. 2010;28:15 (suppl; abstr 8062).
27. Younes A, Connors JM, Park SI, et al. Frontline therapy with brentuximab vedotin combined with ABVD or AVD in patients with newly diagnosed
advanced stage Hodgkin lymphoma). Blood. 2011;118;21 (suppl; abstr 955).
28. Moskowitz C. Risk-adapted therapy for relapsed and refractory lymphoma using ICE chemotherapy. Cancer Chemother Pharmacol. 2002;49:1
(suppl; S9-12)
29. Chen RW, Forman SJ, Palmer J, et al. Brentuximab vedotin (SGN-35)
enables successful reduced intensity allogeneic hematopoietic cell transplantation in relapsed/refractory Hodgkin lymphoma. Blood. 2011;118 (supple;
abstr 664).
EARLY DRUG DEVELOPMENT: CASTING A WIDE

NET VERSUS PRESELECTING A NARROW AUDIENCE
CHAIR
Anthony W. Tolcher, MD
South Texas Accelerated Research Therapeutics (START)
San Antonio, TX
SPEAKERS
Josep Tabernero, MD
Vall dHebron University Hospital
Barcelona, Spain
Bruce Chabner, MD
Boston, MA
Approval of New Agents after Phase II Trials

By Bruce Chabner, MD
Overview: Cancer drug approval has evolved as the understanding of cancer biology, and the ability to select patients
for trials of targeted agents, has matured. The longstanding
reliance on Phase III trials to prove drug efficacy and positive
impact on patient survival may no longer be necessary, as
early trials, particularly the expansion phase of a Phase I trial,
may provide convincing evidence of a high response rate to a
targeted drug in a patient population who has been poorly
ANCER AND AIDS represent unique diseases and

public health challenges. Both have the potential of a
fatal outcome and affect large populations of patients. Cure
or long-term control is difficult when patients present with
advanced disease. Both engender a sense of urgency related
to efforts to develop better therapies. In particular, the AIDS
epidemic prompted a rethinking of traditional paths of drug
testing and approval. For the past 50 years, since the
Kefauver legislation in 1962, drug approval by the United
States Food and Drug Administration (FDA) has required
proof of safety and efficacy through the vehicle of wellcontrolled clinical trials.1 Phase III comparisons of new and
standard treatments, with a survival end point, were the
gold standard for approval of new agents. As a viral infectious disease that spread rapidly throughout the Western
world in the early 1980s, AIDS brought forth a new paradigm: advocates campaigned for immediate access to promising agents. This advocacy led to an early implementation
of compassionate release of azidothymidine (AZT) before its
FDA approval in 1987 and the establishment in 1992 of a
new category of marketing approval referred to as accelerated approval, based on surrogate endpoints of drug efficacy.2 No longer was it necessary for patients to wait for
phase III trials to provide proof of an extension of survival.
Intermediate endpoints, such as viral load for AIDS patients
and clinical response for patients with cancer, could become
useful markers of benefit for randomized, or single-arm
phase II studies.
For the past 20 years, the accelerated approval category
has allowed early access to cancer drugs that showed promise in addressing unmet needs for treatment of potentially
fatal disease, and more than 35 new cancer drugs have
received marketing permission by this new route.3 Of the
agents approved by this mechanism, full approval has subsequently been granted to most (the exact number is not
available on the FDA website), although a small number of
agents have been withdraw for safety (gemtuzumab) or
efficacy (gefitinib in lung cancer and bevicizumab in breast
cancer) reasons. This mechanism has not been without
controversy, as both the FDA and its critics have noted a
failure to complete the required confirmatory trials postapproval in perhaps a third to one-half of accelerated approvals.3
Recent developments in cancer drug discovery have only
heightened interest in earlier approval. The rapid expansion
of knowledge regarding the genetic basis of cancer has
revealed specific molecular targets that underlie common
malignancies. At the same time, major categories of disease,
such as lung and colorectal cancer, once thought to be
relatively homogeneous, are now recognized as encompass-
responsive to conventional therapy. If the new drug produces

no safety signals of great concern, and if a validated biomarker for patient selection has been established and is
readily available, accelerated approval may be achievable
prior to completion of a randomized trial. The advantages, and
potential downside, of rapid approval scenarios will be discussed in this article.
ing a number of unique molecular entities, and in selected

cases, these entities respond to treatments that target the
underlying mutations. Thus, mutated receptor tyrosine kinases and activated intracellular signaling pathways have
become rational targets for experimental cancer therapies
(Table 1), and their early application in carefully selected
subsets of patients has met with impressive success. The
widespread adoption of this targeted approach, called
personalized or precision medicine, has had immediate consequences for cancer drug development and approval.
Recent examples of successful targeted development include crizotinib for nonsmall cell carcinoma of the lung
(NSCLC) with EML-4-ALK translocation4 and vemurafenib
for BRAF-mutated (V600E) melanoma.5 In both cases, specific selection of patients with the appropriate molecular
lesion allowed investigators to establish impressive response rates during the expansion phase of phase I trials,
encompassing 30 to 80 patients treated at the maximum
tolerated dose. In these initial patient populations, the
majority of patients received clinical benefit, response
rates exceeded 50 percent, and the clinical value of the new
agent was clear even at this early stage of development, as
alternative standard therapies were largely ineffective and
seriously toxic. The results of the phase I studies, including
the remarkable trial of GDC-0449 in patients with basal cell
cancer, are shown in Table 1 and illustrate the power of
initiating trials in rationally selected patients. In both cases,
an assay for a reliable biomarker for patient selection was
developed and implemented during the earliest phases of
the first-in-human trials and was validated by the positive
result. The biomarker assay, an essential requirement for
rapid drug development and early approval, must be reliable, reproducible, and ultimately, available nationwide.
The strategy for drug approval differed in the cases of
these two drugs, and these differences are instructive in a
discussion of early approval strategies. In the case of crizotinib, the sponsor undertook two additional trials, even as
the expanded phase I trial continued; the first was a phase
II in previously treated patients with NSCLC and the
EML-4/ALK translocation,6 and the second trial was a
randomized trial comparing the new medication to standard
combination chemotherapy, with crossover at the time of
From the Massachusetts General Hospital Cancer Center, Boston, MA.

Address reprint requests to Bruce Chabner, MD, Massachusetts General Hospital Cancer
Center, 10 North Grove Street, Boston, MA 02114; email: bchabner@partners.org.
1092-9118/10/1-10
e1
BRUCE CHABNER
Table 1. Response Rates in Expanded Cohorts of Recent Phase I Trials of Targeted Drugs in Selected Patient Populations
Response
Tumor
Mutation
Drug
No. Patients
Complete
Partial
Response
Rate (%)
Ref.
Basal Cell Cancer

NSCLC
Melanoma
smoothed in hedgehog pathway

EML4-ALK translocation
BRAF (V600E)
GDC-0449
Crizotinib
Vemurafenib
33
82
32
2
1
2
16
46
24
54
57
81
15
4
5
progression. The endpoint for the latter trial was response

rate, time to progression (TTP), and survival (despite the
crossover design).
Crizotinib received accelerated approval in August of
2011, 3 years after the first patient with EML-4/ALK
NSCLC entered the phase I trial, based on the results of the
82 patient expansion cohort of the phase I trial and the
confirmation of response rates and safety in the 173 phase II
trial. The phase III trial of crizotinib versus chemotherapy
has completed enrollment but has not yet been analyzed.
For vemurafenib, the strategy for obtaining drug approval
was somewhat different. The key registration trial was a
phase III comparing the new agent to a standard, but
ineffective drug, DTIC, which has a response rate of 10 percent.7 In fact, DTIC is the only approved chemotherapy
agent for metastatic melanoma but does not prolong survival. The endpoint was overall survival, and crossover to
the experimental drug at the time of progression was not
allowed. Vemurafenib was approved in September of 2011,
4 years after the first patient with BRAF V600E melanoma
entered the phase I trial for that drug; full marketing
approval for first-line treatment of metastatic melanoma
was based on a survival advantage versus DTIC in the
randomized phase III trial, which reached an early conclusion 1 year after receiving its first patient. The rapid
appreciation of success of the experimental drug in the
phase III trial led to amendment to allow crossover of
patients on the DTIC arm midway through the trial. The
decision to conduct a phase III trial of vemurafenib, without
crossover, provoked substantial criticism in the lay press
and editorial comment in leading medical journals, where
the ethics of comparing a clearly active agent with a drug of
minimal activity (basically a placebo) were questioned. The
decision to do so was attributed to the sponsor, which wished
to obtain exclusivity for first-line use of its drug. Interestingly, crizotinibs accelerated approval specified its use for
both first-line or later treatment of patients with stage III
(not the subject of a trial) or stage IV disease.
What are we to conclude from these recent decisions and
KEY POINTS
e2
A patient selection strategy is essential to rapid drug

approval.
Expansion phases of phase I trials allow early establishment of disease response rates.
Biomarkers for patient selection can be validated in
expanded phase I.
One confirmatory phase II may be sufficient for
accelerated approval.
Postapproval trials may refine the dose, schedule,
sequence, and combination therapies.
strategies? It seems apparent that with a valid biomarker

test in hand and appropriate patient selection, an appropriately targeted therapy can provide strong indications of its
ultimate value in an expanded cohort of subjects during a
phase I trial. A confirmatory phase II trial, which need not
be randomized if an active control is not available, can
provide sufficient evidence to convince regulatory authorities to grant accelerated approval, and the process can be
completed in three years or less. In many types of cancer, a
strong indication of activity, such as a 50 percent or greater
partial and complete response rate coupled with a time to
progression of 4 to 6 months or greater, would represent a
significant step forward and need not require preapproval
confirmation in a phase III trial, particularly if no standard
therapy provides benefit.8 One can imagine many such disease
entities, including grade glioblastomas, metastatic pancreatic
cancer, metastatic cholangiocarcinoma, hepatoma, and various subsets of soft tissue sarcoma, in which standard therapies are minimally active and significantly toxic. Given the
modest toxicity of most targeted agents, if a new agent
demonstrates impressive activity in its early trials, there is
minimal safety risk in granting accelerated approval for these
agents. Confirmatory trials postapproval would be required
and might take the form of randomized comparison of two
dose levels, intermittent compared with continuous therapy,
randomized discontinuation in patients with stable disease,
or randomization to continued therapy beyond progression
in combination with an alternative therapy. Such trials
would provide a better definition of safety, and survival
benefit, although in most cases, it would not be ethical to
deny the new agent to patients on any arm of the trial.
As an example of the potential strategy for a new targeted
agent, we might consider the development of a new drug for
patients with IDH 1 or IDH 2 mutations (Fig. 1). IDH is a
key enzyme in the intermediary metabolism of glucose in the
Krebs cycle, in which isocitrate to alpha ketoglutarate
(AKG). AKG is further utilized as a substrate for the Krebs
cycle and for more than 60 dioxygenase reactions.9,10 The
dioxygenases require AKG as a substrate in reactions that
hydroxylate methylated bases, such as methyl cytosine
residues in DNA, and methyl groups in the histones that
control gene expression; hydroxylation of these sites leads to
demethylation and changes in gene expression. Dioxygenases have other important functions. They participate in
the repair of DNA alkylation through removal of methyl or
ethyl adducts. Recent work has shown that mutations in
IDH 1 and 2 reduce AKG to a new enzymatic product,
2 hydroxyglutarate (2HG), the isomers of which function
either as potent inhibitors of dioxygenases or, in the case of
the R-isomer, activators of the same enzymes.11 Inhibition
of dioxygenase function leads to hypermethylation of DNA
and of histone H3K27.9,10 The IDH mutations, and their
product, 2HG, have transforming activity in preclinical
systems. Experimental and human tumors with underlying
APPROVAL OF NEW AGENTS
IDH mutations show hypermethylation of both DNA and

histone marks. Human tumors that display mutations in
IDH 1 or 2 include a subset of human leukemias (AML),
thyroid cancers, gliomas, chondrosarcomas, and intrahepatic cholangiocarcinomas.12
Two kinds of drugs might be effective in tumors driven by
IDH 1 or 2 mutation. Analogs of AKG/2HG that inhibit
mutant IDH enzymatic activity are an obvious choice, and
are under development by Agios and other pharmaceutical
companies.13 A second strategy would be to develop inhibitors of H3K27 methylation, and such drugs are also nearing
the clinic from Epizyme and others.14 A third strategy would
be to use inhibitors of DNA methylation, although it is not
clear whether the primary transforming event is methylation of DNA or histones. A tightly focused phase I strategy
aimed at relapsed or refractory IDH mutant AML, grade 4
gliomas, chondrosarcomas, or cholangiocarcinoma might
produce convincing early results with these agents. None of
these settings are curable with standard therapy (with the
possible exception of high dose chemotherapy/stem cell
transplant for relapsed AML). Such a trial will require the
participation of multiple institutions to identify the relatively uncommon subjects for these studies. Mutational
analysis has been developed to identify candidates for accrual to these trials. Confirmation of target engagement will
require evaluation of 2HG levels in tumor cells, serum or
urine, or by imaging, and analysis of DNA and histone
methylation before and after treatment. If response rates
and TTP reach impressive levels in any of these subject
tumor subsets, and provided that the safety profile of the
drug is acceptable, there would be an obvious path to
accelerated approval.
Despite the early success of targeted therapies in various
settings, it is apparent that single agent treatment will not
cure metastatic disease. Combination therapies aimed at
mechanisms of resistance will be required to derive longterm benefit. Proving the value of combination trials will
likely require randomized phase III studies to achieve marketing approval. The phase III trial is not antiquated or
obsolete. It simply needs to be adapted to the new paradigm
of rapid, single agent approval. Its continued place in the
sun is assured.
Author
Bruce Chabner*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. United States Food And Drug Administration. Significant Dates in U.S.
Food and Drug Law History. http://www.fda.gov/AboutFDA/WhatWeDo/
History/Milestones/ucm128305.htm. Last accessed 3/7/12. Accessed October
14, 2010.
2. Broder S. The development of antiretroviral therapy and its impact on
the HIV-1/AIDS pandemic. Antiviral Res. 2010;85:1-18.
3. Johnson JR, Ning YM, Farrell A, et al.Accelerated approval of oncology
products: The food and drug administration experience. J Natl Cancer Inst.
2011;103:636-44.
4. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic Lymphoma Kinase
Inhibition in Non-Small-Cell Lung Cancer. N Engl J Med. 2010;363:16931703.
5. Flaherty KT, Puzanov I, Kim KB, et al. Inhibition of mutated, activated
BRAF in metastic melanoma. N Engl J Med. 2010;26:809-19.
6. FDA approves Xalkori with companion diagnostic for type of late-stage
cancer.
http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/
ucm269856.htm. Last Accessed March 7, 2012.
7. Chapman PB, Hauschild A, Robert C, , et al. Improved survival with
vemurafenim in melanoma with BRAF V600E mutation. N Engl J Med.
2011;364:2507-2516.
8. Chabner BA. Early accelerated approval for highly targeted cancer

drugs. N Engl J Med. 2011;364:1087-1089.
9. Chao Lu, Ward PS, Kapoor GS, et al. IDH Mutation impairs histone
demethylation and results in a block to cell differentiation. Nature. Epub 2012
Feb 12.
10. Turcan S, Rohle D, Goenka A, , et al. IDH1 mutation is sufficient to
establish the glioma hypermethylator phenotype. Nature. Epub 2012 Feb. 15.
11. Koivunen P, Lee S, Duncan CG, Lopez G, Lu G, et al. Transformation
by the (R)-enantiomer of 2-hudroxyglutarate linked EGLN activation. Nature.
Epub 2012 Mar 22.
12. Borger DR, Tanabe KK, Fran KC, et al. Frequent mutation of isocitrate
dehydrogenase (IDH)1 and IDH2 in cholangiocarcinoma identified through
broad-based tumor genotyping. Oncologist. 2012;17:72-79.
13. Yen KE, Schenkein DP. Cancer-associated isocitrate dehydrogenase
mutations. Oncologist. 2012;17:72-79.
14. Copeland RA, Solomon ME, Richon VM. Protein methyltransferases as
a target class for drug discovery. Nat Rev Drug Discov. 2009;8:724-732.
15. Von Hoff DD, LoRusso PM, Rudin CM, et al. Inhibition of the hedgehog
pathway in advanced basal-cell carcinoma. N Engl J Med. 2009;361:11641172.
e3
Drug Development in the Era of Personalized

Oncology: From Population-Based Trials to
Enrichment and Prescreening Strategies
By Rodrigo Dienstmann, MD, Jordi Rodon, MD, and Josep Tabernero, MD
Overview: Recent advances in tumor biology and human

genetics along with the development of drugs for specific
targets hold promise for an era of personalized oncology
treatment. Routine use of modern technologies, such as
large-scale genome sequencing, will help to unravel the specific biology of each tumor. Adding a rigorous genomic view
could determine key genetic events, critical dependencies,
and stratification of patients in early clinical trials. Integrating
biomarker development into the early testing of novel agents
might provide clinically relevant therapeutic opportunities for
patients with advanced-stage cancer and also accelerate the

drug-approval process. After recent success stories of therapies targeting driver molecular aberrations in genetically
defined tumor subtypes, innovative clinical trials based on a
strong biologic hypothesis are expected to bring further
excitement to the field. In this article, we describe a new trend
in biomarker-driven early drug development using enrichment
and prescreening strategies. Technical and logistical obstacles that may hinder progress of this approach will be discussed, along with ethical and economic concerns.
RUG DEVELOPMENT in oncology remains a slow,

costly, and inefficient process. Taking advantage of
new science in the design and execution of clinical investigation, with a switch from histology-driven therapy to molecularly driven clinical oncology, is one way of addressing
the high failure rate of developing an innovative anticancer
agent.1 Targeted therapeutics is optimal when applied in the
appropriate molecular context: it indicates a drug that
interferes with the function of a molecule readily identified
in cancer cells but not healthy tissues. This target plays a
critical role in the abnormal growth and/or invasiveness of
the tumor. A relative specificity of the drug exists, leading to
improvement in the therapeutic index over standard cancer
chemotherapy, higher efficacy, and decreased unnecessary
toxicity. Accomplishments of modern technologies in genomics and biomarker detection are providing information that
enables clinical researchers to investigate selective therapies for individual patients with cancer. Hundreds of new
experimental drugs and biologic agents target the products
of aberrant genes that are mutated or abnormally expressed
in human cancers.2 Many of these approaches are based on
concepts such as oncogene addiction, non-oncogene addiction, and synthetic lethality, which target distinctive and
complementary capabilities that enable tumor growth and
metastatic dissemination.3,4
Given this perspective, phase I trials are providing an
arena for early hypothesis testing.5 Molecular biomarkerbased patient selection in early clinical trials has many
possible advantages over an unselected population-based
approach and include: (a) clinically qualify potential predictive biomarkers for molecularly targeted agents (MTAs)
as early as possible; (b) generate valuable information regarding cancer biology; (c) accelerate patient benefit; and
(d) affect the drug development process by assisting in the
goversus-no-go decisions and changing drug approval
registration strategies of promising agents.6,7 In this manuscript, we describe a new trend in early drug development
with enrichment and prescreening strategies for personalized oncology.
nosis into one that was predictive of benefit from the drug,
was a major advance in the field.8 More recently, patients
with advanced gastric and gastroesophageal junction adenocarcinomas overexpressing HER2 have also been shown to
achieve improved survival when trastuzumab is combined
with chemotherapy. Another example is the PARP inhibitor
olaparib, developed for patients with BRCA1/2-mutant cancers, with antitumor activity observed in different malignancies, including breast, ovarian, and prostate.9 Further
excitement for this approach came with the recent regulatory approvals of vemurafenib for advanced melanoma that
harbors BRAF V600E mutation and crizotinib for nonsmall
cell lung cancer (NSCLC) with EML4-anaplastic lymphoma
kinase (ALK) translocation.10,11 Interestingly, responses
with selective BRAF inhibitors were seen in other tumor
types carrying BRAF V600E mutations, such as thyroid
cancer. The same is true for crizotinib in inflammatory
myofibroblastic tumor and anaplastic large-cell lymphomas
harboring ALK translocation. The hedgehog inhibitor vismodegib has also been developed with enrichment for specific
target patient populations in early clinical trials. Clear signs
of activity were seen both in basal cell carcinomas and
medulloblastomas known to have activating mutations in
the pathway components Patched or Smoothened.12 This
highlights that cancers should likely be treated with targeted agents based on their specific molecular alterations
rather than organ of origin.
The impressive response rate in the selected population of
patients included in the particular trials listed in Table 18-16
is in contrast with the approximate 5% response observed
in early-phase trials conducted in unselected patient populations.17,18 Based on preclinical data, potential enrichment
biomarkers already under investigation include PIK3CAactivating mutations and PTEN loss of expression in early
trials with PI3K pathway inhibitors, with focus on breast
and gynecologic tumors. Clinical development of chaperone
(HSP90) inhibitors is now focused on ALK-rearranged
Success Stories
From the Molecular Therapeutics Research Unit, Vall dHebron University Hospital,
Barcelona, Spain.
Address reprint requests to Josep Tabernero, P. Vall dHebron 119-129, Barcelona, Spain
08035; email: jtabernero@vhebron.net.
1092-9118/10/1-10
The success of several targeted agents has shown that

personalized cancer treatments can have an enormous effect. Development of trastuzumab in HER2-amplified breast
cancers, with conversion of a biomarker for negative prog-
168
DRUG DEVELOPMENT AND PERSONALIZED ONCOLOGY

Table 1. Response Rate of Successful Targeted Therapies in Selected Populations Evaluated in Early Clinical Trials
Marker/Population
HER2-overexpressed/amplified breast cancer

CD117-overexpressed GIST
BRCA1/2 mutant breast, ovarian and prostate cancer
BRAF V600E-mutant melanoma
ALK-rearranged NSCLC
Basal cell carcinomas (majority have inactivating
mutations in PTCH1 or activation of SMO)
Medullary thyroid cancer (known to have RET mutations,
MET expression and VEGF activation)
Agent
Mechanism of Action
Response
Reference
Trastuzumab
Trastuzumab-DM1
Imatinib
Olaparib
Vemurafenib
Dabrafenib*
Crizotinib
Vismodegib
Anti-HER2 antibody
Anti-HER2 antibody drug conjugate
c-KIT, PDGFR inhibitor
PARP inhibitor
BRAF inhibitor
BRAF inhibitor
ALK, MET inhibitor
SMO inhibitor (Hh pathway)
12%
44%
54%
47%
75%
60%
57%
58%
8
13
14
9
10
15
11
12
Cabozantinib
MET, VEGFR2, RET inhibitor
29%
16
* GSK2118436.
Abbreviations: GIST, gastrointestinal stromal tumor; PDGFR, platelet-derived growth factor receptor; PARP, poly(ADP-ribose) polymerase; ALK, anaplastic lymphoma
kinase; NSCLC, nonsmall cell lung cancer; MET, mesenchymal epithelial transition; SMO, smoothened; Hh, hedgehog; VEGFR, vascular endothelial growth factor
receptor.
NSCLC and HER2-amplified breast tumors. Mesenchymal

epithelial transition (MET) amplification, mutations, or receptor overexpression have been linked to responses with
MET inhibitors in different tumor types, such as gastric
adenocarcinomas, papillary renal cell carcinoma, and
NSCLC. The same approach is being implemented in phase
I trials evaluating fibroblast growth factor receptor (FGFR)
inhibitors in FGFR amplified tumors, such as breast cancer
and squamous NSCLC.
Challenges
Many investigators have raised concerns over the use of

predictive biomarkers in early clinical trials, including inac-
KEY POINTS
The one-size-fits-all approach for drug development

of molecularly targeted agents in solid tumors has
shown disappointing results.
With the switch from histology-driven to molecularly
driven therapy, phase I trials in clinical oncology are
providing an arena for clinical trial testing.
Genomics-driven early clinical trial enrollment has
many possible advantages over an unselected
population-based approach, including clinical qualification of predictive biomarkers, accelerated patient
benefit, and a positive impact on the drug development process.
Some concerns over the use of biomarkers in early
clinical trials have been raised, including the use of
assays that are not validated or certified, incorrect
patient selection, increased cost, logistical issues related to the prescreening strategies, and ethical issues regarding mandatory tumor biopsies.
For the accelerated approach of drug development of
molecularly targeted agents to be possible, a cooperative environment is necessary, with experienced
academic institutions, regulatory authorities, and
pharmaceutical and diagnostic companies working
together to promote data sharing, standardized procedures, technology exchange, and avoidance of duplicative efforts.
curate assays, high cost, and ethical issues regarding tumor

biopsies. Most importantly, the potential beneficial effects
of the MTA in a more broadly defined population could be
missed on the basis of incorrect patient selection.6 Nevertheless, unselected patient evaluation has a higher risk of
failure in late-stage trials because of the tumors molecular
heterogeneity, unless the prevalence of the predictive biomarker is already known to be high in the disease overall.5
It is clear that many technical, laboratory, and clinical
challenges still remain to be resolved before the widespread
implementation of molecular biomarker-based patient selection in early clinical trials. At first, investigators need to be
aware of the possible reasons for failure of MTAs matched to
specific biomarkers of vulnerability in advanced disease.
The presence of the biomarker may not be representative of
the disease (tumor heterogeneity) or its biology (driver vs.
passenger genetic alteration). Analyzing the tumor sample
from disease diagnosis may not reflect the molecular alterations that drive the metastatic disease as a result of further molecular alterations (clonal evolution), selection pressure from previous treatments, and tumor heterogeneity.19
Driver mutations that promote growth and metastasis are
considered the primary targets for therapeutic intervention.
Nevertheless, passenger molecular dysregulations that
arise through the rapid growth and genetic instability of
tumor cells can confer resistance to specific therapies.
Secondly, the predictive test may not be accurate and the
subpopulation most likely to benefit from MTAs may not be
readily and reliably identified. Methodological problems,
including a threshold (cutoff point) for defining the status
of the potential biomarker, may still be missing. It is also
important to note that the strategy of matching a predictive
biomarker with MTAs will not always be applicable to all
novel therapies, for example, broad-spectrum inhibitors targeting multiple signaling pathways and antiangiogenic
agents. In addition, some MTAs may have substantial activity but no identifiable predictive marker, such as histone
deacetylase (HDAC) and proteasome inhibitors. Therefore,
a strong biologic hypothesis, solid preclinical data, and
tumor models that mimic the clinical disease accurately are
needed in order to assist in the drug development process.5
Thirdly, rapid acquisition of resistance to single agents
is a problem. Complex interacting networks with adaptive
negative feedback and compensatory receptor tyrosinekinase stimulation are imperative in cancer cells. In addition to cross-pathways escape, there is always the possibility
169
DIENSTMANN, RODON, AND TABERNERO
that the core pathway is not completely blocked with the

selected MTA. Furthermore, in tumors that have similar
molecular aberrations, variability in patient outcomes has
been observed, such as selective BRAF inhibitors in BRAFmutant melanoma and colorectal cancer. Single genomic
markers are often found to be unsuitable as biomarkers in
clinical studies, and probably a biomarker signature will
eventually be required to predict a response.20 This points
toward a broader approach when analyzing tumor samples
to include molecular profiling of several cancer genes, such
as mutations, rearrangements, and somatic copy-number
alterations. In addition, for true personalized medicine, one
should also take into consideration not only characteristics
of the tumor but also the relationship host-tumor (tumor
microenvironment and immune response) and host-drug
(metabolism genes and pharmacogenomics).19
Additional Issues to Consider for Genomics-Driven
Early Clinical Trial Enrollment
Do We Need Certified and Validated Biomarkers?
An increasingly large number of putative biomarkers

using innovative sophisticated technologies are being used
in phase I trials. The lack of fully validated and reproducible
assays that can be conducted in appropriately certified
laboratories operating according to Clinical Laboratory Improvement Amendments (CLIA) or Good Clinical Laboratory
Practice (GCLP) standards is a major concern. In order to
deal with this obstacle, De Bono and colleagues propose a
parallel and simultaneous predictive biomarker/clinical
anticancer drug development process: in early clinical trials,
when patient selection is many times based on the best
guess, pharmacodynamic biomarkers must follow very rigorous criteria in order to define proof-of-mechanism. Conversely, predictive biomarkers for selecting patients could
be explored according to less strict standards.21 Successfully
enriching phase I trials with patients whose tumors harbor
specific molecular aberrations that may predict response
can demonstrate proof-of-concept and encourage further
research with a given drug or target. Lack of anticancer
effect in the best-case scenario, provided that sufficient
target inhibition is achieved, may ultimately redefine drug
development strategies. Importantly, delineating a selected
patient population does not restrict the late development of
a specific drug to that subpopulation. If predictive biomarkers are proven robust and potentially useful in early clinical
trials, they can be further clinically qualified through prospective evaluation in large randomized controlled trials
before regulatory approval. In other cases in which the
predictive value is not very high, and, therefore, the clinical
utility of the biomarker is not obvious, a randomized and
stratified phase II trial is needed to assess both the new drug
and the corresponding biomarker.20
How Can We Identify Suitable Candidates?
Targeted agents are developed to treat small subsets of

patient populations, and the operations to perform trials
with the old methods become inefficient. Many difficulties in
recruiting suitable patients have to be addressed. The most
evident is the time to perform molecular analysis to identify
targetable aberrations in the context of an early drug development program.
The traditional approach is to prescreen patients by send-
170
ing their tumor tissue to a central laboratory for analysis

(either the sponsor of the trial or a vendor) just before
considering the enrollment of the patient in a trial with an
MTA. This strategy usually demands availability of large
amounts of tumor tissue, which may be a limitation when
scarce material was used for diagnostic purposes, patients
were enrolled in previous trials, or there is competing
in-house academic research. In addition, because of the
delay during the prescreening process and clinical deterioration at the time of recruitment, patients may miss the
opportunity to receive a promising agent.
The alternative strategy calls for local prescreening at
academic institutions and analyzing tumor samples of patients who are still receiving standard treatment for advanced disease. Molecular selection can be performed at any
time during the disease course. Advantages include the
requirement of only one consent form (based on a protocol
approved by the local institutional review board) and the
reduction in the time from disease progression with standard treatment until enrollment in a clinical trial with an
MTA. Nevertheless, upfront tumor analysis has one big
hurdle: its cost is usually not covered by health care providers, neither national health systems nor private insurance
companies. Building the cost of a broad molecular analysis
into the budget of a specific trial is not feasible because it
precedes inclusion in any given trial. Finally, there is always
the possibility of not having a slot at the time of progression or the patient not being eligible for a particular trial
with matched MTA as a result of inclusion/exclusion criteria
being too strict.
Another important dilemma is the magnitude and type of
molecular analysis that is likely to be informative for definition of tumor vulnerability and corresponding targeted
therapy. Usually, targetable molecular aberrations are examined, including specific mutations, common gene amplifications/translocations, and selected protein-expression
levels. Screening for single mutations has now been replaced
by multiassay platforms, and several are now available both
commercially and at academic centers that evaluate simultaneously for mutations in 40 to 200 genes (Sequenom,
SNaP shot, among others). Such platforms require very little
tissue, can be performed quickly, and are less expensive.
Academic institutions are already starting to run pilot
projects, including comparative genomic hybridization,
microarray-based profiling of gene expression, massively
parallel sequencing, exome, and whole-genome sequencing.22,23 In this regard, there is the issue of whether crossvalidation of molecular profiling results should be performed
at each single institution or at a centralized laboratory
serving many institutions in the same region. In addition,
translating high-throughput sequencing for biomarkerdriven clinical trials for personalized oncology presents
unique logistical challenges, including the development of
an informed-consent process that addresses how to handle
incidental findings, the selection of the results that should
be disclosed to patients, and the implementation of efficient
and integrative computational pipelines for data analysis.22
Should Tumor Biopsies Be Mandatory?
Tissue that is collected as part of a research protocol (at

baseline, during treatment, or at the time of tumor progression) has great potential to advance scientific knowledge
by determining how well a drug is affecting the target tissue,
DRUG DEVELOPMENT AND PERSONALIZED ONCOLOGY

Fig 1. Matching tumor type, molecular aberration, screening platform, and
early clinical trial with targeted agent.
In the first scenario (left to right), a patient with ovarian cancer (whose tumor
may present PI3K pathway aberrations, RAS, RAF or BRCA mutations) will
have molecular prescreening with multiple assays and will be ultimately referred to a corresponding phase I trial.
On the other hand (right to left), if the
screening is performed according to
the molecular target agent available
at the site (e.g., FGFR inhibitor), screening will be focused on different tumor
types, such as HR positive breast cancer
and squamous NSCLC.
Abbreviations: HR, hormone receptor;
NSCLC, non-small cell lung cancer; CRC,
colorectal cancer; IHC, immunohistochemistry; FISH, fluorescence in situ hybridization; CGH, comparative genomic
hybridization; PCR, polymerase chain
reaction.
identifying mechanisms of drug resistance, describing new

oncologic signaling pathways, or establishing predictors of
a favorable or unfavorable outcome.24 To date, many assays
still require fresh, frozen samples for technical reasons.
Biomarker analysis can sometimes directly benefit the patient (for example, in breast cancer relapses presenting a
change in hormonal receptor and HER2 status) and also
future patients with cancer. Mandatory biopsies are acceptable when investigators appropriately weigh the risks
against the necessity of the correlative question. However,
various hurdles must be faced when protocols mandate
biopsies. Ethical concerns have been raised when the participants enrollment in a clinical trial depends on his or her
consent to undergo a research biopsy. Investigators may feel
forced to find suitable on-trial or off-trial options for patients
whose tumors have been biopsied and undergone molecular
profiling. The risks of the procedure also need to be taken
into consideration. However, according to the experience of
large centers, biopsies in early phase clinical trials are safe,
with less than 1.5% risk of serious complications.25
Ongoing Innovative Trials
Clinical trial design needs to be adjusted to a new era of

personalized oncology. Novel approaches include histologyindependent trials (for example, patients with BRCA1/2
mutated tumors, regardless of histology), window-ofopportunity trials (when standard anticancer therapy is
delayed, and patients receive first a matched MTA allowing
chemotherapy-free intervals) or trials that subclassify a
specific disease into discrete molecular categories (such as
the Investigation of Serial Studies to Predict Your Therapeutic Response with Imaging and Molecular Analysis [ISPY 2] trial in breast cancer and the Biomarker-integrated
Approaches of Targeted Therapy for Lung Cancer Elimination [BATTLE] trial in NSCLC).26,27 For these approaches
to be possible, a cooperative environment is necessary, with
experienced academic institutions, regulatory authorities,
and pharmaceutical and diagnostic companies working together in order to promote data sharing, standardized procedures, technology exchange, and avoidance of duplicative
efforts.2 Initiatives, such as Worldwide Innovative Network
(WIN) Consortium, are a great example of this collaboration
toward a common goal. Together, investigators can utilize

structured methods to define when a new agent should cease
further study and when it has demonstrated sufficient
likelihood of potential benefit to proceed to the next phase of
testing.
Preliminary results from personalized medicine programs
suggest that molecular analysis of tumors and the use of
MTAs to counteract the effects of specific aberrations improves the outcomes of affected patients. In a seminal trial,
Von Hoff and colleagues used molecular profiling (based
on immunohistochemistry, immunofluorescence, and microarray analysis) to select a targeted agent and were able to
prove that this approach is feasible. In 27% of the patients,
treatment based on molecular analysis resulted in longer
progression-free survival than the prior unmatched therapy.28 Similar results were seen in an innovative model for
phase I clinical trials performed at the University of Texas
M. D. Anderson Cancer Center. Matching patients with an
MTA in early clinical trials, mainly based on mutation
status of frequently activated genes in cancer (KRAS,
NRAS, BRAF, PIK3CA, EGFR, CKIT) and PTEN protein
expression levels, resulted in longer time to treatment
failure (5.3 vs. 3.2 months) compared with their prior systemic antitumor therapy. In addition, for patients with one
molecular aberration, the response rate was 29% with
matched targeted therapy versus 8% without matching,
highlighting the potential improved outcomes of heavily
pretreated patients enrolled in genomic-driven phase I trials
of MTAs.18 Based on these and other observations, our
institute, the Massachusetts General Hospital, the Royal
Marsden Cancer Center, and the Institute Gustave Roussy,
among others, are already building flexible programs
(adapting the putative predictive biomarkers and the clinical trials available on-site) to serve their phase I trials.
Figure 1 exemplifies different ways of matching tumor type,
molecular aberration, screening platform, and early clinical
trials with MTAs.
Conclusion
Phase I trials in which an MTA can be matched precisely

with a population of patients whose tumors are driven
predominantly by the target of interest are still uncommon.
171
DIENSTMANN, RODON, AND TABERNERO
This situation is not surprising, given that most advanced

human malignancies have complex signaling networks. Nevertheless, the one-size-fits-all approach for drug development of MTAs in solid tumors has shown disappointing
results in recent years. It is no longer viable to develop
targeted drugs without a reliable indicator of a potential
benefit derived from them. The genomic-driven accelerated
approach of drug development is applicable when the appropriate patient population can be evaluated in the phase I
trial. This conceptual shift of clinical development of pairing
therapies and biomarkers will require a change in the design
and execution of clinical trials to one where antitumor
activity of MTAs is evaluated. Hopefully, the decreasing cost

of high-throughput molecular technologies will enable more
comprehensive characterization of the numerous additional
tumor genome alterations, and, therefore, facilitate a personalized oncology approach.29 In addition, alternative techniques to biopsy of the metastatic disease might be
considered in the near future, such as molecular imaging
and genomic analysis of circulating tumor cells, or tumor
DNA detected in peripheral blood.30 In conclusion, biologic
evidence, clinical data, and previous experience propose that
it is time for a change in the current way drug development
of targeted agents is done.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Rodrigo Dienstmann*
Jordi Rodon*
Josep Tabernero
Amgen; BristolMyers Squibb;

Genentech;
Merck KGaA;
Novartis; Onyx;
Pfizer; Roche;
Sanofi
Amgen; Merck
KGaA; Novartis;
Roche; Sanofi
REFERENCES
1. de Bono JS, Ashworth A. Translating cancer research into targeted
therapeutics. Nature. 2010;467:543-549.
2. Mendelsohn J. A national cancer clinical trials system for targeted
therapies. Sci Transl Med. 2011;3:75cm8.
3. Luo J, Solimini NL, Elledge SJ. Principles of cancer therapy: oncogene
and non-oncogene addiction. Cell. 2009;136:823-837.
4. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation.
Cell. 2011;144:646-674.
5. Yap TA, Sandhu SK, Workman P, et al. Envisioning the future of early
anticancer drug development. Nat Rev Cancer. 2010;10:514-523.
6. Carden CP, Sarker D, Postel-Vinay S, et al. Can molecular biomarkerbased patient selection in Phase I trials accelerate anticancer drug development? Drug Discov Today. 2010;15:88-97.
7. Garrido-Laguna I, Hidalgo M, Kurzrock R. The inverted pyramid of
biomarker-driven trials. Nat Rev Clin Oncol. 2011;8:562-566.
8. Baselga J, Tripathy D, Mendelsohn J, et al. Phase II study of weekly
intravenous recombinant humanized anti-p185HER2 monoclonal antibody
in patients with HER2/neu-overexpressing metastatic breast cancer. J Clin
Oncol. 1996;14:737-744.
9. Fong PC, Boss DS, Yap TA, et al. Inhibition of poly(ADP-ribose)
polymerase in tumors from BRCA mutation carriers. N Engl J Med. 2009;
361:123-134.
10. Flaherty KT, Puzanov I, Kim KB, Ribas A, et al. Inhibition of mutated,
activated BRAF in metastatic melanoma. N Engl J Med. 2010;363:809-819.
11. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase
inhibition in non-small-cell lung cancer. N Engl J Med. 2010;363:1693-1703.
12. LoRusso PM, Rudin CM, Reddy JC, et al. Phase I trial of hedgehog
pathway inhibitor vismodegib (GDC-0449) in patients with refractory, locally
advanced or metastatic solid tumors. Clin Cancer Res. 2011;17:2502-2511.
13. Krop IE, Beeram M, Modi S, Jones SF, et al. Phase I study of
trastuzumab-DM1, an HER2 antibody-drug conjugate, given every 3 weeks to
patients with HER2-positive metastatic breast cancer. J Clin Oncol. 2010;28:
2698-2704.
14. Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety of
imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl
J Med. 2002;347:472-480.
15. Kefford R, Arkenau H, Brown MP, et al. Phase I/II study of
GSK2118436, a selective inhibitor of oncogenic mutant BRAF kinase, in
patients with metastatic melanoma and other solid tumors. J Clin Oncol.
2010;28:abstract 8503.
172
16. Kurzrock R, Sherman SI, Ball DW, et al. Activity of XL184 (Cabozantinib), an oral tyrosine kinase inhibitor, in patients with medullary thyroid
17. Horstmann E, McCabe MS, Grochow L, et al. Risks and benefits of
phase 1 oncology trials, 1991 through 2002. N Engl J Med. 2005;352:895-904.
18. Tsimberidou AM, Iskander NG, Hong DS, et al. Personalized medicine
in a phase I clinical trials program: The MD Anderson Cancer Center
Initiative. J Clin Oncol. 2011;29:abstract CRA2500.
19. Gasparini G, Longo R. The paradigm of personalized therapy in
oncology. Expert Opin Ther Targets. 2011 Nov 11. [Epub ahead of print]
20. Coyle VM, Johnston PG. Genomic markers for decision making: what is
preventing us from using markers? Nat Rev Clin Oncol. 2010;7:90-97.
21. Garcia VM, Cassier PA, De Bono JS. Parallel anticancer drug development and molecular stratification to qualify predictive biomarkers: Dealing
with obstacles hindering progress. Cancer Discovery. 2011;1:207-212.
22. Roychowdhury S, Iyer MK, Robinson DR, et al. Personalized oncology
through integrative high-throughput sequencing: A pilot study. Sci Transl
Med. 2011;3:111ra121.
23. Wagle N, Berger MF, Davis MJ, et al. High-throughput detection of
actionable genomic alterations in clinical tumor samples by targeted, massively parallel sequencing. Cancer Discovery. 2012;2:82-93.
24. Olson EM, Lin NU, Krop IE, Winer EP. The ethical use of mandatory
research biopsies. Nat Rev Clin Oncol. 2011;8:620-625.
25. El-Osta H, Hong D, Wheler J, et al. Outcomes of research biopsies in
phase I clinical trials: the M. D. Anderson Cancer Center experience.
Oncologist. 2011;16:1292-1298.
26. Barker AD, Sigman CC, Kelloff GJ, et al. I-SPY 2: an adaptive breast
cancer trial design in the setting of neoadjuvant chemotherapy. Clin Pharmacol Ther. 2009;86:97-100.
27. Kim ES, Herbst RS, Wistuba II, et al. The BATTLE Trial: Personalizing
therapy for lung cancer. Cancer Discovery. 2011;1:44-53.
28. Von Hoff DD, Stephenson JJ Jr., Rosen P, et al. Pilot study using
molecular profiling of patients tumors to find potential targets and select
treatments for their refractory cancers. J Clin Oncol. 2010;28:4877-4883.
29. MacConaill LE, Garraway LA. Clinical implications of the cancer
genome. J Clin Oncol. 2010;28:5219-5228.
30. Devriese LA, Voest EE, Beijnen JH, Schellens JH. Circulating tumor
cells as pharmacodynamic biomarker in early clinical oncological trials.
Cancer Treat Rev. 2011;37:579-589.
IMMUNOTHERAPY WITH THE T-CELL CHECKPOINT

ANTIBODIES: IMPLICATIONS FOR THE
PRACTITIONER
CHAIR
Mario Sznol, MD
Yale Cancer Center
New Haven, CT
SPEAKERS
Jeffrey S. Weber, MD, PhD
H. Lee Moffitt Cancer Center & Research Institute
Tampa, FL
F. Stephen Hodi, MD
Boston, MA
Practical Management of Immune-Related

Adverse Events from Immune Checkpoint
Protein Antibodies for the Oncologist
By Jeffrey S. Weber, MD, PhD
Overview: Monoclonal antibodies directed against immune

checkpoint proteins, such as cytotoxic T-lymphocyte
antigen-4 (CTLA-4) or programmed death-1 (PD-1), can boost
endogenous immune responses directed against tumor cells.
Recently, ipilimumab was approved by the U.S. Food and Drug
Administration (FDA) for the treatment of metastatic melanoma, and the anti-PD-1 antibody BMS-936558 has shown
promising results in patients with melanoma, non-small cell
lung cancer, and renal cell cancer. During treatment with
these antibodies, a unique set of toxicities occur called
immune-related adverse events (irAEs). These irAEs may occur at any time during treatment and include colitis characterized by a mild to moderate but occasionally severe and
persistent diarrhea. Hypophysitis, hepatitis, pancreatitis, iri-
HE DOCUMENTED efficacy of the anti-CTLA-4 antibody ipilimumab that led to its FDA approval for
treatment of metastatic melanoma suggests that dissemination of knowledge about and familiarity with its unique side
effects is important for the oncology community.1,2 Treating
physicians must be able to quickly detect and adequately
treat the toxicities of ipilimumab. The management of these
toxicities requires the collaborative efforts of a multidisciplinary team led by a physician and including the nursing/
midlevel staff who are routinely in contact with the patient.
In addition, understanding irAE management is important
for consulting specialists, such as gastroenterologists, endocrinologists, hepatologists, dermatologists, surgeons, and
others. Therapeutic algorithms have been developed for the
management of irAEs and are part of the package insert for
ipilimumab, yet it is felt that a review of the management
of these novel toxicities might be useful for those who treat
patients with melanoma.3
The irAE: What Is It? When Does It Happen?
Where Does It Happen?
docyclitis, lymphadenopathy, neuropathies, and nephritis

have also been reported with ipilimumab, and a subset of
those side effects has also been observed with BMS-936558.
Patient and physician education as well as good patient
caretaker communication are keys to limiting the morbidity of
irAEs. Early recognition of these irAEs and initiation of treatment are critical to reduce the risk of complications, since
virtually all irAEs are reversible with the use of steroids and
other immune suppressants. The onset of grade 3 to 4 irAEs
correlated with treatment response in some ipilimumab studies. This article provides detailed description and recommendations for practicing oncologists to manage the common
irAEs associated with antibodies against immune checkpoint
blockade.
related adverse events occurred after a period of 2 to 3 weeks

from treatment (1 dose). Thereafter, GI and hepatic adverse
events began to occur within 6 to 7 weeks (2 to 3 doses), and
finally endocrinologic side effects were noted after an average of 9 weeks (3 to 4 doses) since the last dose of ipilimumab. The frequencies of grade 3 to 4 ipilimumab-related
irAEs increased with dose and were not observed at 0.3
mg/kg. Few irAEs were seen at 1 mg/kg, and they were
present in 10% of patients at the 3 mg/kg dose level, rising to
over 20% of patients at 10 mg/kg.7 The long half life of
ipilimumab may be consistent with slow resolution of many
irAEs. The primary approach to grade 1 to 2 irAEs is
supportive and symptomatic care; for higher grade irAEs,
steroids by mouth or parenterally are given, and either
skipping a dose or stopping therapy is appropriate. Recurrent grade 2 irAEs may also mandate skipping a dose or the
use of steroids. Algorithms for toxicity described below,
available online in the ipilimumab package insert, describe
detailed management guidelines that are tailored to the
severity and nature of the particular irAE.
Colitis
Early in ipilimumabs development, it became clear that it

induced dose-related, immune-related, or inflammatory side
effects. The most common systems affected were the skin,
gut, liver, and pituitary. Immunohistochemistry of affected
skin and gut revealed infiltration by CD4 and CD8 T cells,
and highly activated effector cells correlated with side effect
intensity.4 Elevated inflammatory cytokines in the sera, as
well as rapid resolution of some irAE symptoms with use of
the tumor necrosis factor-alpha antibody infliximab, suggested that cytokine release by activated T cells was associated with irAEs.5 Allergic and anaphylactic reactions during
infusions were rare, since ipilimumab is a human antibody.
In a pooled analysis of 325 patients treated with ipilimumab
10 mg/kg four times every three weeks, drug-related overall
adverse events were observed in 84.6% of patients, of which
immune-related adverse events of any grade accounted
for 72.3%.6 Grade 3 or 4 irAEs were observed in 25.2% of
patients, mainly in the gastrointestinal (GI) tract (12%),
liver (7%), skin (3%), and endocrine organs (3%). These
adverse events exhibited a characteristic timing. Skin-
174
Diarrhea occurs in up to 44% of patients receiving

ipilimumab.2,8-12 Diarrhea of grade 3 or 4 (7 diarrheal bowel
movements above baseline in 24 hours) was reported in 18%
of patients at the dose of 10 mg/kg.2 Diarrhea of grades 3/4
can cause weakness, imbalanced electrolytes, and weight
loss. Diarrhea can also be associated with colitis, which, if
unattended, can lead to obstruction and bowel perforation.
To date, very few cases of bowel perforation leading to
colectomy have been reported with ipilimumab therapy.8,13
Unlike chronic inflammatory bowel diseases (IBD), such
as Crohns disease or ulcerative colitis, colitis associated
with ipilimumab predominantly involves the descending
From the Moffitt Cancer Center, Tampa, FL.

Address reprint requests to Jeffrey S. Weber, MD, PhD, Moffitt Cancer Center, 12902
Magnolia Dr., SRB-2, Tampa, FL; email: jeffrey.weber@moffitt.org.
1092-9118/10/1-10
CHECKPOINT PROTEIN INHIBITOR TOXICITY MANAGEMENT
colon.12 Neutrophilic infiltrates are seen in 46% of patients,

lymphocytic infiltrates in 15%, and a mixed neutrophiliclymphocytic infiltrate in 38%.12,14
Low-grade diarrhea (grade 1, an increase of 2 over baseline in 24 hours) should be treated symptomatically using
loperamide, oral hydration, and electrolyte substitution. The
American Dietary Association colitis diet is useful. With
persistent or higher-grade diarrhea, bacterial or parasitic
infection, viral gastroenteritis, or the first manifestation of
an IBD must be ruled out by examination for stool leukocytes, stool cultures, and a Clostridium difficile titer. Grade
2 diarrhea can be treated with the addition of oral diphenoxylate hydrochloride and atropine sulfate four times daily
and budesonide 9 mg daily. Endoscopy is recommended to
confirm or rule out colitis with persistent grade 2 diarrhea or
grades 1 to 2 diarrhea with bleeding. Endoscopy can generally be done safely in patients with the GI side effects from
ipilimumab, although there is the risk of perforation with
biopsies. The presence and nature of colitis may change
ones therapeutic management, since diffuse ulceration and
bleeding in the setting of grade 2 diarrhea may mandate a
course of oral steroids and skipping a dose and can also
represent an increased risk for the development of a bowel
perforation.
For grade 3 or 4 diarrhea (7 or more increase over baseline
in 24 hours), treatment with ipilimumab should be permanently discontinued and intravenous steroids and replenishment of fluid and electrolytes intravenously should be
instituted. Intravenous methylprednisolone 125 mg should
be given. Oral dexamethasone 4 mg every four hours or
prednisone 1 to 2 mg/kg/daily can be given thereafter,
followed by a taper and discontinuation over the next 6
weeks. Generally, there is a significant improvement of GI
symptoms within 1 to 2 weeks. Steroid therapy must be
tapered over at least 4 weeks to ensure complete resolution
of symptoms. In patients with diffuse and severe ulceration
and/or bleeding, a tapering of up to 6 to 8 weeks is appropriate since rapid tapering can result in recurrence or
worsening of symptoms. The use of opiates and other analgesics may mask pain associated with colitis induced by
ipilimumab. A low threshold for concern about GI complications, including perforation and obstruction, should be maintained. A low threshold for imaging with plain films or
computed tomography (CT), as well as a surgical consult
especially in any patients admitted to the hospital for GI
irAE managementis appropriate.
KEY POINTS
Immune-related adverse events (irAEs) represent a

unique spectrum of toxicities with checkpoint protein
inhibition.
Colitis, hepatitis, endocrinopathies, and dermatologic
toxicities are seen with ipilimumab and the key to
proper management of these irAEs is good patientcaretaker communication.
irAEs are generally managed successfully with steroids and other immune suppressants.
irAEs are also observed with the anti-PD-1 antibody
BMS-936558.
If intravenous steroids followed by high-dose oral steroids

does not decrease symptoms within 48 to 72 hours, treatment with infliximab at 5 mg/kg every 2 weeks is an
alternative.5,15 Once relief of symptoms is achieved, which
can be very rapid and dramatic, it should be discontinued
and a prolonged steroid taper over 45 to 60 days should be
instituted. There may be a waxing and waning of the GI
toxicities of ipilimumab during the course of steroids. As
steroids are tapered, symptoms may worsen, mandating a
retapering of steroids starting at a higher dose of 80 or 100
mg, a more prolonged taper, and additional use of infliximab. Ipilimumab should be permanently discontinued for
grade 3 or 4 diarrhea or colitis.
Dermatologic Toxicity
A diffuse, erythematous maculopapular rash that can be

intensely pruritic was observed in 47% to 68% of patients,
starting an average of 3 to 4 weeks after ipilimumab.1-3,7-9
In 4% of patients, it was severe. Ipilimumab-induced rashes
can be quite focal or even patchy. Microscopic examination
shows a perivascular lymphocytic infiltrate that extends
deep into the dermis in most cases.4 Immunohistochemical
staining showed that CD4-positive and Melan-A-specific
CD8-positive T cells were in close proximity to apoptotic
melanocytes, suggesting that an immune response was
directed against melanocytes.4 This is consistent with a
reported 11% rate of vitiligo with ipilimumab.7 Most skin
eruptions and pruritus associated with ipilimumab are generally managed symptomatically and usually do not require
skipping a dose or discontinuation. Topical glucocorticosteroids (e.g., betamethasone 0.1% cream) or urea-containing
creams in combination with oral antipruritics (e.g., diphenhydramine HCl or hydroxyzine HCl) are recommended.3 For
grade 3 dermatologic irAEs, one should hold a dose and treat
with a 3 to 4-week tapering course of oral steroids, starting
at 1 mg/kg prednisone or dexamethasone 4 mg four times
orally daily. Ipilimumab can be held for moderate to severe
skin toxicity but should be permanently discontinued for
severe, life-threatening skin toxicity and steroids initiated
at 1 to 2 mg/kg prednisone orally or its equivalent tapering
over not less than 30 days. Rare cases of toxic epidermal
necrolysis, as well as Stevens-Johnson syndrome, both in less
than 1% of patients, have been reported with ipilimumab,
and several patients with those conditions have died.
Endocrinopathies
Hypophysitis is an uncommon complication of treatment

with ipilimumab with an incidence of approximately
1.5%.16,17 Behavioral change, fatigue, severe headaches,
blurring or diplopia, myalgias, loss of appetite, or nausea
and vomiting have been seen, but symptoms can be vague.
Visual changes and/or headaches in patients with metastatic melanoma should also prompt concern for possible
central nervous system or ocular metastases and should be
evaluated by magnetic resonance imaging (MRI) of the brain
with gadolinium. Hypophysitis can present as a diffuse,
heterogenous enlargement of the pituitary on a brain MRI
but can be completely normal.18 When hypophysitis with
pituitary dysfunction is suspected, blood tests including
thyroid-stimulating hormone (TSH), free T4, adrenocorticotropic stimulating hormone, cortisol, leutinizing hormone,
and follicle-stimulating hormone should be obtained in
women, and the first four plus testosterone in men. Typically
175
JEFFREY S. WEBER
the anterior pituitary axis is involved, affecting thyroid,

gonadal, and adrenal function, but isolated axis dysfunction
can be seen. Hypophysitis will cause low free T4 as well as
TSH, but with peripheral thyroiditis leading to hypothyroidism, the TSH will be elevated even though T3 and free T4
will be low.
Hypophysitis with clinically significant adrenal insufficiency and hypotension, dehydration, and electrolyte abnormalitiessuch as hyponatremia and hyperkalemia
constitutes adrenal crisis. Hospitalization and intravenous
steroids with mineralocorticoid activity, such as methylprednisolone, should be initiated while waiting for laboratory results. Infection and sepsis should be ruled out with
appropriate cultures and imaging. Prednisone 1 mg/kg by
mouth should be administered if patients are clinically
stable. Steroids can usually be tapered over 30 days to
achieve physiologic replacement levels. Thyroid hormone
and/or testosterone replacement therapy may not be permanent, as the need for those hormones may wane over months
in some patients. Cortisone replacement may also not be
permanent in a modest portion of patients. With grade 2
endocrinopathies, ipilimumab therapy can be safely continued, in our experience, but proper informed consent should
be obtained from the patient before continuing treatment
since continued deterioration of endocrine function may
occur. Grade 3 to 4 endocrinopathies require discontinuation
of ipilimumab.
Isolated thyroid dysfunction can also be seen with ipilimumab, initially presenting as autoimmune thyroiditis
with elevation of free T4, evolving to a state of burnt-out
hypopthyroidism. Current recommendations in the ipilimumab package insert include checking baseline TSH and
free T4 and subsequent monitoring every 3 weeks during
induction ipilimumab treatment and every 3 months for 6
months following completion of therapy. Isolated adrenal
insufficiency unrelated to hypophysitis can also be seen.19
Hepatitis
Hepatotoxicity has been observed in 3% to 9% of patients

receiving ipilimumab.1,2,6-9 It usually presents as an asymptomatic increase of transaminases and bilirubin, although
some patients also have fevers and malaise. Liver biopsies
have shown a diffuse T-cell infiltrate consistent with
immune-related hepatitis. This must be differentiated from
progressive metastases in the liver, as well as other etiologies such as viral hepatitis or another drug-specific toxic
reaction. One should perform a standard workup to rule out
viral hepatitis, disease progression, and other drug-related
causes for abnormal liver functions. The current algorithm
for the management of a hepatotoxicity irAE contains the
recommendation that for grades 3 to 5 toxicity, one should
use high-dose intravenous glucocorticosteroids for 24 to 48
hours, followed by an oral steroid taper with dexamethasone
in a dosage of 4 mg every 4 hours or prednisone at 1 to 2
mg/kg tapered over not less than 30 days. Liver function
blood tests, if elevated beyond eight times normal, should
be checked every other day until they begin to drop, then
weekly until they are normal and mandate stopping ipilimumab. If serum transaminase levels do not decrease 48
hours after initiation of systemic steroids, consideration
should be given to the use of oral mycophenolate mofetil 500
mg every 12 hours. Infliximab because of its potential for
hepatotoxicityshould be avoided in this setting. One re-
176
port describes a patient treated with antithymocyte globulin

after developing severe hepatotoxicity from ipilimumab despite high-dose steroids and mycophenylate mofetil.20 A
waxing and waning picture may be seen with anti-CTLA-4
induced hepatitis, and several courses of tapering steroids
may be required. A very significant rate of hepatotoxicity
was seen when ipilimumab was combined with dacarbazine
and may be related to hepatic toxicity associated with
dacarbazine that is exacerbated by ipilimumab.2 Ipilimumab should be permanently discontinued for grade 3 to 4
hepatotoxicity.
Current guidelines recommend evaluation of serum markers of hepatic function at baseline and before each dose, as
well as periodically (every three months) after completion of
induction ipilimumab.
The Role of Surgery in Managing irAEs
With proper management of colitis, the likelihood that

obstruction or perforation of the bowel will occur is low, but
rarely colitis cannot be controlled by conservative measures
such as high-dose intravenous steroids, hydration, or repeated injections of infliximab.21 If colitis is resistant to
those measures, patients must be made nil per os (NPO)
except for required oral medicines, have a central venous
catheter inserted, and have complete bowel rest with total
parenteral nutrition (TPN) often for a period of 2 to 3
months.3 Rarely diarrhea and symptoms of colitis may
persist despite a trial of NPO and TPN while steroids are
unable to be tapered. In that case, if a colonoscopy confirms
recurring inflammation, then surgery to perform an ileostomy and put the colon at complete bowel rest is a reasonable option. In the experience of the author, a 3 to 4 month
period suffices to allow inflammation to resolve, the colonoscopic picture to improve, and diarrhea to disappear. At that
time, the ileostomy may be surgically reversed with some
assurance that the colitis has completely resolved after
CT scanning of the abdomen and repeat colonoscopy. A
partial or even complete colectomy may be necessary in very
rare cases of ipilimumab-induced colitis refractory to all
measures. Bowel perforation is rare occurring in less
than 1% of patients but must be considered.13 The risk of
death with ipilimumab is low, at 1% or less, often related to
complications of GI irAEs.1,2 It should be noted that patients
previously treated with high-dose interleukin-2 appear to
have an increased risk of intestinal perforation with ipilimumab, based on a small experience at one institution.
irAEs Observed with PD-1 Antibodies
In phase I and ongoing phase II trials, the spectrum of side

effects of PD-1 antibody BMS-936558 has been shown to
be qualitatively similar to ipilimumab but with fewer doselimiting immune-related side effects. In an initial singledose, phase I trial, no dose-limiting toxicity up to 10 mg/kg
was observed, and the side effects ranged from lymphopenia
to myalgias to fatigue of grades 1 and 2.22 One patient
developed grade 3 inflammatory colitis after five doses,
which responded to steroids and infliximab. One patient
developed grade 2 hypothyroidism requiring hormone replacement. Two patients developed grade 2 polyarticular
arthropathies requiring oral steroids; all three of the latter
could be considered irAEs. In an ongoing phase I/II trial of
every two weekly BMS-936558 with a peptide vaccine, no
maximum tolerated dose was reached up to 10 mg/kg and
CHECKPOINT PROTEIN INHIBITOR TOXICITY MANAGEMENT
hypothyroidism of grade 2 was observed; one patient developed grade 3 colitis, and one patient had dose-limiting optic
neuritis, both of which were felt to be irAEs.
Another PD-1 antibody, CT-011, has been tested in patients with hematologic malignancies at doses up to 6 mg/kg
with dose-limiting side effects.23 Only diarrhea was observed in more than one patient, along with drug-induced
weakness, rash, and flushing of grades 1 and 2.
Are irAEs Associated with Clinical Benefit?
A number of investigators have suggested that the occurrence of grade 3 to 4 irAEs is associated with overall
responses to ipilimumab. In patients with melanoma and
renal cancer who received ipilimumab at doses ranging from

3 to 9 mg/kg every 3 weeks, there was a clear association
between development of irAEs, chiefly GI- and endocrinerelated, as well as response to ipilimumab.24 In several
small adjuvant trials with ipilimumab, dose-limiting grade
2, 3, and 4 irAEs were both dose-related and associated with
prolonged relapse-free survival.25 The results of phase II
studies in metastatic melanoma have been mixed, with
several trials showing a modest association of dose-limiting
irAEs and overall response rate, but others contradicting
that observation. Interestingly, the use of steroids to treat
irAEs has not been shown to impact on the benefit of
ipilimumab.8,9
Author
Jeffrey S. Weber
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Altor
BioScience;
Celldex; Genesis
Biopharma
Altor
BioScience;
Celldex; Genesis
Biopharma
Honoraria
Bristol-Myers
Squibb; Celldex;
Genentech;
Merck;
Prometheus
Research
Funding
Expert
Testimony
Other
Remuneration
Novartis
REFERENCES
1. Hodi FS, ODay SJ, McDermott DF, et al. Improved survival with
ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010;363:
711-723.
2. Robert C, Thomas L, Bondarenko I, et al. Ipilimumab plus dacarbazine
for previously untreated metastatic melanoma. N Engl J Med. 2011;364:25172526.
3. Weber J. Ipilimumab: Controversies in its development, utility, and
autoimmune adverse events. Cancer Immunol Immunother. 2009;58:823-830.
4. Hodi FS, Mihm MC, Soiffer RJ, et al. Biologic activity of cytotoxic T
lymphocyte-associated antigen 4 antibody blockade in previously vaccinated
metastatic melanoma and ovarian carcinoma patients. Proc Natl Acad Sci
U S A. 2003;100:4712-4717.
5. Johnston RL, Lutzky J, Chodhry A, et al. Cytotoxic T-lymphocyteassociated antigen 4 antibody-induced colitis and its management with
infliximab. Dig Dis Sci. 2009;54:2538-2540.
6. Lebbe C, ODay S, Chiarion Sileni V, et al. Analysis of the onset and
resolution of immune-related adverse events during treatment with ipilimumab in patients with metastatic melanoma. Presented at Perspectives in
Melanoma XII, The Hague, Netherlands, October 2-4, 2008.
7. Wolchok JD, Neyns B, Linette G, et al. Ipilimumab monotherapy in
patients with pretreated advanced melanoma: A randomised, double-blind,
multicentre, phase 2, dose-ranging study. Lancet Oncol. 2010;11:155-164.
8. Phan GQ, Yang JC, Sherry RM, et al. Cancer regression and autoimmunity induced by cytotoxic T lymphocyte-associated antigen 4 blockade in
patients with metastatic melanoma. Proc Natl Acad Sci U S A. 2003;100:
8372-8377.
9. Attia P, Phan GQ, Maker AV, et al. Autoimmunity correlates with tumor
regression in patients with metastatic melanoma treated with anticytotoxic
T-lymphocyte antigen-4. J Clin Oncol. 2005;23:6043-6053.
10. Beck KE, Blansfield JA, Tran KQ, et al. Enterocolitis in patients with
cancer after antibody blockade of cytotoxic T-lymphocyte-associated antigen
4. J Clin Oncol. 2006;24:2283-2289.
11. Weber J, Thompson JA, Hamid O, et al. A randomized, double-blind,
placebo-controlled, phase II study comparing the tolerability and efficacy of
ipilimumab administered with or without prophylactic budesonide in patients
with unresectable stage III or IV melanoma. Clin Cancer Res. 2009;15:55915598.
12. Oble DA, Mino-Kenudson M, Goldsmith J, et al. Alpha-CTLA-4 mAbassociated panenteritis: A histologic and immunohistochemical analysis.
Am J Surg Pathol. 2008;32:1130-1137.
13. Smith FO, Goff SL, Klapper JA, et al. Risk of bowel perforation in
patients receiving interleukin-2 after therapy with anti-CTLA 4 monoclonal
antibody. J Immunother. 2007;30:130.
14. Berman D, Parker SM, Siegel J, et al. Blockade of cytotoxic

T-lymphocyte antigen-4 by ipilimumab results in dysregulation of gastrointestinal immunity in patients with advanced melanoma. Cancer Immun.
2010;10:11.
15. ODay SJ, Maio M, Chiarion-Sileni V, et al. Efficacy and safety of
ipilimumab monotherapy in patients with pretreated advanced melanoma:
A multicenter single-arm phase II study. Ann Oncol. 2010;21:1712-1717.
16. Blansfield JA, Beck KE, Tran K, et al. Cytotoxic T-lymphocyteassociated antigen-4 blockage can induce autoimmune hypophysitis in patients with metastatic melanoma and renal cancer. J Immunother. 2005;28:
593-598.
17. Dillard T, Yedinak CG, Alumkal J, et al. Anti-CTLA-4 antibody therapy
associated autoimmune hypophysitis: Serious immune related adverse events
across a spectrum of cancer subtypes. Pituitary. 2010;13:29-38.
18. Carpenter KJ, Murtagh RD, Lilienfeld H, et al. Ipilimumab-induced
hypophysitis: MR imaging findings. AJNR Am J Neuroradiol. 2009;30:17511753.
19. Min L, Vaidya A, Becker C. Ipilimumab therapy for advanced melanoma is associated with secondary adrenal insufficiency: A case series.
Endocr Pract. Epub 2011 Dec 2.
20. Chmiel KD, Suan D, Liddle C, et al. Resolution of severe ipilimumabinduced hepatitis after antithymocyte globulin therapy. J Clin Oncol. 2011;
29:e237-e240.
21. Chin K, Ibrahim R, Berman D, et al. Treatment guidelines for the
management of immune-related adverse events in patients treated with
ipilimumab, an anti-CTLA4 therapy. Ann Oncol. 2008;19 (suppl 8):787.
22. Brahmer JR, Drake CG, Wollner I, et al. Phase I study of single-agent
anti-programmed death-1 (MDX-1106) in refractory solid tumors: Safety,
clinical activity, pharmacodynamics, and immunologic correlates. J Clin
Oncol. 2010;28:3167-3175.
23. Berger R, Rotem-Yehudar R, Slama G, et al. Phase I safety and
pharmacokinetic study of CT-011, a humanized antibody interacting with
PD-1, in patients with advanced hematologic malignancies. Clin Cancer Res.
2008;14:3044-3051.
24. Downey SG, Klapper JA, Smith FO, et al. Prognostic factors related to
clinical response in patients with metastatic melanoma treated by CTLassociated antigen-4 blockade. Clin Cancer Res. 2007;13:6681-6688.
25. Sarnaik AA, Yu B, Yu D, et al. Extended dose ipilimumab with a
peptide vaccine: Immune correlates associated with clinical benefit in patients with resected high-risk stage IIIc/IV melanoma. Clin Cancer Res.
2011;17:896-906.
177
TARGETING CRITICAL MOLECULAR ABERRATIONS

EARLY IN THE COURSE OF SOLID TUMORS:
IS IT ABOUT TIME?
CHAIR
Houston, TX
SPEAKERS
Hagop Kantarjian, MD
Houston, TX
San Francisco, CA
Treatment of Chronic Myelogenous Leukemia

as a Paradigm for Solid Tumors: How
Targeted Agents in Newly Diagnosed Disease
Transformed Outcomes
By Jason R. Westin, MD, Hagop Kantarjian, MD, and Razelle Kurzrock, MD
Overview: Although chronic myelogenous leukemia (CML) is

rare, with approximately 5000 new cases in the United States
annually, it may be the poster child for the future of oncology.
Imatinib mesylate, a selective Bcr-Abl tyrosine kinase inhibitor
(TKI), transformed the course of CML from a rapidly fatal
disease (median survival, 3 to 6 years) to a functionally
curable, indolent disease with an estimated median survival of
more than 25 years. This transformation can be attributed to
several key factors: the identification of a causal and actionable molecular aberrationBCR-ABL; the development of a
potent and selective Bcr-Abl TKIimatinib; and, importantly
the application of imatinib in the earliest phase of CML. In
contrast, imatinib, if used in CML blastic phase, improves
median survival to only about 1 year. Similar to CML blastic
ARGETED THERAPY in cancer started long before the

imatinib era in chronic myelogenous leukemia (CML).
Examples of targeted therapies include hormone therapy
(tamoxifen) in estrogen- and progesterone-receptor positive
breast cancers, monoclonal antibodies in leukemias and
lymphomas (rituximab, gemtuzumab ozogamycin), and others. However, the era of targeted therapy in cancer is
identified with imatinib as a result of several factors: 1) deciphering the Philadelphia chromosome (Ph)-associated
and BCR-ABL-associated molecular abnormalities in CML;
2) defining a causal association between the BCR-ABL
molecular events and the development of CML in animal
models, thus proving that Bcr-Abl is a legitimate target for
therapeutic interventions; 3) discovering a relatively nontoxic, orally available selective Bcr-Abl inhibitor, imatinib
mesylate; and 4) demonstrating the clinical efficacy of imatinib. After positive clinical trials were reported, media
outlets suggested this new magic bullet could represent A
New Hope For Cancer and signal the dawn of a new phase
in the war on cancer.1
Left untreated, CML will ineluctably progress from the
chronic phase to the accelerated phase, and eventually
transform to CML blastic phase.2 This evolution is analogous to the natural disease course believed to occur in most
cancers. In the time before imatinib, therapies for CML
improved blood counts but achieved complete cytogenetic
responses (0% Ph-positive [Ph] metaphases in the bone
marrow) in only 15% of patients and improved median
survival from 3 years with busulphan or hydroxyurea to only
57 years with interferon alpha-based therapies. In contrast, when treating the earliest phases of CML with a
therapyimatinibspecific for the critical genetic abnormality, the cumulative complete cytogenetic response rate
was greater than 80%, and the estimated 10-year survival
increased from less than 20% to 85% or more. Indeed, at
present, the estimated median survival of patients with
CML exceeds 25 years.3-7 Because the median age of patients with CML at the time of diagnosis is nearly 60 years,
it is now estimated that most patients with CML have a
normal life expectancy if treated appropriately with Bcr-Abl
phase, metastatic solid malignancies have undergone genetic

evolution, and their molecular aberrations are complex. As a
result, resistance is common and eradication is difficult. The
key to the dramatic improvement in the survival of patients
with CML involved using imatinib in newly diagnosed disease,
before blastic transformation. We hypothesize that metastatic
solid tumors are analogous to CML blastic phase, and that to
achieve improvements in solid tumor outcomes similar to
those seen in CML, application of targeted agents to newly
diagnosed disease may be required to prevent disease transformation (i.e., metastases). Targeting driver mutations at the
time of diagnosis may be critical to the goal of markedly
changing the outlook for patients with cancer.
tyrosine kinase inhibitors (TKIs). These patients are therefore functionally cured. Imatinib changed the course of
CML from a previously fatal disease (median survival, 3 to 6
years) to a functionally and molecularly curable disorder.
Patients diagnosed with chronic-phase CML today may be
reassured, with our current knowledge, that most can expect
to live decades and will likely die with, and not of, CML
provided they comply with TKI therapy, there is a reasonable observation for progression of the disease, and an
appropriate change is made to second- and third-generation
TKIs if disease resistance evolves.
Imatinib was approved by the Food and Drug Agency in
2001 for treatment of Ph CML. During the decade to
follow, a sea change occurred in oncology with the approval
of dozens of agents to be used for targetable cancerassociated molecular aberrations in both hematologic and
solid tumor malignancies. Unfortunately, in solid tumors,
these new targeted therapies have to date not achieved the
expectations set by imatinib in CML. Therefore, we believe a
review of the CML experience is required to potentially
apply the lessons learned to solid tumors.
Lessons from the CML Experience
In hindsight, the positive results of imatinib therapy in

CML seem now preordained. However, at the start of the
imatinib trials, a common prediction shared by some CML
experts was that 1) CML is a heterogeneous disease with
multiple molecular abnormalities with increasing prevalence in advanced CML; 2) although early trials appeared
favorable, CML cells will ultimately develop resistance and
eventual late disease transformation even with early Bcr-
From the Department of Leukemia; Department of Lymphoma and Myeloma; Department

of Investigational Cancer Therapeutics, University of Texas M. D. Anderson Cancer Center,
Houston, TX.
Address reprint requests to Razelle Kurzrock, MD, M. D. Anderson Cancer Center, 1400
Holcombe Blvd., Houston, TX 77030; email: rkurzro@mdanderson.org.
1092-9118/10/1-10
179
WESTIN, KANTARJIAN, AND KURZROCK

Table 1. Comparison of Response Rates with Matched Targeted Agents in CML and Solid Tumors
CML
CML
CML
CML
blastic phase
accelerated phase
relapsed chronic phase
Newly-diagnosed chronic phase
Complete Cytogenetic
Response Rate (%) with
Bcr-Abl TKIs*
References
0 to 30
17 to 24
1349
65 to 80
811, 17
12
3, 24
3,14,15
Melanoma: relapsed metastatic V600E BRAF mutation

Melanoma: untreated metastatic V600E BRAF mutation
Lung Cancer: relapsed, advanced, ALK-rearranged
Response Rate (%)

with vemurafenib
(BRAF inhibitor)
Partial
75
47#
PR rate with crizotinib
(ALK inhibitor)
56%
Complete
6
1#
CR rate with crizotinib
(ALK inhibitor)
1%
26
32
27
Abbreviations: CML, chronic myelogenous leukemia; TKI, tyrosine kinase inhibitor.

* Bcr-ABl TKIs include imatinib, dastatinib and nilotinib, #: Response rate reported with median follow up of 3.8 months, tumor regression rate was approximately
95%.
Abl suppression; and 3) the early enthusiasm regarding the

positive results with imatinib should be tempered because
historic experiences in oncology have often resulted in disappointing long-term results.
The long-term positive experience with imatinib in CML
has in fact been beyond our original expectations. Currently,
the annual mortality related to CML is reduced from the
historic rate of 10% to 20% to less than 1%. We also learned
multiple important lessons that were unanticipated based
on previous cancer treatment experience: 1) the importance
of early therapy with targeted agents to achieve optimal
outcome; 2) the importance of an optimal biologic dose of the
targeting agent, rather than a classical maximum-tolerated
dose; 3) the rate of transformation was low, and if it
occurred, it did so early rather than late; 4) the development
of mutations in the targeted kinase as a resistant mechanism to TKIs; and 5) rationally designed and more potent
KEY POINTS
180
Chronic myelogenous leukemia (CML) is the poster

child for the future of oncology, having been transformed from a disease that was rapidly fatal (median
survival of 3 to 6 years) to one that is functionally
cured (estimated survival of 25 years).
In CML blastic phase (advanced stage), responses to
imatinib are transient, development of resistance
common, and survival is still only 1 year.
Targeted therapies in solid tumors are typically evaluated in patients with advanced disease and achieve
responses similar to imatinib in CML blastic phase.
The effectiveness of targeted therapies in solid tumors may be underestimated by evaluation at a stage
analogous to CML blastic phase.
In order to realize the same response outcomes in
solid tumors as have occurred in CML, it may be
necessary to use the high successful strategy of treating the earliest stage of the disease with targeted
therapy.
TKIs can be developed to overcome some of the mechanisms

of CML resistance to imatinib.
Importance of Treating Newly Diagnosed CML
The use of targeted agents in the earliest phases of the

disease is likely the most important lesson from the CML
experience that could be directly applied to other solid
tumors. With imatinib and other TKIs (nilotinib, dasatinib),
therapeutic outcome is most strongly associated with the
phase of CML (ie, chronic versus blastic phase) during which
TKI therapy is introduced (Table 1). In patients with blastic
phase CML, TKI therapy produces complete cytogenetic
response rates of only 10% to 20%, with transient cytogenetic responses and a median survival that increased from a
median of 3 6 months to 9 18 months. Invariably, in most
patients with blastic-phase CML resistance develops rapidly
to TKI therapy and patients die from complications of their
disease.8-11 These poor responses to TKI are assumed to be
related to the multiple additional resistance mechanisms
(cytogenetic clonal evolution, mutations, and multiple parallel and additional molecular abnormalities causing CML
transformation).
Even delaying therapy moderately so that it is given in
late chronic phase or accelerated phase markedly compromises complete cytogenetic response rates. For instance,
patients with newly diagnosed chronic phase CML have a
complete cytogenetic response rate of over 80% with Bcr-Abl
TKI treatment.7 The cytogenetic response rates drop to
approximately 40% in patients with previously treated
chronic phase CML.3 By the time disease is in acceleratedphase CML, complete cytogenetic responses occur in approximately 20% of cases.12,13
In addition, deeper levels of (molecular) responses are
achieved at a greater frequency in early chronic-phase CML
compared with more advanced disease, attesting to the
increased efficacy of TKI therapies in early chronic-phase
CML. Achieving a major molecular response (defined as
BCR-ABL transcript levels less than 0.1%) is noted in 40% to
70% of patients with newly diagnosed CML, with higher
molecular response rates observed with the more potent
second-generation TKIs (eg, dasatinib and nilotinib) compared with imatinib.14,15 The major molecular response rate
rates were lower in accelerated-phase CML (20% to 30%),
TARGETED TREATMENT OF NEWLY DIAGNOSED DISEASE
and less than 10% in blastic-phase CML. Similarly, achievement of undetectable BCR-ABL levels (more than 4 4.5 log
reduction of CML burden) was observed in 20% to 50% of
patients with newly diagnosed CML treated with TKI therapy but was uncommon in advanced CML.
and acceptable toxicities and established these agents at the

optimal biologic dosages as new standards of care in myelodysplastic syndrome.19
Progression to Blastic Phase of CML Is Rare with
TKI Treatment and Occurs Early Rather than Late
Optimal Dose of Targeted Therapy
Historically, most anticancer agents have been tested at

their maximum-tolerated dose, typically defined as a dose
below the one that produced severe toxicities in less than
one third of the patients. The assumption that the
maximum-tolerated dose is optimal is based on observations
that the highest possible tolerable dose would result in the
greatest antitumor results. This assumption may not apply
to targeted therapies, as was observed with both TKIs in
CML and the response rates of targeted therapies in phase
I trials.16 The phase I study of imatinib noted rare complete
cytogenetic responses at doses below 300 mg orally daily and
frequent complete cytogenetic responses at doses ranging
from 400 to 800 mg per day (on average, there were higher
responses with 800 mg than 400 mg).17 Interestingly, no
further improvement in complete cytogenetic responses
were seen at doses of 1000 mg daily and above, although
these doses were reasonably tolerated by patients. The
phase II study proceeded with imatinib doses of 400 mg
orally daily based on the imatinib plasma levels that were
anticipated to produce maximum anti-CML activity. Higher
doses of imatinib (800 mg daily) were later observed to
induce higher rates of complete cytogenetic and major molecular responses, but the additional benefit (as measured by
survival) was controversial. The experience with nilotinib
indicated that despite the long half-life of the drug, gastrointestinal absorption plateaued at a dose of 400 mg, resulting in schedules of 400 mg orally twice daily, implemented into phase II studies in late chronic-phase CML.
The randomized studies of nilotinib compared with imatinib
in newly diagnosed CML (Evaluating Nilotinib Efficacy
and Safety in clinical Trials-Newly Diagnosed patients
[ENEST-nd] trial) also compared a lower dose schedule of
nilotinib, 300 mg twice daily, with the approved standard
dose of 400 mg twice daily, demonstrating the equal efficacy
of the lower dose schedule and the benefit of a lower cost
and toxicity.14 Nilotinib 300 mg orally twice daily is now the
standard dose schedule for newly diagnosed CML. The early
studies with dasatinib suggested twice-daily schedules
(based on the short half-life of the drug) and doses of 140 mg
daily to be optimal. Subsequent randomized trials of four
different schedules established the single-daily 100 mg dose
of dasatinib to be the optimal schedule (equal efficacy, lower
toxicity), which is now used as the standard of care in
CML.15 These studies suggest that treatment with optimal
biologic dosages, as opposed to maximum-tolerated doses, of
TKIs in CML is best.
This finding has subsequently been validated by the
experience with epigenetic therapy (decitabine, azacitidine)
in myelodysplastic syndrome. Following early studies of
these drugs in the 1980s at the maximum-tolerated clinical
dosages, the drugs were almost abandoned until the 1990s,
when the concept of epigenetic therapy was developed.18
Using optimal biologic dosages of the drugs, decitabine 50 to
100 mg per m2 per course (instead of 1000 to 2500 mg per m2
per course) and azacitidine 250 to 525 mg per m2 per course
(instead of higher dosages), resulted in high efficacy results
The expectation during the International Randomized

Study of Interferon and STI571 (IRIS) trial was that leukemic cells would develop resistance and transform into
blastic-phase CML either at a uniform annual rate or later
in the course of therapy. This expectation was based on
the assumption that resistance develops under a timeassociated selective pressure from TKIs. Surprisingly, the
opposite was in fact observed.3 The annual transformation
rate during imatinib therapy was low, but if transformation
occurred, it happened early (2% to 4% annually in the first 3
years, and less than 1% annually thereafter). These data are
now interpreted to suggest that some patients with newly
diagnosed chronic-phase CML may harbor a small subset(s)
of CML clones with a priori TKI resistance, silently displaying the molecular resistance mechanisms of transformation
that would not be influenced by imatinib therapy. In these
patients transformation will develop within the first 23
years of therapy. In contrast, most patients with newly
diagnosed chronic-phase CML have clones that do not harbor intrinsic transformation mechanisms at diagnosis. In
these patients, imatinib therapy will ultimately reduce the
transformation rate to less than 1% per year.
Mutations as a Mechanism of Resistance in CML
During imatinib therapy, clinical resistance will develop

at a rate of 2% to 4% annually. Mutations at the BCR-ABL
kinase domain were noted in 30% to 40% of patients with
resistance in chronic phase and in more than 50% of patients
with resistance in transformation.20 Of note, a portion of
these mutations would alone be insufficient to result in
resistance to imatinib therapy (suggesting that other mechanisms of resistance were operational). Other mutations had
intermediate to high resistance to imatinib but were sensitive to second-generation TKIs. A pan-TKI resistant mutation, T315I (also referred to as gatekeeper mutation), was
detected in 20% of mutations (5% to 10% of resistant cases),
was resistant to second-generation TKIs (eg, dasatinib,
nilotinib, bosutinib), but was sensitive to several thirdgeneration TKIs including ponatinib.21 This experience suggests that in CML, and possibly in solid tumors, resistance
to targeted therapy may develop through the selective pressure of the targeted agent, resulting in mutations at the
target site that prevent the targeted agent from exerting
its effect. This phenomenon has now been observed with
FLT3 inhibitors (development of FLT3 point mutations
associated with resistance to FLT3 inhibitors) and in patients with solid tumors treated with targeted agents (eg,
development of mutations in the epidermal growth factor
receptor [EGFR] associated with resistance to EGFR TKI
therapy).22,23
Rationally Designed More Potent TKIs Overcome CML
Resistance to Imatinib
The development of mutations at the Bcr-Abl kinase

domain sites in TKI-treated CML had been predicted in
vitro from selection pressure models that identified several
181
50%
50%
100%
Change in Tumor Size
Survival Outcome
% Surviving
???
Time
Fig 1. Left panel shows typical waterfall plot for successful targeted agent in metastatic solid tumors. Many patients show tumor regression
but few achieve complete remission.
Upper right hand panel shows typical survival curve for patients with metastatic solid tumors treated with an active targeted agent. Even with
improved survival, most patients die of their disease, with survival gains generally measured in weeks to months. Survival curves for patients
with CML blast phase remain similar, even in the imatinib era.
Lower right hand panel shows typical survival curve for patients with newly-diagnosed CML chronic phase treated with imatinib or other
Bcr-ABl TKIs. High rates of long-term survival are achieved.
mutations later confirmed in vivo. This discovery resulted in

the development of a novel generation of TKIs that are
highly effective in suppressing the mutated CML clones
and have demonstrated clinical efficacy in patients with
imatinib-resistant CML. Second-generation TKIs resulted
in complete cytogenetic response rates of 50% or more in
such patients; these responses were durable and reduced
the post-imatinib resistance mortality rate from 10% to 15%
annually to 5% or less.24,25 These results encouraged the
front-line use of these agents, which when compared with
imatinib, produced higher rates of complete cytogenetic
response, major molecular response, and complete molecular
response.14,15 More importantly, they were associated with
lower rates of transformation to the accelerated and blastic
phases of CML.14,15 Together, these data suggest that the
earlier use of more potent TKIs could further suppress the
inherent mechanisms of resistance of CML clones at diagnosis in at least some patients. Furthermore, the successful
development of potent T315I inhibitors demonstrated their
high efficacy among patients with T315I-mutated CML.
Such experiences should also be considered in the context of
research and solid tumors.
Why Are We Not Achieving Similar Results in
Solid Tumors?
The development of imatinib for the treatment of CML is

considered a paradigm for targeted therapy. Despite numerous promising drugs and herculean efforts by expert oncologists, to date the CML/imatinib success remains
unmatched. At least two possibly valid conclusions can be
reached: 1) the CML/imatinib experience is singular (i.e.,
CML is so genetically simple and homogeneous that it is not
a relevant comparison to solid tumors) or 2) the use of
targeted therapy in solid tumors has not yet followed the
highly successful strategy used by CML investigators.
Although targeted therapies in solid tumors have not yet
transformed disease outcomes to match a normal life span,
they certainly cannot be described as ineffective. Therapies
targeting the driving molecular abnormalities in solid tu-
182
mors, such as vemurafenib for V600E BRAF mutations in

melanoma and crizotinib in ALK-rearranged lung cancer,
have resulted in substantial tumor response rates in preliminary trials. In a landmark phase I trial, Flaherty and
colleagues26 demonstrated an overall response rate of 81%
(26 of 32; two complete responses and 24 partial responses)
with the BRAF inhibitor vemurafenib given at a dose of 960
mg twice daily in the dose-expansion cohort of heavily
pretreated patients with mutated V600E BRAF metastatic
melanoma. Perhaps even more impressive, the doseescalation phase demonstrated a 69% response rate (11 of
16; one complete response and 10 partial responses) with
vemurafenib at doses of 240 mg or more twice daily. Unfortunately, disease control was transient in six of the 11
patients with response, and disease progressed within 9
months after the initiation of therapy. In the expansion
phase of the phase I trial of the ALK inhibitor crizotinib, the
response rate for patients with advanced, pretreated ALKrearranged lung cancer was 57% (47 of 82; one complete
response and 46 partial responses).27 These therapies were
largely well tolerated, especially when compared with standard cytotoxic chemotherapy. Although exciting, the data
indicate that these targeted therapies do not commonly
result in complete responses, that cures are rare, and that
patients who have response will likely have relapse. At first
glance, these results may be perceived as disappointing
compared with the CML experience. However, on closer
observation, they have striking parallels. A portion of patients with advanced disease, either CML or solid tumors,
benefit from targeted therapy, but resistance inevitably
develops (Fig. 1).
Importance of Early Therapy
In CML, the timing of targeted therapy is critical: patients

with newly-diagnosed CML do much better than those
with advanced or transformed disease (Table 1). As in CML,
it is generally accepted that newly-diagnosed solid tumors
are more responsive to therapy than previously treated or
relapsed cancers. Because of genomic instability, faulty
DNA repair, or various selection pressures, advanced cancers accumulate multiple genetic and/or molecular derangements over time, conferring various phenotypic advantages
on each subpopulation.28,29 Metastatic tumors from breast,
lung, and gastrointestinal cancers have more genetic abnormalities than do their primary tumors, and often display
considerable heterogeneity compared with primary lesions,
including discordant clinically utilized biomarkers.30,31
A counter argument to the conventional wisdom that solid
tumors are genetically more complex and heterogenous than
CML is the efficacy of targeted agents (e.g., vemurafenib and
crizotinib) in selected patients with advanced disease. In the
phase III vemurafenib trial, the overwhelming majority of
previously untreated patients with advanced melanoma had
tumor regression, although 48% of the evaluable patients
met criteria for response.32
An example of response rates of targeted therapy in
advanced solid tumors degrading with increasing stage of
disease is lapatinib, an oral dual EGFR-erbB2 TKI. When
evaluated in patients with newly diagnosed metastatic
HER2-positive breast cancer, lapatinib achieved an response rate of 24%.33 Among patients with previously
treated, metastatic HER2-positive breast cancer, the response rate declined to 7%.34 These trials cannot be directly
compared but do suggest an improvement in response rates
with earlier use of targeted therapy, similar to the CML
experience.
Thus, although it appears likely that targeted agents
perform better in untreated patients than patients with
relapsed metastatic disease, we believe that this experience
cannot be compared with newly diagnosed CML, as untreated metastatic solid tumors are analogous to untreated
CML blastic phase. Interestingly, the results attained in
metastatic solid tumors (either untreated or previously
treated) are remarkably similar to those achieved in patients with CML blastic phase (either untreated or previously treated): 1) many patients have some regression; 2) a
partial or complete response is achieved in a small subset of
patients ; 3) resistance develops in most patients; and 4) the
increase in survival can be measured in months, not years.
Mutations as a Mechanism of Resistance in
Solid Tumors
Although other pathways may be relevant for disease

progression, much of the resistance to imatinib in CML is
due to BCR-ABL kinase domain mutations.35,36 Resistance
to targeted therapy of solid tumors is less well characterized,
but preliminary evidence suggests a similar phenomenon
may occur, i.e., mutation of the original critical target.37,38 It
is not clear if this dependence on the initial oncologic driver
is more pronounced early, or is constant throughout the
disease course. In CML blastic phase, the minimal and
transient efficacy of Bcr-Abl TKIs, including those able to
overcome many common mutations, would be an arguement
for a substantial change in disease biology away from initial
driver dependence with disease progression.
When a crizotinib-resistant model of ALK-rearranged
lung cancer was generated with serial exposure, the most
common mechanisms of resistance were amplification of
the EML4-ALK gene and a gatekeeper mutation within
the kinase domain.37 As with CML, this finding suggests
that the driving abnormality remains critical to the survival
of the cancer cells, even in disease resistant to targeted

therapy.
In contrast, early studies of vemurafenib resistance in
V600E mutant melanoma cell lines have not identified
mutations in BRAF.39 In lieu of mutating the kinase domain, these vemurafenib-resistant cells were able to activate the MAPK pathway in a BRAF-independent fashion,
but remained sensitive to targeting the pathway farther
downstream.
The use of EGFR inhibitors in EGFR-mutated lung cancer
provides corroborating evidence of both resistance mechanisms. Mutations in exon 19 and 21 of EGFR are known to
be associated with increased sensitivity to EGFR TKIs. In
previously untreated patients with advanced lung cancer
who have an EGFR mutation, treatment with an EGFR
inhibitor typically yields median progression-free survivals
of approximately 9 months.40 When resistance develops, the
most common mechanisms are the development of a secondary point mutation in EGFR (T790M), which reduces the
ability of TKIs to inhibit EGFR, or amplification of the MET
gene.38,40,41
Together, these preliminary data suggest that a portion of
solid tumors may gave resistance mechanisms to targeted
therapy similar to those in CML, even while there is continued dependence on the initial genetic driver. If so, a CMLlike approach of developing second-generation targeted
agents with activity in resistant populations with mutated,
yet functional target kinase, or in targeting the critical
pathway at multiple points, could demonstrate efficacy by
either resensitizing resistant disease or blocking development of resistance.
Conclusion
Cancer claims more than half a million lives annually and

is now the leading cause of death under age 85 in the United
States.42 Globally, an estimated 12.7 million new cancer
diagnoses and 7.6 million cancer deaths occurred in 2008,
and these numbers are expected to nearly double in the next
25 years.43,44 In order to make a substantial impact on this
global health crisis, we believe it is beneficial to re-examine
one of oncologys greatest success storiesimatinib in CML.
Like solid tumors, CML accumulates new genetic aberrations throughout its course and eventually reaches a stage
where therapy targeted against its driving aberration has
markedly reduced effectiveness. It is our contention that
CML blastic phase is the equivalent of metastatic disease in
solid tumors. Both CML blastic phase and metastatic solid
tumors are difficult to eradicate because of their molecular
complexity. Deciphering this complexity and determining
how to best treat it is perceived as a staggeringly complicated task. Indeed, although the survival of patients with
CML blastic phase has increased in the imatinib era, it
remains at about 1 year. Yet, outcomes in chronic phase
CML have been transformed, with overall survival from
diagnosis increasing from approximately 4 years to more
than 25 years. As seen in CML, the true leap forward in
outcomes occurred only after investigators used Bcr-Abl
TKIs in the earliest stage of the disease. Importantly,
delaying therapy even slightly, to later chronic phase and
certainly to accelerated phase CML, is associated with a
precipitous fall in response rates and much poorer outcomes.
It is therefore plausible, if the CML paradigm holds true,
that administering targeted therapies to patients with ad-
183
vanced solid tumors, even if they are untreated, will continue to result in only incremental improvements. Newly
diagnosed metastatic disease is already genetically complex
and likely analogous to newly diagnosed CML blastic phase,
and not to newly diagnosed chronic phase CML. Timing may
therefore be a crucial missing component in the optimal
application of targeted therapies to solid tumors. The les-
sons of CML suggest that once a targeted therapy has shown

some activity in advanced disease, it may have the potential
for a high rates of functional cures but only if given at the
time of diagnosis of early stage disease. The current strategy
of using targeted therapy for patients with metastatic solid
tumors may therefore underestimate, perhaps dramatically,
the ultimate effectiveness of active targeted agents.
Author
Jason R. Westin*
Hagop Kantarjian
Razelle Kurzrock
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Novartis
AstraZeneca;
Health
Advances;
Johnson &
Johnson; Merck;
SAIC-Frederick
Amgen;
AstraZeneca;
Enzon; Exelixis;
Genentech;
Johnson &
Johnson;
Maxygen;
Merck; Myriad;
Nereus; Pfizer;
Pharmion;
Roche; SAICFrederick
Research
Funding
Bristol-Myers
Squibb; Cyclacel;
Genzyme;
Novartis; Pfizer
Amgen;
AmpliMed;
AstraZeneca;
Callisto;
Centocor Ortho
Biotech;
CuraGen; Enzon;
Exelixis;
GlaxoSmithKline;
Globomax;
Merck; Myriad
Genetics;
Nereus;
Novartis; Pfizer;
Roche
Expert
Testimony
Other
Remuneration
Exelixis; Nova
Research Co.;
Pfizer; Roche
REFERENCES
1. Lemonick MD, Park A. New hope for cancer. TIME. May, 28, 2001.
2. Kurzrock R, Gutterman JU, Talpaz M. The molecular genetics of
Philadelphia chromosome-positive leukemias. N Engl J Med. 1988 Oct
13;319:990-998.
3. Deininger M, OBrien, SG, Guilhot F, et al. International randomized
study of interferon vs STI571 (IRIS) 8-year follow up: sustained survival and
low risk for progression or events in patients with newly diagnosed chronic
myeloid leukemia in chronic phase (CML-CP) treated with imatinib. Blood.
2009;114:22 (suppl; abstr 1126).
4. Druker BJ, Guilhot F, OBrien SG, et al. Five-year follow-up of patients
receiving imatinib for chronic myeloid leukemia. N Engl J Med. 2006;355:
2408-2417.
5. Kantarjian HM, Talpaz M, OBrien S, et al. Survival benefit with
imatinib mesylate versus interferon-alpha-based regimens in newly diagnosed chronic-phase chronic myelogenous leukemia. Blood. 2006;108:18351840.
6. Bjorkholm M, Ohm L, Eloranta S, et al. Success story of targeted therapy
in chronic myeloid leukemia: a population-based study of patients diagnosed
in Sweden from 1973 to 2008. J Clin Oncol. 2011;29:2514-2520.
7. Kantarjian H, OBrien S, Jabbour E. Improved survival in chronic
myeloid leukemia since the introduction of imatinib therapy: a singleinstitution historical experience. Blood. 2012;119:1981-1987.
8. Kantarjian HM, Cortes J, OBrien S, et al. Imatinib mesylate (STI571)
therapy for Philadelphia chromosome-positive chronic myelogenous leukemia
in blast phase. Blood. 2002;99:3547-3553.
9. Cortes J, Kim DW, Raffoux E, et al. Efficacy and safety of dasatinib in
imatinib-resistant or -intolerant patients with chronic myeloid leukemia in
blast phase. Leukemia. 2008;22:2176-2183.
10. Giles FJ, Kantarjian HM, le Coutre PD, et al. Nilotinib is effective in
imatinib-resistant or -intolerant patients with chronic myeloid leukemia in
blastic phase. Leukemia. Epub 2011 Dec 13.
11. Sawyers CL, Hochhaus A, Feldman E, et al. Imatinib induces hematologic and cytogenetic responses in patients with chronic myelogenous leukemia in myeloid blast crisis: results of a phase II study. Blood. 2002;99:35303539.
12. Talpaz M, Silver RT, Druker BJ, et al. Imatinib induces durable
hematologic and cytogenetic responses in patients with accelerated phase
184
chronic myeloid leukemia: results of a phase 2 study. Blood. 2002;99:19281937.

13. Guilhot F, Apperley J, Kim DW, et al. Dasatinib induces significant
hematologic and cytogenetic responses in patients with imatinib-resistant or
-intolerant chronic myeloid leukemia in accelerated phase. Blood. 2007;109:
4143-4150.
14. Kantarjian HM, Hochhaus A, Saglio G, et al. Nilotinib versus imatinib
for the treatment of patients with newly diagnosed chronic phase, Philadelphia chromosome-positive, chronic myeloid leukaemia: 24-month minimum
follow-up of the phase 3 randomised ENESTnd trial. Lancet Oncol. 2011;12:
841-851.
15. Kantarjian HM, Shah NP, Cortes JE, et al. Dasatinib or imatinib in
newly diagnosed chronic phase chronic myeloid leukemia: 2-year follow-up
from a randomized phase 3 trial (DASISION). Blood. Epub 2011 Dec 9.
16. Jain RK, Lee JJ, Hong D, et al. Phase I oncology studies: evidence that
in the era of targeted therapies patients on lower doses do not fare worse. Clin
Cancer Res. 2010;16:1289-1297.
17. Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safety of a specific
inhibitor of the BCR-ABL tyrosine kinase in chronic myeloid leukemia.
N Engl J Med. 2001;344:1031-1037.
18. Jabbour E, Issa JP, Garcia-Manero G, et al. Evolution of decitabine
development: accomplishments, ongoing investigations, and future strategies.
Cancer. 2008;112:2341-2351.
19. Fenaux P, Mufti GJ, Hellstrom-Lindberg E, et al. Efficacy of azacitidine
compared with that of conventional care regimens in the treatment of
higher-risk myelodysplastic syndromes: a randomised, open-label, phase III
study. Lancet Oncol. 2009;10:223-232.
20. Shah NP, Nicoll JM, Nagar B, et al. Multiple BCR-ABL kinase domain
mutations confer polyclonal resistance to the tyrosine kinase inhibitor imatinib (STI571) in chronic phase and blast crisis chronic myeloid leukemia.
Cancer Cell. 2002;2:117-125.
21. Santos FP, Quintas-Cardama A. New drugs for chronic myelogenous
leukemia. Curr Hematol Malig Rep. 2011;6:96-103.
22. Heidel F, Solem FK, Breitenbuecher F, et al. Clinical resistance to the
kinase inhibitor PKC412 in acute myeloid leukemia by mutation of Asn-676
in the FLT3 tyrosine kinase domain. Blood. 2006;107:293-300.
23. Wheeler DL, Dunn EF, Harari PM. Understanding resistance to EGFR

inhibitors-impact on future treatment strategies. Nat Rev Clin Oncol. 2010;
7:493-507.
24. Hochhaus A, Baccarani M, Deininger M, et al. Dasatinib induces
durable cytogenetic responses in patients with chronic myelogenous leukemia
in chronic phase with resistance or intolerance to imatinib. Leukemia.
2008;22:1200-1206.
25. Kantarjian HM, Giles FJ, Bhalla KN, et al. Nilotinib is effective in
patients with chronic myeloid leukemia in chronic phase after imatinib
resistance or intolerance: 24-month follow-up results. Blood. 2011;117:11411145.
BRAF in metastatic melanoma. N Engl J Med. 2010:363:809-819.
inhibition in non-small-cell lung cancer. N Engl J Med. 2010;363:1693-1703.
28. Bown NP. Chromosome studies of solid tumours. J Clin Pathol.
1992;45:556-560.
29. Nowell PC. Cytogenetics of tumor progression. Cancer. 1990: 65:21722177.
30. Barlogie B, Johnston DA, Smallwood L, et al. Prognostic implications of
ploidy and proliferative activity in human solid tumors. Cancer Genet Cytogenet. 1982;6:17-28.
31. Curigliano G, Bagnardi V, Viale G, et al. Should liver metastases of
breast cancer be biopsied to improve treatment choice? Ann Oncol. 2011;
22:2227-2233.
32. Chapman PB, Hauschild A, Robert C, et al. Improved survival with
vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med.
2011;364:2507-2516.
33. Gomez HL, Doval DC, Chavez MA, et al. Efficacy and safety of lapatinib
as first-line therapy for ErbB2-amplified locally advanced or metastatic
34. Blackwell KL, Burstein HJ, Storniolo AM, et al. Randomized study of
Lapatinib alone or in combination with trastuzumab in women with ErbB2-
positive, trastuzumab-refractory metastatic breast cancer. J Clin Oncol.

2010;28:1124-1130.
35. Gorre ME, Mohammed M, Ellwood K, et al. Clinical resistance to
STI-571 cancer therapy caused by BCR-ABL gene mutation or amplification.
Science. 2001;293:876-880.
Cancer Cell. 2002;2:117-125.
37. Katayama R, Khan TM, Benes C, et al. Therapeutic strategies to
overcome crizotinib resistance in non-small cell lung cancers harboring the
fusion oncogene EML4-ALK. Proc Natl Acad Sci U S A. 2011:108:7535-7540.
38. Pao W, Miller VA, Politi KA, et al. Acquired resistance of lung
adenocarcinomas to gefitinib or erlotinib is associated with a second mutation
in the EGFR kinase domain. PLoS Med. 2005;2:e73. doi:10.1371/journal.
pmed.0020073.
39. Villanueva J, Vultur A, Herlyn M. Resistance to BRAF inhibitors:
unraveling mechanisms and future treatment options. Cancer Res. 2011;71:
7137-7140.
40. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatinpaclitaxel in pulmonary adenocarcinoma. N Engl J Med. 2009:361:947-957.
41. Sequist LV, Martins RG, Spigel D, et al. First-line gefitinib in patients
with advanced non, small-cell lung cancer harboring somatic EGFR mutations. J Clin Oncol. 2008;26:2442-2449.
42. Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011: the impact
of eliminating socioeconomic and racial disparities on premature cancer
deaths.. CA Cancer J Clin. 2011: 61:212-236.
43. Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA Cancer
J Clin. 2011:61:69-90.
44. Ferlay J, Shin HR, Bray F,et al. GLOBOCAN 2008 v1.2, Cancer
incidence and mortality worldwide: IARC CancerBase No. 10 [Internet].
Lyon, France: International Agency for Research on Cancer; 2010. http://
globocan.iarc.fr. Accessed Dec. 15, 2011.
185
Targeting Molecular Aberrations in Breast

Cancer: Is It about Time?
By Laura Esserman, MD, MBA, Christopher Benz, MD, and Angela DeMichele, MD
Overview: Breast cancer is not one homogeneous disease; it

is a heterogeneous cancer with remarkable genomic variation
and variable risk for systemic spread, time to recurrence, and
response to treatment. Although it is increasingly clear that a
substantial proportion of newly diagnosed patients with
breast cancer carry only minimal risk of developing metastatic
disease, our inability to accurately prognosticate leads to
systemic overtreatment. The recent introduction of multigene
predictors for baseline risk assessment and treatment response in breast cancer subsets heralds the beginning of
REAST CANCER is the first solid tumor to demonstrate the effectiveness of targeted therapy. The discovery more than three decades ago that the steroid
hormone receptors (HRs), estrogen receptor (ER) and progesterone receptor (PR), are overexpressed in the majority of
human breast cancers led to the eventual understanding
that endocrine therapy only benefits women with either ERor PR-positive disease. Initially thought to provide clinical
benefit only to postmenopausal patients with breast cancer,
ER-targeted agents such as tamoxifen were later shown to
reduce disease burden, decrease recurrence rates, and improve disease-free survival in premenopausal patients,1
demonstrating that targeting the ER pathway is a biologically basednot an age-basedtherapeutic approach.
Another successful example of breast cancer receptor
targeting is the HER2/neu story. The discovery by Slamon
and colleagues 25 years ago that a substantial proportion
(approximately 20%) of breast cancers possess genomic amplification of the HER2/neu oncogene and overexpress its
membrane-bound protein receptor, and that this is associated with worse outcome,2 spurred development of HER2
receptortargeted therapies. The first of these anti-HER2
therapeutics was the humanized monoclonal antibody, trastuzumab, that was approved for the treatment of metastatic
HER2-positive breast cancer3 after 13 years, and later
approved for treatment in the adjuvant setting on the basis
of its significant improvement in disease-free survival and
overall survival in patients with HER2-positive disease.4
The total time from identification that the receptor was a
critical driver to drug approval in the adjuvant setting was
18 years.
The subsequent introduction of prognostic and predictive
breast cancer biomarkers and multigene assays has had
some impact on our ability to determine who is most at risk
for developing metastatic recurrence and who would benefit
most from interventions such as systemic adjuvant chemotherapy. Bernie Fisher put forward the hypothesis that, for
most women, breast cancer is a systemic disease.5 A series of
large randomized clinical trials and meta-analyses have
proven conclusively that adjuvant chemotherapy can reduce
future recurrence of metastatic disease and markedly improve patient survival, verifying Fishers hypothesis. The
same data, however, have provided us with the sober realization that only some women benefit from this aggressive
systemic therapy. Many women do not need such aggressive
therapy because their tumors have indolent growth charac-
186
personalized cancer care and the transition to precision

medicine. At the same time, rapid clinical adoption of these
predictors illustrates the voracious unmet need for better and
more finely tuned prognostic and predictive tools. Recent
advances in clinical trial designs have enabled the testing of
novel agents in combination with standard therapy in the
neoadjuvant setting, with a goal of identifying the specific
biomarker-drug pairs that should be advanced for confirmatory trials and accelerated approval on the basis of response
to therapy.
teristics and would not likely recur in their lifetime; for

others, their tumors are not innately sensitive to the empirically determined drug combinations typically used in adjuvant treatment.
With the advent of widespread mammographic screening,
some aggressive breast cancers are detected at an earlier
(and more curable) clinical stage, but more common is the
detection of indolent tumors that are less likely to present a
life-threatening problem.6,7 We once provided treatment
with chemotherapy and endocrine therapy for any woman
with a tumor larger than 1 cm;1 there are now multigene
predictors that have enabled a more tailored approach.8,9
Chemotherapy is actually more effective against highly
proliferative tumors and less effective against low-grade,
slowly proliferating tumors, and commercially approved
multigene predictors that commonly interrogate proliferation genes have recently been introduced into clinical practice to tailor individual therapy.10 Interestingly, currently
approved multigene predictors have been unable to prognosticate the outcome of women diagnosed with HR-negative
tumors, which are usually more proliferative than ER- or
PR-positive tumors, but whose risk of dissemination may be
more dependent on yet-to-be-understood host factors. This is
just one of several obvious areas of unmet clinical need, in
which challenges and opportunities exist to expand the use
of gene profiling for developing novel predictors, and to tailor
existing and emerging therapies for those who will receive
the greatest clinical impact.11
The translational research community has produced several validated gene signatures, and many more gene signature candidates that can potentially push the field forward
have been described. Such signatures may answer questions
about response to our current medicationsincluding powerful biostatistical approaches that combine and order several predictors, such as hormone therapy sensitivity and
chemotherapy sensitivity ultimately allowing us to iden-
From the Carol Franc Buck Breast Care Center, University of California, San Francisco;
Cancer and Developmental Therapeutics Program, Buck Institute for Research on Aging,
Novato, CA; Breast Cancer Program, Abramson Cancer Center, University of Pennsylvania,
Philadelphia, PA.
Address reprint requests to Laura Esserman, MD, MBA, University of California, San
Francisco, 1600 Divisadero St., 2nd Floor, Box 1710, San Francisco, CA 94115; email:
laura.esserman@ucfsmedctr.org.
1092-9118/10/1-10
MOLECULAR ABERRATIONS IN BREAST CANCER
tify patients who may be refractory to therapy and at

highest risk for metastatic progression.12 In addition, we
now have access to powerful assay platforms and informatics tools that can generate and integrate tens of thousands of
gene expression data from very small slices of archival
tumor. This, in turn, creates the opportunity to validate
promising gene signatures and integrate them into a userfriendly, personalized clinical breast cancer application that
provides prognostic and predictive tools to all patients with
all types of breast cancer.
There are several important questions that have the
potential to be addressed by using new and emerging candidate signatures. We discuss them in the following sections.
Evidence of Very-Low-Risk Tumors and Tools
for Prediction
Screening has clearly led to an increase in the number of

detected cancers.13 Many of these tumors are likely to be
indolent, and the term IDLE tumors has been suggested to
differentiate tumors that are life-threatening and those that
are not.13 As shown by Esserman and vant Veer, it is
possible to identify molecular evidence of such tumors by
defining an ultra-low threshold for the 70-gene profile that
could identify tumors that, in the absence of any adjuvant
therapy, would not result in distant progression. By using
two large European data sets, they showed that with the
advent of screening, the incidence of these IDLE tumors has
increased, and that in women with screen-detected cancers,
30% could fit the definition of IDLE tumors and potentially
not need additional therapy.14 Validation studies are currently in progress. A tool that successfully predicted indolent disease would have an enormous and beneficial impact
KEY POINTS
Breast cancer is made up of several different diseases

with remarkable variation in the risk and timing of
recurrence.
There are targeted treatments in breast cancer that
are established on the basis of inhibiting overexpressed estrogen (ER) or HER2 receptors, and these
treatment strategies are being further refined by
using multigene subclassifiers.
Proliferative lesions are most likely to benefit from
chemotherapy, and the genes turned on by activated
ER are emerging as predictors for sensitivity to
endocrine therapy. Additional signatures are being
developed to predict resistance to either or both
endocrine and chemotherapies.
Existing and emerging predictive and prognostic biomarkers are being integrated into clinical trials to
help focus treatments on patient subpopulations
most likely to benefit from systemic therapy.
New and adaptive clinical trial designs can improve
the efficiency of new drug evaluation, particularly in
the context of biomarkers, and have already begun to
define a new regulatory path that will accelerate the
ability of targeted drug and biomarker pairs to be
used in the adjuvant setting.
for women, reducing the use of toxic therapies in women who

do not need them.
Evidence that Timing of Risk Is Related to
Tumor Biology
There is now a body of literature supporting the notion

that the timing of recurrence is a critical feature of the
biology of HR-positive breast cancer. By using a data set
from the late 1970s from Guys Hospital in London, before
the advent of adjuvant therapy and screening, we were able
to newly characterize 349 cases using more modern scoring
of ER and PR (HR), HER2, and grade to develop new tools to
understand the best order of integrating predictors of outcome. The routinely used staging, histologic and immunohistochemical features of primary breast cancers, provide
important information about overall relapse risk and timing
of future metastatic recurrence. By using a risk-partitioning
algorithm, we can identify the order in which the standard
clinical information can separate out groups of patients by
the degree and timing of risk. By using a classification and
regression tree approach, which identifies the order of importance of the clinical variables, the first split in the data is
by node status (0 vs. any). For the node-negative patients,
the groups then split by HR-positive versus triple-negative
(both ER and PR negative) and HER2positive disease.
HR-positive disease was split by low grade versus others.
The low-grade tumors appear to have late but not early
recurrence risk.15 The risk-partitioning method identifies
the critical dependencies among variables and opportunities
where molecular information can clearly improve our ability
to fine-tune predictions of both early and late risk, as well as
the benefits of chemotherapy. Curated and analyzed published breast cancer data sets can serve as an important
resource to validate candidate predictors for use in Clinical
Laboratory Improvement Amendments (CLIA)-certified
laboratories.16-18 Although our analyses have shown that
tumor proliferative capacity is predictive of early breast
cancer metastasis but only for HR-positive breast cancer15-17we have also uncovered new predictors of late
metastatic recurrence by HR-positive breast cancer that
need additional validation with respect to routine adjuvant
endocrine therapy use.15
We have clearly established that the timing of metastatic
recurrence is a critical feature of tumor biology. For HRnegative tumors and HER2-positive tumors, recurrence risk
is early (within the first 5 years), but for HR-positive breast
cancer, the risk can extend for many years. Others have
demonstrated this as well. The conclusion that emerges is
that HR-positive tumors have a recurrence risk that can
span more than 20 years from the time of diagnosis. For
these patients, signatures that are based largely on the
expression of genes driving cell proliferation are predictive
only of early recurrence. Fortunately, investigators are now
focusing on finding predictors that can be used to identify
HR-positive breast tumors with higher likelihood of recurrence. Dr. Liu and colleagues have described a 91-gene
signature to look for early recurrence as well as late recurrence, between 5 and 20 years after diagnosis.19,20
Sensitivity to Endocrine Therapy
Prediction of HR-positive patients most likely to benefit

from endocrine therapy remains a major clinical challenge to
187
ESSERMAN, BENZ, AND DEMICHELE
date. With the advent of gene expression profiling, several

gene signatures of endocrine sensitivity, including the
HOXB13:IL17RB ratio21,22 and the Sensitivity to Endocrine
Therapy (SET index),23 have been identified. In addition,
the estrogen-regulated gene GREB1 has been shown to be a
critical regulator of hormone-dependent breast cancer
growth and may serve as a marker of endocrine therapy
sensitivity24 Bianchi and colleagues from Italy recently
presented evidence that a signature of four estrogen-related
genes and 12 mitotic kinase genes has the potential to
predict resistance/sensitivity to hormone therapy.20 The fact
that many signatures are emerging on the basis of the same
biologic principles, and can be generated from the tissue
sample at the time of diagnosis, suggests that these types of
signatures are likely to validate and find their way into
clinical trials and practice. By using the neoadjuvant approach, even more discriminatory signatures are likely to
emerge. These predictors will not only identify women at
risk, but we hope will also suggest targeted strategies, based
on timing of treatments as well as novel agent combinations,
that will substantially improve outcomes specifically for the
groups at risk.
Identifying Risk and Response to Therapy in
HR-Negative Tumors
For HR-negative tumors and HER2-positive tumors, the

risk of recurrence is early, largely in the first 5 years. In
these tumors, proliferation does not seem to be a driver, as
the majority of tumors are indeed proliferative. In this
population, immune-based signatures seem to have the most
predictive signatures, and they can predict risk within the
first 5 years.15 Importantly, there is evidence that in some
patients with node-negative, HR-negative disease, risk is
quite low as a result of high immune responsiveness. This
has been shown independently in several studies.25 We have
also recently identified immune- and cytokine-based gene
expression levels capable of identifying early-stage HRnegative/triple-negative breast cancer at the lowest risk for
developing early or late metastatic recurrence.18 Both of
these signatures have been translated to CLIA-certified
RNA-based assay platforms and preliminarily validated
with our retrospective collection of formalin-fixed paraffinembedded breast tumor derived RNA samples,18,25 and
have been shown to be able to predict HR-negative and
triple-negative metastatic outcome.15,18 There is clearly a
spectrum of risk that is modified by the expression of
immune-related genes, suggesting that there is an important component of immune responsiveness that drives outcome in HR-negative tumors. This conclusion is further
supported by the identification of the T-cell/macrophage
signature that confers chemoresistance, described in the
following section.
Immunity and Sensitivity to Chemotherapy: T-Cell and
Macrophage Signatures
Leukocyte infiltration into breast tumors has been observed in the context of both protumorigenic inflammation
and anticancer immunosurveillance. Thus, beyond just the
composition of the immune infiltrate, the immune context of
these leukocytes (their distribution, density, architecture,
and interactions) must be analyzed to understand their
prognostic significance. Although the mechanisms involved
in leukocyte infiltration into tumors are poorly character-
188
ized, a recent study found high endothelial venules (HEVs),

blood vessels normally found in lymphoid tissues specialized
in lymphocyte recruitment, in a variety of solid tumors.26
High densities of HEVs in breast cancers correlated with a
more pronounced infiltration by T cells and conferred a
lower risk of relapse and significantly longer metastasisfree, disease-free, and overall survival rates.26
T lymphocytesin particular CD8-positive T lymphocytesare a crucial component of cell-mediated immunity.
Several studies have examined the prognostic value of
tumor-infiltrating lymphocytes (TILs) and/or CD8-positive
TILs in breast cancer.27,28 Interestingly, high TILs are
associated with better responses to chemotherapy and improved disease-specific survival, particularly in HR-negative
tumors.
We, and others, have shown that high numbers of tumorassociated macrophages (TAMs) are associated with high
grade, HR-negative breast cancers and poor outcomes.29,30
Recently, we proposed a gene expression signature related to
cytotoxic T-cell (Tc) responses and major histocompatibility
complex class II expression that might serve as a surrogate
for an anticancer immune microenvironment in breast cancer.31 Others have proposed combinations of macrophage
and T-cell markers as markers of poor outcomes on the basis
of both mouse and human studies.32,33 These results illustrate the importance of understanding the immune context
of leukocytes within the tumor microenvironment, and the
opportunity we may have to improve outcomes by targeting
the immune microenvironment as part of our therapeutic
strategy.
Trials that Focus on Development of Drug-Biomarker
Pairs and Response to Treatment in the
Neoadjuvant Setting
Increasingly, it is clear that all drugs will not benefit all

patients with breast cancer. As new targeted drugs are being
tested, more and more trials are incorporating biopsies
before and after treatment to both determine predictors to
therapy response and also assess pharmacodynamics. Biopsy at the time of presentation of metastatic disease is
increasingly becoming the standard of care, given that 20%
to 35% of metastatic lesions are discordant from the primary
tumor on the basis of HR or HER2 status.34,35 However, this
is problematic on several fronts. First, it is challenging to
perform biopsies of metastatic lesions: There are associated
risks depending on the location of these lesions. Second, the
standard setting for testing new drugs is in patients with
metastatic disease, often when they have experienced failure with first- and second-line treatments. Novel therapies
may be less effective once tumors have progressed through
several treatment regimens, and serial biopsies performed
with subsequent lines of therapy compound the risk. Finally,
effective agents have resulted in longer control of metastatic
disease, but rarely cure. Those same agents, when administered at the time of primary diagnosis, have resulted in cure.
Trastuzumab is the prime example of this phenomenon.4,5
Determining appropriate end points in the metastatic setting that predict curative potential when administered at
the time of diagnosis has been challenging.
One approach to determining which biologically defined
tumors are most likely to benefit from which therapy is to
introduce new agents at the time of diagnosis. This approach
is feasible because neoadjuvant therapy is increasingly
Fig. 1. Recurrence-free survival (RFS) stratified by response to therapy, complete pathologic response (pCR) for the population of 172
patients, excluding all patients receiving neoajuvant trastuzumab overall and by hormone
receptor (HR)-positive/HER2-negative subset,
HR-negative/HER2-negative (triple-negative)
subset, and HER2-positive subset. All patients
received neoadjuvant chemotherapy with a
doxorubicin regimen followed by a taxane.
RFS for all patients (excluding those receiving
trastuzumab, 37 patients) is shown on the left
and on the right for the receptor subsets. The
solid line represents patients who achieved a
pCR and the dotted lines represent patients
who did not. HR-positive/HER2-negative tumors are shown in blue, triple negative tumors
are shown in red, and HER2-positive tumors
are shown in green. Reprinted with permission. American Society of Clinical Oncology.
All rights reserved.40
being used as a standard of care.36 The order of therapy

(surgery or systemic therapy first) does not affect the efficacy
of therapy.37,38 However, sequencing systemic therapy first
allows the measurement of response to therapy and the safe
capture of tissue specimens during the course of treatment.
The I-SPY TRIAL and others have demonstrated that response to therapy is predictive of ultimate outcome and that
those with either complete or near-complete response to
chemotherapy in fact have much better outcomes than those
who have large residual cancer burdens.39-41 Importantly,
this relationship between response to treatment and outcome is even stronger by tumor subset (Fig. 1)whether
looking by receptor subtype40 or by molecular subset.42
We now have the opportunity to test emerging new agents
in the neoadjuvant setting for women at high risk for
recurrence. To do this efficiently, we need clinical trial
designs that take the heterogeneity of disease into account,

a paradigm for identifying the most effective agents and
their associated predictive biomarkers and a more rapid
path to getting such drugs to market. The I-SPY 2 TRIAL
design is an adaptive clinical trial focused on women at the
time of primary diagnosis with tumors 2.5 cm or larger. The
first step of the process is to generate a molecular profile
from core biopsy of the primary tumor; those women found to
be at high risk for early recurrence, on the basis of a
high-risk classification by the 70-gene profile, are then
randomly assigned to either standard therapy or standard
therapy in combination with a novel agent. The philosophy
of the I-SPY 2 TRIAL is to accelerate the testing of promising new targeted agents by enabling their first phase II
assessment to be in the neoadjuvant setting. To accomplish
this, the I-SPY 2 TRIAL has established a precompetitive
189
ESSERMAN, BENZ, AND DEMICHELE
consortium to facilitate the testing of agents against a

backdrop of comprehensive tumor profiling, with the ability
to use the adaptive design process to enrich enrollment on
the basis of standard markers. The biomarker design of the
trial allows the evaluation of three levels of biomarkers. The
first are standard or investigational device exemption (IDE)
approved, which are integral markers used in the adaptive
randomization. These include hormone and HER2 receptors,
the U.S. Food and Drug Administration (FDA)approved
70-gene profile, other measures of HER2 overexpression
(fluorescent in situ hybridization and expression array) and
volume on magnetic resonance imaging. The second level
comprises qualifying biomarkers. These are biomarkers
that have promise in predicting response to a specific agent,
but require validation in a CLIA-approved testing environment. This includes cell line predictors that predict response
and can be read by expression array, or other specific
phosphoprotein that is thought to be a key target of a
particular agent. These markers, if they prove to be predictive of response, would potentially be used as integral
markers in future trials of the agent and the data generated
from their evaluation in the I-SPY 2 TRIAL can be used to
support an IDE application. The last layer is the exploratory
biomarkers. These include hypothesis-generating markers
from new platforms such as RNAseq. The platforms being
used to characterize tumors in the trial include Agilent

Technologies (Palo Alto, CA) and Affymetrix (Santa Clara,
CA) expression arrays, reverse phase protein arrays,
genome-wide association study (GWAS), and RNAseq.
The goal of the trial is to graduate agents and biomarker
pairs with a predicted likelihood of success in phase III
trials.7 Going forward, if an agent is shown to improve the
chance of obtaining a pathologic complete response and this
observation can be confirmed in a separate neoadjuvant
trial, the FDA is willing to consider awarding accelerated
approval on the basis of these data.43 This provides a
regulatory path forward for accelerating successful agents to
the neoadjuvant or adjuvant settings, in which they are
likely to provide the greatest benefit. Indeed, it is about time
that we design our trials around the biomarkers that characterize risk and response to treatment. The I-SPY 2 TRIAL
is one such effort directed at more rapidly and efficiently
finding better options for patients with breast cancer. Furthermore, this platform provides the opportunity to validate
several of the emerging markers described in the first
portion of this article, and to help us move faster toward the
time when we will be able to provide precision medicine
approaches to reduce the suffering and death associated
with a diagnosis of breast cancer.
Author
Laura Esserman
Christopher Benz*
Angela DeMichele*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Agendia
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Early Breast Cancer Trialists Collaborative Group (EBCTCG). Effects of
chemotherapy and hormonal therapy for early breast cancer on recurrence
and 15-year survival: an overview of the randomised trials. Lancet. 2005;365:
1687-1717.
2. Slamon DJ. Proto-oncogenes and human cancers. N Engl J Med. 1987;
317:955-957.
3. Slamon DJ, Leyland-Jones B, Shak S, et al. Use of chemotherapy plus a
monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med. 2001;344:783-792.
chemotherapy for operable HER2-positive breast cancer. N Engl J Med
2005;353:1673-1684.
5. Fisher B. Breast-cancer management: alternatives to radical mastectomy. N Engl J Med. 1979;301:326-328.
6. Esserman LJ, Shieh Y, Rutgers EJ, et al. Impact of mammographic
screening on the detection of good and poor prognosis breast cancers. Breast
Cancer Res Treat. 2011;130:725-734.
7. Barker AD, Sigman CC, Kelloff GJ, et al. I-SPY 2: an adaptive breast
cancer trial design in the setting of neoadjuvant chemotherapy. Clin Pharmacol Ther. 2009;86:97-100.
8. van de Vijver MJ, He YD, vant Veer LJ, et al. A gene-expression
signature as a predictor of survival in breast cancer. New Engl J Med.
2002;347:1999-2009.
9. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence of
tamoxifen-treated, node-negative breast cancer. New Engl J Med. 2004;351:
2817-2826.
10. Paik S, Tang G, Shak S, et al. Gene expression and benefit of
chemotherapy in women with node-negative, estrogen receptor-positive
11. vant Veer LJ, Bernards R. Enabling personalized cancer medicine
through analysis of gene-expression patterns. Nature. 2008;452:564-570.
190
12. Hatzis C, Pusztai L, Valero V, et al. A genomic predictor of response

and survival following taxane-anthracycline chemotherapy for invasive
breast cancer. JAMA. 2011;305:1873-1881.
13. Esserman L, Shieh Y, Thompson I. Rethinking screening for breast
cancer and prostate cancer. JAMA. 2009;302:1685-1692.
14. Kok M, Linn SC, Van Laar RK, et al. Comparison of gene expression
profiles predicting progression in breast cancer patients treated with tamoxifen. Breast Cancer Res Treat. 2009;113:275-283.
15. Esserman LJ, Moore DH, Tsing PJ, et al. Biologic markers determine
both the risk and the timing of recurrence in breast cancer. Breast Cancer Res
Treat. 2011;129:607-616.
16. Tutt A, Wang A, Rowland C, et al. Risk estimation of distant metastasis
in node-negative, estrogen receptor-positive breast cancer patients using an
RT-PCR based prognostic expression signature. BMC Cancer. 2008;8:339.
17. Yau C, Wang Y, Zhang Y, et al. Young age, increased tumor proliferation and FOXM1 expression predict early metastatic relapse only for
endocrine-dependent breast cancers. Breast Cancer Res Treat. 2011;126:803810.
18. Yau C, Esserman L, Moore DH, et al. A multigene predictor of
metastatic outcome in early stage hormone receptor-negative and triplenegative breast cancer. Breast Cancer Res. 2010;12:R85.
19. Liu MC, Dixon JM, Xuan JJ, et al. Molecular signaling distinguishes
early ER positive breast cancer recurrences despite adjuvant tamoxifen.
Cancer Res. 2011;71 (suppl; abstr S1-8).
20. Bianchini G, Pusztai L, Iwamoto T, et al. Molecular Tumor Characteristics Influence Adjuvant Endocrine Treatment Outcome. Cancer Res. 2011;71
(suppl; abstr S1-7).
21. Sgroi DC, Haber DA, Ryan PD, et al. RE: A two-gene expression ratio
predicts clinical outcome in breast cancer patients treated with tamoxifen.
Cancer Cell. 2004;6:445.
22. Ma XJ, Wang Z, Ryan PD, et al. A two-gene expression ratio predicts

clinical outcome in breast cancer patients treated with tamoxifen. Cancer
Cell. 2004;5:607-616.
23. Symmans WF, Hatzis C, Sotiriou C, et al. Genomic index of sensitivity
to endocrine therapy for breast cancer. J Clin Oncol. 2010;28:4111-4119.
24. Rae JM, Johnson MD, Scheys JO, et al. GREB 1 is a critical regulator
of hormone dependent breast cancer growth. Breast Cancer Res Treat.
2005;92:141-149.
25. Teschendorff AE, Caldas C. A robust classifier of high predictive value
to identify good prognosis patients in ER-negative breast cancer. Breast
Cancer Res. 2008; 10:R73.
26. Martinet L, Garrido I, Filleron T, et al. Human solid tumors contain
high endothelial venules: association with T- and B-lymphocyte infiltration
and favorable prognosis in breast cancer. Cancer Res. 2011;71:5678-5687.
27. Denkert C, Loibl S, Noske A, et al. Tumor-associated lymphocytes as an
independent predictor of response to neoadjuvant chemotherapy in breast
28. Mahmoud SM, Lee AH, Paish EC, et al. The prognostic significance of
B lymphocytes in invasive carcinoma of the breast. Breast Cancer Res Treat.
Epub 2011 Jun 14.
29. Mukhtar RA, Moore AP, Nseyo O, et al. Elevated PCNA tumorassociated macrophages in breast cancer are associated with early recurrence
and non-Caucasian ethnicity. Breast Cancer Res Treat. 2011;130:635-644.
30. Mahmoud SM, Lee AH, Paish EC, et al. Tumour-infiltrating macrophages and clinical outcome in breast cancer. J Clin Pathol. 2012;65:159-163.
31. Mukhtar R, Wolf D, Tandon V, et al. PCNA tumor associated
macrophages are associated with M1 and M2 gene expression, and confer poor
prognosis in the absence of anti-tumor immune environment. Cancer Res.
2011;71 (suppl; abstr P4-09-19).
32. DeNardo DG, Barreto JB, Andreu P, et al. CD4() T cells regulate
pulmonary metastasis of mammary carcinomas by enhancing protumor
properties of macrophages. Cancer Cell. 2009;16:91-102.
33. DeNardo DG, Brennan DJ, Rexhepaj E, et al. Leukocyte complexity in
breast cancer predicts overall survival and functionally regulates response to
chemotherapy. Cancer Discov. 2011;1:54-67.
34. Wilking U, Karlsson E, Skoog L, et al. HER2 status in a populationderived breast cancer cohort: discordances during tumor progression. Breast
35. Amir E, Miller N, Geddie W, et al. Prospective study evaluating the
impact of tissue confirmation of metastatic disease in patients with breast
36. Buchholz TA, Lehman CD, Harris JR, et al. Statement of the science
concerning locoregional treatments after preoperative chemotherapy for
breast cancer: a National Cancer Institute conference. J Clin Oncol 2008;26:
791-797.
37. Bear HD, Anderson S, Smith RE, et al. Sequential preoperative or
postoperative docetaxel added to preoperative doxorubicin plus cyclophosphamide for operable breast cancer: National Surgical Adjuvant Breast and
Bowel Project Protocol B-27. J Clin Oncol. 2006;24:2019-2027.
38. Fisher B, Brown A, Mamounas E, et al. Effect of preoperative chemotherapy on local-regional disease in women with operable breast cancer:
findings from National Surgical Adjuvant Breast and Bowel Project B-18.
J Clin Oncol. 1997;15:2483-2493.
39. Carey LA, Dees EC, Sawyer L, et al. The triple negative paradox:
primary tumor chemosensitivity of breast cancer subtypes. Clin Cancer Res.
2007;13:2329-2334.
40. Esserman LJ, Berry DA, DeMichele A, et al. Pathologic complete
response predicts recurrence-free survival more effectively by cancer subset:
results from the I-SPY 1 TRIAL (CALGB 150007/150012; ACRIN 6657).
J Clin Oncol. In Press.
41. Symmans WF, Peintinger F, Hatzis C, et al. Measurement of residual
breast cancer burden to predict survival after neoadjuvant chemotherapy.
J Clin Oncol, 2007;25:4414-4422.
42. Esserman LJ, Berry DA, Cheang MC, et al. Chemotherapy response
and recurrence-free survival in neoadjuvant breast cancer depends on biomarker profiles: Results from the I-SPY 1 TRIAL (CALGB 150007/150012;
ACRIN 6657). Breast Cancer Res Treat. Epub 2011 Dec 25.
43. Esserman LJ. Woodcock J. Accelerating identification and regulatory
approval of investigational cancer drugs. JAMA. 2011;306:2608-2609.
191
ETHICAL CHALLENGES OF HEALTH CARE REFORM

FOR ONCOLOGY PROVIDERS: A DEBATE
CHAIR
Beverly Moy, MD, MPH
Boston, MA
SPEAKERS
Jeffrey M. Peppercorn, MD, MPH
Durham, NC
Amy P. Abernethy, MD
Durham, NC
Core Elements of the Patient Protection and

Affordable Care Act and Their Relevance to
the Delivery of High-Quality Cancer Care
By Beverly Moy, MD, MPH, Amy P. Abernethy, MD,
and Jeffrey M. Peppercorn, MD, MPH
Overview: The Affordable Care Act (ACA) contains many

provisions that affect cancer care. The provisions of health
care reform aim to improve access to quality cancer care,
particularly among the most vulnerable Americans. However,
health care reform also offers many challenges and opportu-
N MARCH 2010, President Obama signed into law the

Patient Protection and Affordable Care Act (ACA).1,2 The
ACA contains provisions that have important implications
for cancer care.3 Its implications for patients are sweeping
and have the potential to expand access to care and improve
cancer care among vulnerable groups. The ACA, although
not perfect, aims to make cancer care more comprehensive,
affordable, and accessible. This article provides an overview
of the ACA in relation to oncology, discusses the ethical
challenges for the oncology provider, and summarizes issues
relevant to various stakeholders in the oncology community.
Overview of Cancer Care and the
Affordable Care Act
Insurance Reforms
In 2008, there were an estimated 46.3 million uninsured

Americans, equaling 15% of the United States population,
plus an additional 25 million underinsured Americans.4,5
According to the Congressional Budget Office, the ACA is
expected to expand health insurance coverage to 32 million
individuals by 2019.6
Medicaid
The ACA expands Medicaid to individuals with incomes

up to 133% of the federal poverty level (FPL), thereby adding
16 million to 20 million individuals to the Medicaid roster.
The ACA also standardizes Medicaid benefits by guaranteeing a minimum package of essential services. However,
oncology provider participation in Medicaid is in jeopardy,
particularly in states that are economically poorer as a
result of low reimbursement levels. Currently, some states
are reporting additional cuts in Medicaid payments, further
reducing the value of this coverage. Also of major concern is
that there is convincing evidence that adult patients with
cancer with Medicaid have similarly poor clinical outcomes
compared with those of uninsured patients.7-10 Therefore,
Medicaid expansion would not necessarily be expected to
improve cancer outcomes among vulnerable populations as
the system currently stands. Given the anticipated expansion in the Medicaid population, the quality of care under
the Medicaid program may be further compromised.
Health Insurance Exchanges
By 2014, health insurance exchanges must be established

in all states. These exchanges are designed to provide
individuals or small businesses the opportunity to shop for
health insurance.
e4
nities that affect every stakeholder in oncology. This article

summarizes the ACA provisions relevant to oncology, discusses the ethical implications for the oncology caregiver,
and describes the effects on specific oncology stakeholders.
Elimination of Coverage Barriers
The ACA prohibits insurers from denying coverage to

children with preexisting medical conditions and allows
young adults up to age 26 to remain on their parents health
plans. Starting in 2014, all insurers will have to accept
all applicants, irrespective of preexisting conditions such as
cancer, and renew coverage. The law prohibits canceling
coverage, eliminates the lifetime amount insurance will
pay for certain conditions, and restricts annual limits. Patients with cancer and survivors of cancer will not be denied
coverage on the basis of their preexisting diagnosis of
cancer.
Medicare Donut Hole
The ACA helps to close the so-called Medicare donut hole

so that seniors do not face a costly gap in prescription drug
coverage. This is especially relevant in cancer because many
modern anticancer therapies are now in oral form and can be
prohibitively expensive. Each eligible senior will receive a
one-time, tax-free $250 rebate check.
Pediatric Cancer
In aggregate, childhood cancer is the sixth most common

cancer in the United States. The ACA specifically mandates
that Medicaid-eligible children who voluntarily opt for hospice services will no longer be required to forego curative
services, such as active cancer treatment, to receive hospice
care coverage.
Cancer Prevention and Survivorship
Early detection of cancer through adherence to recommended screening examinations leads to decreased mortality from most cancers. The ACA requires all health plans to
cover preventive services that receive an A or B rating
from the U.S. Preventive Services Task Force (USPSTF).
These treatments must be covered with no deductibles or
copays and with no maximums allowed. However, the ACA
does not expressly require insurers to cover follow-up testing of abnormalities found in a cancer screening examination. The ACA also does not comment on reimbursement for
From the Massachusetts General Hospital Cancer Center, Boston, MA; and Duke
Comprehensive Cancer Center, Durham, NC.
Address reprint requests to Beverly Moy, MD, MPH, Massachusetts General Hospital,
55 Fruit St., YAW 9A, Boston, MA 02114; email: bmoy@partners.org.
1092-9118/10/1-10
HEALTH CARE REFORM AND ONCOLOGY
development and implementation of cancer survivorship

care plans.11
Cancer Clinical Trials
The ACA mandates coverage of routine costs for patients

who participate in cancer clinical trials. Insurers are prohibited from dropping or limiting coverage for participants in
cancer clinical trials.
Improving Quality and Lowering Costs
The ACA establishes a national pilot program to encourage hospitals, doctors, and other providers to work together
to improve the coordination and quality of patient care.
Under payment bundling, hospitals, doctors, and providers
are paid a flat rate for an episode of care such as a cancer
diagnosis, rather than the current fee-for-service system.
The entire oncology team is compensated with a bundled
payment that provides incentives to deliver health care
services more efficiently while maintaining or improving
quality of care. Bundled payments may be the most striking change resulting from the ACA that oncologists will
experience.
ACA Impact on Oncology
The stated goals of ACA are to expand access and control

health care costs. However, although the expansion of access
is relatively easy to quantify, the impact on health care costs
is less clear. On the surface, increasing demand for any good
or service typically results in higher costs in a free market
unless supply is similarly increased. Both expanded coverage through the ACA and the projected increase in cancer
cases as a result of demographic shifts in the U.S. population
(e.g., aging) are likely to increase demand for oncology goods
and services over time. In contrast, on the supply side, an
oncology physician workforce shortage is projected,12 novel
drugs with complex production requirements may be scarce
after initial approval,13 and shortages of commonly used
generic drugs are already widespread and projected to
increase over time.14 One of the key questions then is how,
in this context, the impetus to control costs will affect care in
the clinic.
To the extent that the ACA provides oncologists with more
resources to care for patients, the ACA may ease ethical
tensions. An oncologist faced with a patient in clinic with a
treatable disease or symptoms has an ethical obligation to
provide care or to help the patient obtain care elsewhere.
Oncology practices are currently faced with the dilemma of
how to care for patients without adequate resources.15 Few
would argue that such differences in access or outcomes on
the basis of ability to pay alone are morally acceptable.
However, there is legitimate debate regarding how society
KEY POINTS
The ACA contains provisions that are relevant to

oncology.
Health care reform poses ethical implications for
oncology providers.
Specific stakeholders in the oncology community will
face a variety of relevant issues.
can best address such disparities and the extent to which the
ACA will really help. At the center of the decades-old
controversy over health care reform in the United States
is the question of whether the goals of providing access
and improving health outcomes are best achieved through
a) government-provided health care, as is the dominant
model in the British National Health Service, b) governmentfinanced health care, as in Canada, c) increased government
regulation of health care and health insurance, as established by the ACA, or d) deregulation and greater freedom
within the medical care and medical insurance market
place, as proposed by some critics of the recent health care
law.
At the level of the individual provider, regardless of the
basis for insurance coverage, any increase in the amount of
resources available to care for an individual patient presenting with cancer should provide a greater opportunity to meet
our ethical obligation to provide care. There are several
other features of the new law that will enhance our opportunity to provide cancer care. As described herein, insurers
will be unable to refuse coverage to patients with preexisting
conditions and there will be no lifetime cap on coverage
critical issues for patients with cancer. There will be no cost
sharing for recommended cancer screenings such as mammography and cervical cancer screening. In addition, in a
provision that has received little attention, the law establishes a federal right to timely independent external appeal
of adverse coverage decisions. This could be particularly
important for patients with rare cancers or rare presentations of cancer requiring off-label therapy.
Ethical Challenges in Cost Control
The predominant ethical challenges facing oncologists will

likely come from increasing pressure on oncologists to act
both in their patients interest and simultaneously, in the
interest of society, the government, the insurer, or the
accountable care organization (ACO) in helping control
costs. These pressures are likely to be most acute in the
newly established ACOs, in which there are direct financial
incentives for controlling the cost of health care, but will
likely extend to consideration of care for all patients. Not
only is the entire success of this experiment in expanded
health care coverage dependent on financial sustainability,
but oncologists are likely to directly feel the impact of any
ongoing failure to shift the cost curve.
The law establishes a new Independent Advisory Board
charged with controlling Medicare spending. Given that
most potential remedies for rising costs are explicitly prohibited, such as rationing, cuts in services, cost sharing, or
changes in hospital reimbursement, it appears likely that
rising costs will be met with cuts in reimbursement to
providers. One of the issues likely to face providers is simply
how (and whether) to practice in an era of lowered reimbursement. Will we protect revenue at the expense of time
with patients or provision of ancillary services? Will this
affect individual treatment decisions? There may be very
personal answers to these questions for each oncologist.
The goal, of course, is to provide higher-quality care at
lower cost by focusing on the value of the care we provide.16
In one of few studies to demonstrate this potential, Neubauer and colleagues found that adherence to lung cancer
guidelines can reduce costs substantially with no detriment
in survival.17 The potentially high costs associated with
e5
MOY, ABERNETHY, AND PEPPERCORN
non evidence-based cancer care have also been documented.18 However, it is likely that in some settings we will
face real trade-offs between marginal benefits of an intervention and additional cost. For example, one year of trastuzumab for HER2-positive breast cancer reduces the risk of
recurrence by close to 50% compared with chemotherapy
alone, but at an additional cost of approximately $50,000 per
patient.19 Recent data in both the neoadjuvant and metastatic breast cancer settings suggest that additional interventions above and beyond trastuzumab, such as lapatinib
and pertuzumab, may further improve outcomes for some
patients.20,21 If additional benefit is confirmed in large
adjuvant randomized trials, one can imagine a scenario in
which we are forced to decide on further improving outcomes
versus doubling or tripling the cost of therapy. In fact,
throughout oncology, novel interventions are improving outcomes, but in many cases such progress comes at a substantial price.13,22
If, as is currently the case, interventions continue to be
approved on the basis of marginal benefits in efficacy, how
will oncologists decide which interventions truly bring
meaningful clinical benefit? What role will patient preference or shared decision making have in an era of increased
pressure to control costs? If nothing else, the recent decision
by the U.S. Food and Drug Administration (FDA) to withdraw approval for bevacizumab in breast cancer demonstrated lack of consensus on what constitutes value in
oncology among both clinicians and patients.
Many of these tensions exist within the current health
care system. They may be magnified if there is an assumption of universal access to cancer care that is not matched
by the actual reimbursement for services. Will patients with
different forms of insurance within the new system be
treated differently? Will treatment decisions vary for patients who are within or outside of an ACO?
In the setting of expanded access and pressure to control
costs, oncologists may face ethical challenges related both to
how they practice and whom they are willing to treat. With
more insured patients there may be greater demand for
oncology services, but as noted above, it is expected that
important differences in the adequacy of coverage for cancer
care will persist. Will the expanded cohort of patients with
Medicaid and continued low reimbursement levels paradoxically push a greater number of oncologists to close their
practices to Medicaid? Where low reimbursement for a small
percentage of patients might be subsidized within a large
practice, the potential economic impact of a larger share of
such patients might prove to be too great a risk in some
settings.
In addition, the ACO model of rewarding physicians and
practices for improving the quality, coordination, and cost of
care may create a disincentive to treat more complex patients who may require care outside of the expected norms.
On the basis of unusual disease presentation, tolerance of
standard therapies, or comorbidities, some patients may
require oncologists to either step outside of the standard
care plans and consider options that might be more expensive than the norm. As noted herein, this will even be
facilitated under the new law through a strengthened right
to appeal coverage decisions. There would seem to be a clear
incentive to both provide treatment for patients with less
complex disease, whose expected costs of care are likely to
e6
be at or below the average, and to constrain our decision

making for individual patients.
There are clearly some pitfalls ahead, primarily in the
area of cost containment, and no easy solutions. The question will be how to find the balance in our practices between
advocacy for our patients interests and preferences and
some effort to consider wise and efficient use of resources.
Failure to strike this balance may leave us with a tragedy
of the commons in which individual decisions to ignore
societal costs of care results in unsustainability of the entire
system, and threatens our ability to guarantee high-quality
cancer care to all patients.
The ACA and Oncology Stakeholders
As demonstrated in the preceding section, the ACA promises to change the landscape of health care for virtually
every stakeholder. Insurance reforms will inevitably lead
to changes that will dramatically affect the composition of
populations of patients with cancer around the country. A
focus on health service delivery innovation, comparative
effectiveness research (CER), and reducing health care disparities will shift the profile of cancer research.
Oncology Workforce
In addition to the ethical considerations and the projected

increased demand for oncology care outlined above, improvements in access to health care will affect the oncology
workforce. Support for new models of health care delivery
under the ACA will help meet escalating demand for cancer
care. The ACA established the Center for Medicare &
Medicaid Innovation (CMI) to test innovative payment and
service delivery models to reduce Medicare and Medicaid
expenditures while improving quality of care. Beginning
with its Health Care Innovation Challenge program in 2012,
CMI will provide funding to demonstrate rapidly deployable
new models for reducing total costs of care while improving
quality and health outcomes; proposed models will likely
address access and volume in the context of resource constraints.
ACA encourages models, such as the medical home and
ACO, which are designed to optimize care and largely focus
on community-based delivery through which the majority
of patients access cancer care. A common theme among
the various emerging delivery models is coordination of
care. Models for which ACA authorizes funding include
(1) community-based collaborative care networksconsortia
of health care providers that include a safety-net hospital,
have a joint governance structure, and provide comprehensive coordinated and integrated services to low-income
populations; (2) interdisciplinary, interprofessional teams of
health care providers who work with primary care providers
to provide integrated community-based care; and (3) community health centers, which serve an estimated one in
three low-income people and one in four low-income minority individuals. Provisions pertaining to medical homes
allow states, under Medicaid, to make medical assistance
payments at an enhanced federal match to teams of providers. Clearly, under the ACA, coordinated care and clinician
teamwork will increasingly become the norm.
Provider organizations are encouraged, through the ACA,
to expand and diversify their workforce. Through the Centers for Disease Control and Prevention (CDC), the law
HEALTH CARE REFORM AND ONCOLOGY
Under the ACA, payers such as health insurers will lose a

substantial degree of selectivity. As stated previously, they
will be required to accept all new applications for health
insurance and to renew individuals coverage, regardless of
preexisting conditions. They will no longer be allowed to
cancel coverage or to set a lifetime amount that a policy will
pay for certain conditions, and annual limits will be restricted. Health insurance exchanges will create an environment of open competition, although the extent to which
plans will differ remains to be seen.
a part of ARRA. Critical to the Office of the National

Coordinators strategy is to develop the information substrate needed to support a system of rapid learning health
care, a goal well aligned with the rapid learning cancer
care agenda of the oncology community.23 The HITECH Act
will affect the cancer community broadly, predominantly
through the requirement to transition to electronic medical
records and the use of data for quality monitoring, and will
affect the cancer research community quite specifically
through the provision of clinical data to support CER,
annotation of biospecimens, and clinical trials.
Aspects of the ACA will aid oncology researchers by
helping overcome barriers to clinical research. As noted
above, to remove an economic barrier and thereby facilitate
participation in clinical trials, ACA prohibits health insurers from denying coverage of routine costs associated with
clinical trial participation and from discriminating against
patients participating in clinical trials. Because rates of
racial and ethnic minority participation in cancer clinical
trials are disproportionately low,24 this provision may prove
particularly beneficial to the quality and generalizability of
cancer research. Health services research will be facilitated
by a requirement that all federally funded public health and
health care programs collect data on race, ethnicity, sex,
primary language, disability, and geographic data at the
lowest level (i.e., state, local, institutional) where they can
be aggregated; oncologists seeking to study disparities will
benefit from this provision.
Academic oncologists will see further evidence of federal
commitment to reducing disparities as the National Center
on Minority Health and Health Disparities (NCMHHD)
becomes a new institute within the National Institutes of
Health (NIH). Created by the ACA, this institute will hold
responsibility for planning and coordinating all NIHsupported health disparities research. The law also moves
the Office of Minority Health from the Public Health Service
to HHS, and establishes parallel offices in several HHS
agencies including the CDC, FDA, and Agency for Healthcare Research and Quality (AHRQ)further highlighting
health disparity research as a national priority while endeavoring to coordinate approaches to studying disparities.
Cancer Researchers
Conclusion
Following on the heels of the American Reinvestment and

Recovery Act (ARRA) of 2009, the ACA extends the impact of
health care reform on medical research. Academic oncologists will continue to see federal funds designated for CER,
which is characterized by direct comparison of existing
interventions, enrollment of real world populations (i.e.,
patients as they are typically seen in day-to-day clinical
care), and diversity of study design. The mantra of getting
the right care to the right patient at the right time guides
CER. Substantial ARRA funding has already supported
CER studies. Building on CER momentum ignited by ARRA,
the ACA established the Patient-Centered Outcomes Research Institute (PCORI); this nonprofit, nongovernmental
entity promotes CER by identifying research priorities,
establishing and maintaining a research agenda, and providing funding for research.
Further, the CER agenda will be facilitated by a national
transition to electronic medical records, data availability,
and data use, supported by the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009,
Health care reform has the potential to improve cancer

care in the United States by making it more accessible,
affordable, and comprehensive. The law contains many
provisions that may meaningfully change clinical outcomes,
especially for the most vulnerable Americans diagnosed
with cancer. The vast majority of patients with cancer will
now have insurance coverage and will no longer face the risk
of losing coverage or reaching a lifetime cap on benefits,
regardless of changes in employment or personal wealth.
However, the ACA may exacerbate existing practical and
ethical challenges in oncology practice as we attempt to
deliver the highest-quality cancer care to an expanded pool
of patients on a financially sustainable basis. This law, and
the opportunities and challenges that accompany it, affect
every stakeholder in cancer care, from patient to clinicians
to payers to researchers. Going forward, it is imperative that
the entire oncology community has a clear vision for maintaining and improving the quality of care within the new
framework for access, coordination, and cost control established under ACA.
authorizes support of community health workers at hospitals, federally qualified health centers, and other public or
nonprofit entities. The ACA also authorizes grants for various workforce diversity measures including recruitment of
individuals from under-represented, disadvantaged, or rural
backgrounds into health professions; community-based
training and education, with an emphasis on primary care
in underserved areas; and community-based field placements or preceptorships.3 The law encourages development
of curricula for cultural competency programs, and inclusion
of competency measures within quality measurement systems.3
Hospital and Private Practice Administration
The ACA will have substantial administrative impact on

providers (both organizations and individual practitioners).
Hospitals and physicians will be required to collect and
report quality data, with penalties imposed for not reporting
or not meeting quality standards. These requirements, although initially burdensome, may encourage provider participation in important quality initiatives such as ASCOs
national quality improvement program and clinical registry,
the Quality Oncology Practice Initiative (QOPI). The ACA
reflects a national trend toward linking health care financing to quality metrics. For example, under ACA, providers of
pediatric medical care who meet specified requirements can
be recognized by states as ACOs, and will receive incentive
payments based on metrics specified by the U.S. Department of Health and Human Services.
Insurers
e7
MOY, ABERNETHY, AND PEPPERCORN
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Beverly Moy
GlaxoSmithKline
(U); Pfizer (U)
Amy P. Abernethy
Bristol-Myers
Squibb; Helsinn
Jeffrey M. Peppercorn
GlaxoSmithKline (I)
Stock
Ownership
Honoraria
Amgen; Novartis
AVEO; Bayer;
Genentech
Research
Funding
Expert
Testimony
Other
Remuneration
Amgen; BioVex;
Bristol-Myers
Squibb; DARA;
Eisai; Helsinn;
KangLaiTe; Lilly;
Pfizer
Genentech;
Novartis
REFERENCES
1. Patient Protection and Affordable Care Act. Public Law 111-148. http://
www.gpo.gov/fdsys/pkg/PLAW-111publ148/html/PLAW-111publ148.htm. Accessed February 17, 2012.
2. Health Care and Education Reconciliation Act of 2010. Public Law
111-152. http://www.gpo.gov/fdsys/pkg/PLAW-111publ152/html/PLAW-111
publ152.htm. Accessed February 17, 2012.
3. Moy B, Polite BN, Halpern MT, et al. American Society of Clinical
Oncology policy statement: opportunities in the patient protection and affordable care act to reduce cancer care disparities. J Clin Oncol. 2011;29:38163824
4. DeNavas-Walt C, Proctor BD, Smith JC. US Census Bureau, Current
Population Reports. Income, Poverty, and Health Insurance Coverage in the
United States: 2008. Washington, DC: U.S. Government Printing Office,
Economics and Statistics Administration, 2009.
5. Schoen C, Collins SR, Kriss JL, et al. How many are underinsured?
trends among U.S. adults, 2003 and 2007. Health Aff (Milwood). 2008;27:
w298-w309.
6. Elemendorf DW. Letter to the Honorable Nancy Pelosi. http://www.
cbo.gov/ftpdocs/113xx/doc11379/AmendReconProp.pdf. Accessed February 16,
2011.
7. Ayanian JZ, Kohler BA, Abe T, et al. The relation between health
insurance coverage and clinical outcomes among women with breast cancer.
N Engl J Med. 1993;329:326-331.
8. Kelz RR, Gimotty PA, Polsky D, et al. Morbidity and mortality of
colorectal carcinoma surgery differs by insurance status. Cancer. 2004;101:
2187-2194.
9. Roetzheim RG, Gonzalez EC, Ferrante JM, et al. Effects of health
insurance and race on breast carcinoma treatments and outcomes. Cancer.
2000;89:2202-2213.
10. Roetzheim RG, Pal N, Gonzalez EC, et al. Effects of health insurance
and race on colorectal cancer treatments and outcomes. Am J Pub Health.
2000;90:1746-1754.
11. From Cancer Patient to Cancer Survivor: Lost in Translation. Washington, DC: The National Academies Press, 2005.
e8
12. Hortobagyi GN, American Society of Clinical Oncology. A shortage of

oncologists? The American Society of Clinical Oncology workforce study.
J Clin Oncol. 2007;25:1468-1469.
13. Peppercorn J, Armstrong A, Zaas DW, et al. Rationing in urologic
oncology: lessons from sipuleucel-T for advanced prostate cancer. Urol Oncol.
Epub 2012 Feb 3.
14. Gatesman ML, Smith TJ. The shortage of essential chemotherapy
drugs in the United States. N Engl J Med. 2011;365:1653-1655.
15. Ward E, Jemal A, Cokkinides V, et al. Cancer disparities by race/
ethnicity and socioeconomic status. CA Cancer J Clin. 2004;54:78-93.
16. Schnipper LE, Meropol NJ, Brock DW. Value and cancer care: toward
an equitable future. Clin Cancer Res. 2010;16:6004-6008.
17. Neubauer MA, Hoverman JR, Kolodziej M, et al. Cost effectiveness of
evidence-based treatment guidelines for the treatment of non-small-cell lung
cancer in the community setting. J Oncol Pract. 2010;6:12-18.
18. de Souza JA, Polite BN, Zhu S, et al. Utilization and costs of nonevidence-based (non-EBM) antineoplastic agents in patients with metastatic
colon cancer (mCC). J Clin Oncol 2009;29 (suppl; abstr 6002)
19. Kurian AW, Thompson RN, Gaw AF, et al. A cost-effectiveness analysis
of adjuvant trastuzumab regimens in early HER2/neu-positive breast cancer.
J Clin Oncol. 2007;25:634-641.
20. Baselga J, Bradbury I, Eidtmann H, et al. Lapatinib with trastuzumab
for HER2-positive early breast cancer (NeoALTTO): a randomised, openlabel, multicentre, phase 3 trial. Lancet, Epub 2012 Jan 16.
21. Baselga J, Cortes J, Kim SB, et al. Pertuzumab plus trastuzumab plus
docetaxel for metastatic breast cancer. N Engl J Med. 2012;366:109-119.
22. Fojo T, Grady C. How much is life worth: cetuximab, non-small cell lung
cancer, and the $440 billion question. J Natl Cancer Inst. 2009;101:10441048.
23. Abernethy AP, Etheredge LM, Ganz PA, et al. Rapid-learning system
for cancer care. J Clin Oncol 28:4268-4274
24. Park ER, Weiss ES, Moy B. Recruiting and enrolling minority patients
to cancer clinical trials. Community Oncology. 2007;4:254-257.
INTERNATIONAL VARIATION IN UNDERSTANDING

ONCOLOGISTS PROFESSIONAL DUTIES TOWARD
PATIENTS, FAMILIES, AND THEMSELVES
CHAIR
Amy P. Abernethy, MD
Durham, NC
SPEAKERS
New York University
New York, NY
Simon Wein, MD
Davidoff Cancer Center, Rabin Medical Center
Petah Tikva, Israel
A Balanced Approach to Physician

Responsibilities: Oncologists Duties
toward Themselves
By Amy P. Abernethy, MD
Overview: Although critical to the provision of best patient

care, physician self-care is an underattended aspect of responsibility in the medical professions, including oncology.
Neglecting self-care bears negative consequences for the
individual oncologist, ranging from burnout and fatigue to interpersonal and relationship stress, addiction, and disruptive
behavior. It may also contribute to medical errors, disinterest
in or depersonalization of patient care, and lower quality of
care. Because of its effect on physicians, patients, and the
health care environment, physician self-care is increasingly
recognized as an important professional responsibility.
FTEN CONSTRUED as ones obligation toward others

or used interchangeably with the concepts of reliability
and dependability, responsibility is a quality that defines
both our competence and the caliber and character of the
care we provide.
The notion of responsibility is inherent in medicine. We
enter the profession taking the Hippocratic Oath, which,
whether in its original or modernized form, lays out a set of
responsibilities to which the new physician agrees to abide.
Medical training inculcates in us a sense of responsibility
first and foremost toward our patients and secondarily toward
our colleagues, discipline, institution, and scientific field.
Less well defined, if even recognized, have been the physicians responsibilities to himself or herself. Selfless service
has been held up as an ideal in health care, whereas a focus
on ones self has been viewed with skepticism, even stigma.
Is a definition of responsibility that heavily emphasizes
duties toward others, often at the expense of the self and its
reasonable needs, realistic or even beneficial for all involved? A disproportionate focus on others, without adequate attention to self, results in negative consequences for
not only the oncologist but also patients and other stakeholders in the health care environment.
Evidence and Consequences of Inadequate Self-Care
Much as a computer that is run continuously, without

periodic shut-downs and reboots to allow for system updates,
does not run at optimal efficiency using the latest programs,
or much as an athlete who trains continuously, without
taking rest days for muscle recovery and mental recharge,
fails to achieve or sustain peak performance, the physician
who neglects to take care of himself or herself will not
function at his/her best with respect to patient care, research,
education and training, or administrative tasks. This neglect ultimately manifests in burnout and/or impairment.
A familiar concept, burnout results from tension over time
between what the individual is doing and is capable of doing
and what he or she is expected to do. It is experienced as a
progressive erosion of the soul. In physicians, burnout has
been called the silent anguish of healersfelt as anguish,
perhaps, because of a painful awareness that what we feel
called to do as medical professionals diverges from what we
are actually accomplishing and because of an escalating
inability to care about, or even to meet, our responsibilities
as professionals and healers.
Nonetheless, professional obligations, competing demands

on time, and personal priorities conspire to prevent a large
proportion of oncologists from adequately attending to selfcare in even simple ways, such as getting sufficient exercise
and sleep. This chapter discusses the need for physician
self-care and the repercussions of not meeting this fundamental responsibility. Self-care is described in the context of three
life domains: professional, personal (physical, psychological,
mental, and spiritual), and interpersonal (relationships, family,
social, and community). Strategies are provided for caring for
the self in each domain.
Physician burnout is characterized by three main dimensions: (1) emotional exhaustion, cynicism, and/or depersonalization in relationships with coworkers and/or patients,1
manifesting as detachment from ones job; (2) low sense of
personal accomplishment; and (3) perceived clinical ineffectiveness.2 Any or all of these factors might be present in the
affected physician.3 The gradual deterioration that leads to
physician burnout typically starts in medical school, residency, or fellowship, although its symptoms may not present
until midcareer.4 Signs and symptoms of burnout are
many3,5: overwhelming physical and emotional exhaustion,
overidentification or overinvolvement with patients, irritability and hypervigilance, sleep problems, social withdrawal,
professional and personal boundary violations, poor judgment, perfectionism and rigidity, questioning the meaning
of life, interpersonal conflicts, avoidance of emotionally difficult clinical situations, numbness and detachment, difficulty in concentrating, frequent illness (e.g., headaches and
gastrointestinal disturbances), and immune system impairment.
Although attention to these signs and symptoms, in ones
self and in colleagues, is the first-line approach to identifying burnout, this symptom can be diagnosed with validated
assessment instruments. The most commonly used and
well-recognized tool for this purpose is the 22-item Maslach
Burnout InventoryHuman Services Survey, which generates subscales for emotional exhaustion, depersonalization,
and personal accomplishment.6 Using this instrument as
the criterion for burnout assessment, studies have documented that half to two-thirds of oncologists experience
emotional exhaustion.7,8 Among practicing physicians more
generally, studies have reported a 46% to 80% prevalence of
moderate to high levels of emotional exhaustion, a 22% to
93% prevalence of moderate to high levels of depersonalization, and a 16% to 79% prevalence of low to moderate levels
of personal achievement.9 Burnout does not only affect the
From the Division of Medical Oncology, Department of Medicine, and the Duke Cancer
Institute, Duke University Medical Center, Durham, NC.
Address reprint requests to Amy P. Abernethy, MD, Duke University Medical Center, Box
3436, Durham, NC 27710; e-mail: amy.abernethy@duke.edu.
1092-9118/10/1-10
e9
AMY P. ABERNETHY
midcareer physician; in one cross-sectional study, more than

three-fourths of internal medicine residents exhibited burnout.10
Risk factors for physician burnout have been described
across medical disciplines. Demographic factors associated
with increased risk of burnout include younger age and
junior status, female sex, and unmarried status.11 Personality factors predisposing physicians to burnout include high
level of motivation and intense investment in ones profession12; personality traits, such as intensity, impulsivity, and
compulsiveness; previous mental health problems, particularly depression; and wishful thinking as ones coping style.
Environmental or contextual risk factors include work overload; perception of a lack of control over ones workload,
when that workload exceeds capacity; lack of social support
from colleagues; dissatisfaction with resources at work;
practice within a high-risk medical specialty (such as oncology); work stressors, including reimbursement issues,
complex regulations, and interactions with insurers; organizational concerns and management style; and family stress.
The risk of burnout increases proportionally with time spent
in direct patient care; time away and a mix of job responsibilities are protective.8 Care for people who are dying and
disease that is not responsive to aggressive treatment, both
prominent in oncology, are sources of burnout that accumulate over time, especially when the physician feels illequipped to provide expert end-of-life care or experiences
a sense of failure when disease progresses. Spirituality and
religiosity have been found to exert a protective effect
against burnout.
Multiple dire consequences result from physician burnout
and from inadequate care of the self more generally. Among
physicians, substance abuse rates of 10% to 15% have been
reported,13 although physicians tend to use alcohol and
prescription medications rather than illicit substances.14
Physician fatigue is cited in the Institute of Medicines
seminal report, To Err is Human, as a factor contributing to
the tens of thousands of medical errors occurring in the
United States each year.15 A prospective, longitudinal study
found increased burnout and reduced empathy to be associated with increased likelihood of committing a self-perceived
KEY POINTS
e10
Self-care encompasses all strategies implemented to

support ones own optimal health and well-being.
Because failure to attend to self-care negatively affects patients both directly and indirectly (e.g., depersonalization leading to lower empathy or physician
exit from the profession), self-care is a vital responsibility of every oncologist.
Self-care can be conceptualized as a set of activities
that one uses on a regular basis to promote fulfillment and wellness in the professional, personal, and
interpersonal domains of life.
A vast array of options for self-care, and supporting
facilitators, are available in each domain.
Self-care strategies must be personalized, with sensitivity to cultural nuances and other individuallyspecific determinants of best approach.
medical error in the subsequent 3 months.16 Failure of

physician self-care to establish well-being is evidenced in
high suicide rates. A meta-analysis that included 25 studies
reported that male physicians commit suicide at a rate 40%
higher than that of men in the general public, and female
physicians do so at 130% the rate of women in general.17
Among physicians 55 years and younger, higher perceived
stress is associated with (1) lower satisfaction levels, which
in turn are related to intention to quit, decrease work hours,
change specialty, or leave direct patient care, and (2) poorer
mental health, which in turn is related to intention to leave
direct patient care.18 Other potential consequences of
inadequate physician self-care include impaired job performance, poor physical health (e.g., headaches, sleep disturbance, hypertension, and cardiac events), poor emotional
health and well-being (e.g., irritability, fatigue, anxiety, and
depression), and relational and marital difficulties.
Ultimately, it is not only the oncologists self that is
affected. Physically, emotionally, mentally, or spiritually
depleted physicians can neither take good care of patients
nor model good health or a lifestyle conducive to health for
their patients. Physicians tend to counsel patients in ways
that are consistent with their own health habits,19 and if
these habits are not sound, patients receive bad advice.
Furthermore, as oncologists, we are not just caretakers of
people with cancer, we are also members of our own families
and circles of care; physicians experiencing burnout cannot
contribute as well to family life and other valued relationships, and consequences can be disastrous for all involved.
Professional Recognition of Physician Burnout and
Need for Self-Care
The American Medical Association (AMA) Code of Medical

Ethics, Opinion 9.0305, states, To preserve the quality of
their performance, physicians have a responsibility to maintain their health and wellness, construed broadly as preventing or treating acute or chronic diseases, including
mental illness, disabilities, and occupational stress. A variety of resources are available to support physicians attention to their own health. The AMA produces the Physicians
Guide to Personal Health Toolkit, which covers healthy
eating, physical activity, reduced alcohol consumption, and
smoking cessation, as well as the AMA Physician Health
e-Letter. The biennial International Conference on Physician
Health, sponsored by the AMA, Canadian Medical Association, and British Medical Association, presents research and
original presentations on topics such as burnout and peer
support, physician health as linked to quality and patient
safety, resilience and work-life balance, and physical and
mental health.
Despite the availability of supportive resources, backed by
injunctions from respected agencies, many oncologists remain unclear about, neglect, or postpone active strategies to
take optimal care of themselves. In a 2009 survey of California physicians (n 1,875), 53% reported moderate to
severe stress, 35% reported no or occasional exercise, 34%
reported getting 6 hours or less of sleep per night, and 27%
never or occasionally ate breakfast.20
Conceptual Framework
Clarity in conceptualizing what, specifically, is entailed in

physician self-care may facilitate individuals responsible
SELF-CARE FOR ONCOLOGISTS
Fig. 1. Consequences of self-care or no

self-care for self and patients.
planning and decisions. The World Health Organization

defines wellness as the optimal state of health of individuals and groups and explains that there are two focal
concerns: the realization of the fullest potential of an individual physically, psychologically, socially, spiritually, and
economically, and the fulfillment of ones role expectations in
the family, community, place of worship, workplace, and
other settings.21 Thus defined, a physicians duties to himself or herself span the totality of life experience and affect
not only himself or herself but also those with whom he or
she interacts.
The oncologists responsibilities to self can be conceptualized in three domains: professional, personal, (physical,
psychological, mental, and spiritual), and interpersonal (relational, family, social, and community) (Fig. 1). The following sections briefly describe each domain and provide
pragmatic approaches to self-care within each. Because
many of the basic elements of self-care are straightforward
and likely familiar to most oncologists, only simple guidelines and principles are presented, with discussion of facilitators for their implementation. There is no one-size-fits-all
approach; each oncologist, knowing himself or herself best,
must craft an approach that fits well with his or her own
preferences, characteristics, and circumstances.
Duties to Self within the Three Domains of
Responsibility
Professional. Most, if not all, oncologists enter medicine

with a strong sense of professional responsibility. Upholding
this initial commitment is integral to positive self-esteem
and a sense of personal integrity. Activities to maintain and
further develop ones professional competence, such as conferences and symposia, and to pursue ones own professional
interests, such as committee work and research, offer mechanisms for cultivating professional competence. Ensuring a
mix of responsibilities and some time away from patient care
is important. Goal setting in the professional domain is a
method of articulating personal standards of excellence and
holding ones self accountable to personally meaningful
progressa form of self-care.
Because caring for people at or near the end of life can be

especially burdensome, exacting a toll in terms of the individuals mental and emotional resources, specific strategies
to build palliative care skills can help with self-care. Some
strategies are to hone ones skills and confidence in delivering bad news; obtain palliative carespecific training
through conferences and training videos; develop relationships with palliative care colleagues and call on them
frequently; debrief after difficult cases and losses, and take
time to acknowledge that this is hard; recognize that some
patients affect us more than others, and that this is normal;
and finally, when a death occurs, take time to reflect, allow
yourself to grieve, attend the funeral, reconnect with the
family, write a letter, and smile at the gift of knowing the
person when alive.
In the professional setting, we must care for ourselves and
the teams within which we work. A few strategies for care of
self and others in the work environment are to manage the
volume of time spent in patient care; ensure that the team
has mechanisms in place to debrief and share stressful
experiences; allow staff time to get together outside the
workplace; role model good personal and professional habits
(e.g., time off for important family events and prioritization
of exercise) and promote these for all members of the team;
and be attuned to any signs of burnout among team members and educate them to be alert to these signs in self and
others. A well-functioning, personally healthy team provides
an important support network for the oncologist which can
undergird self-care for each individual and for the team as a
whole.
Personal. Table 1 presents recommendations for personal
self-care in the domains of physical, psychological and cognitive, and spiritual wellness. Although most oncologists
will agree to the needs listed and to the corresponding
recommendations and although it is important to promote
these to patients, far fewer actually succeed in integrating
these or other self-care strategies into daily living. (Table 1
is intended not to provide a comprehensive list of self-care
needs, recommendations, or strategies but rather to offer
an initial set of practical suggestions. Knowing himself or
e11
AMY P. ABERNETHY
Table 1. Recommendations for Personal Self-Care
Domain
Need
Physical
Sleep
Recommendations
Aim for 79 hours per night

Try sleep hygiene before resorting to sleep
medications
Sleep in a totally darkened room
Maintain consistent bedtime and waking hours
Minimize bedroom stimulation
Avoid working in the bedroom
Nutrition
Exercise
Psychological and
cognitive
Balance
Eat a diet featuring lean proteins (e.g., poultry, fish,

eggs, and unsweetened yogurt), vegetables, fruits,
nuts and seeds, legumes (e.g., black beans), and
moderate portions of whole grains
Avoid trans and hydrogenated fats
Avoid stimulants and substances that affect
metabolism (e.g., sugars, caffeine, and alcohol)
Supplement to ensure adequate vitamins and
antioxidants
Spread caloric intake across the day, beginning with
breakfast
Drink at least eight 8-oz glasses of water per day
Get at least 30 minutes of exercise on 3 days or
more per week
Mix types of exercise to maintain strength (e.g.,
bodyweight exercises), flexibility (e.g., yoga), and
cardiovascular conditioning (e.g., lap swimming)
Choose the outdoors, when possible
Be aware of the valued domains in your life (e.g.,

work, family, friends, and external commitments)
Take stock of what fulfills you in each domain
Allocate time to (and prioritize) outside interests
(e.g., creative, intellectual, athletic, and community)
Stress reduction and

relaxation
Practice a relaxation technique (e.g., meditation and

deep breathing)
Use complementary alternative medicine approaches
Engage in a relaxing sport or hobby
Spend time outdoors
Mental rejuvenation
Feed your sense of humor

Play games
Activate your sense of humor
Take time away from workcompletely away
Facilitating Strategies
Set a black out time beyond which you dont

answer the telephone or email
Discontinue work at a set time each day
Keep electronics (e.g., television, laptop, and
mobile telephone) out of the bedroom
Buy a weeks lunches from the salad bar on the

weekend or have healthy lunches delivered
Set up a healthy breakfast the night before

Pack healthy snacks (e.g., nuts and fruit)
Carry a water bottle
Make use of grocery stores list-filling services,
including online shopping
Monitor your coffee and alcohol intake
Work out with a personal trainer
Exercise before work
Make standing exercise dates with friends

Take a refreshing walk or jog outside
Work with a life coach
Write down realistic goals; post them prominently
and review frequently
Set aside 10 minutes before bedtime or in the
early morning to journal, meditate, or reflect
Schedule a regular massage
Make friend/family activities fun and light (e.g.,
watch a comedy)
Walk (even briefly) under the stars at days end
Take regular vacations

Find a brain game (e.g., sudoko or cross-word
puzzle) that appeals to you and make it a
regular practice
Join a drop-in or online book club with a realistic
schedule
Engage in activities that transport you to another

realm (e.g., reading fiction)
Spiritual
Connection to a higher
power
If you are religious, attend services at a church,

temple, or synagogue
Pray, meditate, or engage in a meaningful spiritual
practice regularly
Spend time outdoors
Find sources of beauty to appreciate
Sign up to usher to ensure your attendance at

services
Take minibreaks at points throughout the day to
listen within
Visit a gallery or art museum
Listen to classical music on your commute
Read a poem at bedtime
Sense of meaning and

purpose
Consider what makes you feel most fulfilled
During 1 week, take notes of times when you feel

fulfilled and purposeful
Plan to incorporate this/these things into daily life
on a regular basis
Share this plan with someone who cares about
you to support accountability to self
Identify ways that you can attend to these aspects of

your life
Prioritize these commitments
Reward yourself for living your values
e12
SELF-CARE FOR ONCOLOGISTS
herself best, each oncologist is encouraged to adapt this sort

of list into a realistic and effective self-care plan.)
Given oncologists hectic schedules and competing demands, these recommendations can be challenging to implement. It is useful, therefore, to identify and enlist
facilitators. For example, having a regularly scheduled personal trainer can help ensure that strength training occurs;
making a standing weekly commitment to walk, run, or bike
with a friend can help ensure cardiovascular exercise. Engagement of a housecleaning service can liberate weekend
time for hobbies and other personally fulfilling activities. A
life coach can help establish time management practices
that support balance and identify pragmatic ways to reach
lifestyle and balance goals. Something as simple as purchasing a water bottle and carrying it at all times can help
remind a busy physician to stay well hydrated and set a
positive example for staff and patients.
Stress reduction warrants special mention as key to avoiding burnout. There are many options for alleviating or
reducing stress, defined by individual character and interests. Spending time outdoors; meditating; participating in a
favorite sport; pursuing a hobby, such as playing a musical
instrument; simple relaxation, such as reading a novel in the
evening; and activating ones sense of humor all represent
effective approaches to stress reduction while also meeting
self-fulfillment needs. A complementary and alternative
medicine practitioner, such as a good acupuncturist or
massage therapist, can become a valued ally in stress
reduction. Resources such as meditation, relaxation, or
guided imagery CDs are readily available and can make
stress reduction both convenient and enjoyable. There are
many forms of deep relaxation, including listening to ocean
waves or soothing music, sitting or walking meditation,
biofeedback, transcendental meditation, visualization,
mindfulness meditation, neurofeedback, and progressive
muscle relaxation. The restful alertness (relaxation response) induced by these interventions differs neurophysiologically from the sleep state; it enhances mental clarity
and lowers body catechol levels, cortisol level, oxygen demand, blood pressure, and pulse.
Although more than 90% of Americans believe in God or a
universal higher power,22 spiritual self-care is not commonly recognized as one of the oncologists duties. Some
individuals may believe that the stress management recommendations listed in this chapter adequately address their
needs in the spiritual domain. Others, however, must engage in a regular spiritual practice (e.g., daily prayer or
meditation or weekly attendance at religious services) to feel
that they are attending well to their spiritual dimension.
Finding meaning and higher purpose, beyond individual
satisfaction and gain, in ones profession can support a sense
of spiritual fulfillment. Self-care in this deeply personal
domain requires self-knowledge and a certain level of intimacy with ones self; cultivation of inner connectedness in
whatever way the individual chooses can satisfy self-care
needs and support well-being.
Interpersonal. Like all people, oncologists live within a
social context, a web of relationships. The extent to which
the oncologist attends to his or her responsibilities to self
therefore has an effect on others. Family commitments to
significant others, spouses, and/or children require, above
all else, timea commodity in short supply in the oncology
profession. Inability or failure to meet commitments (im-
plicit or explicit) to ones family can induce considerable

dissatisfaction and distress in all involved. A vital part of
oncologists self-care is thus attending to relationships.
In spending time with others, both quality and quantity
are important; no level of quality time, however exceptional
in quality, can compensate for spending the bare minimum
quantity of time with loved ones. Critical is the ability, in
whatever time is available, to be truly present with others,
rather than multitasking with mobile device in hand or
being so exhausted as to barely engage in conversation.
Strategies for ensuring that relationship responsibilities are
met include family rituals, such as game night or Saturday
breakfast out, and a regular date night with a spouse or
partner. Making a commitment to a family dinner at least 5
or 6 nights per week can be considered an evidence-based
intervention; studies have shown that family dinners are
associated with healthier eating patterns in older children
and adolescents,23 healthy adolescent development, and less
engagement in risk behaviors.24
Social support is an important facilitator of well-being in
busy individuals and those with plenty of time for socializing. Friends and acquaintances provide an emotional outlet
and an opportunity to give and receive ideas, thoughts,
feedback, concern, and affirmation. Although technology
and social networking provide endless opportunities for
connecting with others, there is no substitute for face-to-face
encounters in conveying a true sense of support.
Recognizing burnout in ones self and colleagues and
taking responsible action are important interpersonal duties
of each oncologist. Signs and symptoms include the following: frequent irritability or anger, depression, disengagement and depersonalization, apathy, chronic fatigue and/or
depressive symptoms, and substance abuse (e.g., alcohol,
recreational drugs, and inappropriately used pharmaceutical drugs). These are neither normal nor acceptable attributes of an oncologist. On identifying these symptoms, seek
a psychologists or counselors help for ones self or, in the
case of a colleague, either refer the individual to a practitioner who can provide direct help or report the concern to a
senior supervisor or associate, as appropriate.
In medicine, a system for accountability and attention to
each other is often lacking. When a colleague is demonstrating unhealthy behaviors, ones responsibility and course of
action are often unclear. Whom should we tell? How can we
enlist help? Can we ensure appropriate follow-up? Although
clear lines of authority and accountability exist in other
disciplines, such as nursing, we in medicine are frequently
unsure where to turn. Although lack of line management
affords us a valued measure of autonomy and flexibility, it
also leaves us unclear of how to maneuver when a colleague
is troubled. In the absence of clear directives, oncologists
should take seriously the responsibility to attend to each
other and should discuss within their professional teams a
system for encouraging one anothers well-being and intervening when warning signs emerge. Until such systems
become explicit aspects of the oncologists workplace, it will
remain important to share a common understanding of this
responsibility.
Different Cultural Approaches to Duties to Ones Self
An individuals exploration into and definition of his or her

duties to self will depend not only on personal factors, such
as individual values, preferences, capacities, and circum-
e13
AMY P. ABERNETHY
stances (e.g., stage of career, family needs, and social configuration), but also on his or her cultural background. There
is a paucity of literature examining cross-cultural differences in self-care between U.S. and foreign-born physicians
and between members of racial/ethnic majority compared
with minority populations. Studies have made cross-cultural
comparisons of attitudes toward self-care between the
United States and Germany and explored differences in
U.S. and Swedish constructs such as self-concept and wellbeing.25 For current purposes, it is important to be aware
that ones own concept of self, definition of personal wellbeing in the domains discussed herein, and preferred selfcare strategies will be tempered by individual background
family, upbringing, culture, and life experiences.26
Conclusion
Self-care is not purely a private matter; the extent to

which oncologists attend to, or deny, the needs of their selves
has repercussions in the personal, interpersonal, and professional spheres. Burnout is a prevalent and serious phenomenon, directly related to inadequate self-care. This
chapter presents a flexible framework to help the oncologist
consider, and responsibly attend to, duties toward himself or
herself. Currently, metrics are not used to track oncologists

responsibility in caring for themselves; rather, performance
measures typically evaluate patient care and other externally focused duties. With the understanding that we have a
de facto honor system with respect to self-care, it is incumbent on every oncologist to adopt strategies for ensuring his
or her health and well-being. Short-term and long-term
benefits will be evident in multiple domains, including
physical, mental, and psychological health, job satisfaction,
and quality of life. In addition, our patients will benefit from
having healthier, more present and available physicians.
Acknowledgments
Complete Funding Disclosure: As of February 2012, Dr.
Abernethy has research funding from the U.S. National Institutes of Health, U.S. Agency for Healthcare Research and
Quality, Robert Wood Johnson Foundation, Pfizer, Eli Lilly,
Bristol Meyers Squibb, Helsinn Therapeutics, Amgen, Kanglaite, Alexion, Biovex, DARA Therapeutics, Novartis, and
Mi-Co; these funds are all distributed to Duke University
Medical Center to support research. In the last 2 years she has
had nominal consulting agreements ($10,000) with Helsinn
Therapeutics, Amgen, and Novartis.
Author
Amy P. Abernethy
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Bristol-Myers
Squibb; Helsinn
Honoraria
Amgen; Novartis
Research
Funding
Expert
Testimony
Other
Remuneration
Amgen; BioVex;
Bristol-Myers
Squibb; DARA;
Eisai; Helsinn;
KangLaiTe; Lilly;
Pfizer
REFERENCES
1. Maslach C, Leiter MP. Early predictors of job burnout and engagement.
J Appl Psychol. 2008;93:498-512.
2. Maslach C. Job burnout: new directions in research and intervention.
Curr Direct Psychol Sci. 2003;12:189-192.
3. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annu Rev Psychol.
2001;52:397-422.
4. Spickard A Jr, Gabbe SG, Christensen JF. Mid-career burnout in
generalist and specialist physicians. JAMA. 2002;288:1447-1450.
5. Vachon ML. Staff stress in hospice/palliative care: a review. Palliat Med.
1995;9:91-122.
6. Maslach C, Jackson SE. The measurement of experienced burnout. J
Occup Behav. 1981;2:99-113.
7. Grunfeld E, Whelan TJ, Zitzelsberger L, Willan AR, Montesanto B,
Evans WK. Cancer care workers in Ontario: prevalence of burnout, job stress
and job satisfaction. CMAJ. 2000;163:166-169.
8. Allegra CJ, Hall R, Yothers G. Prevalence of burnout in the U.S.
Oncology community: results of a 2003 survey. J Oncol Pract. 2005;1:140-147.
9. Chopra SS, Sotile WM, Sotile MO. Physician burnout. JAMA. 2004;291:
633.
10. Shanafelt TD, Bradley KA, Wipf JE, Back AL. Burnout and selfreported patient care in an internal medicine residency program. Ann Intern
Med. 2002;136:358-367.
11. Kearney MK, Weininger RB, Vachon ML, Harrison RL, Mount BM.
Self-care of physicians caring for patients at the end of life: Being connected . . . a key to my survival. JAMA, 2009;301:1155-1164, E1.
12. Leiter MP, Maslach C. A mediation model of job burnout. In: Antoniou
CC, Cooper CL (eds). Research Companion to Organizational Health Psychology. Cheltenham, United Kingdom: Edward Elgar Publishing; 2005.
13. Gastfriend DR. Physician substance abuse and recovery: what does it
mean for physicians-and everyone else? JAMA. 2005;293:1513-1515.
14. Hughes PH, Brandenburg N, Baldwin DC Jr., et al. Prevalence of
substance use among US physicians. JAMA. 1992;267:2333-2339.
e14
15. Institute of Medicine. To Err is Human: Building a Safer Health

System. Washington, DC: National Academy of Sciences; 1999.
16. West CP, Huschka MM, Novotny PJ, et al. Association of perceived
medical errors with resident distress and empathy: a prospective longitudinal
study. JAMA. 2006;296:1071-1078.
17. Schernhammer ES, Colditz GA. Suicide rates among physicians: a
quantitative and gender assessment (meta-analysis). Am J Psychiatry. 2004;
161:2295-2302.
18. Williams ES, Konrad TR, Scheckler WE, et al. Understanding physicians intentions to withdraw from practice: the role of job satisfaction, job
stress, mental and physical health. Health Care Manage Rev. 2001;26:7-19.
19. Schwartz JS, Lewis CE, Clancy C, Kinosian MS, Radany MH, Koplan
JP. Internists practices in health promotion and disease prevention. A
survey. Ann Intern Med. 1991;114:46-53.
20. Bazargan M, Makar M, Bazargan-Hejazi S, Ani C, Wolf KE. Preventive, lifestyle, and personal health behaviors among physicians. Acad Psychiatry. 2009;33:289-295.
21. World Health Organization. Health Promotion Glossary Update. Geneva, Switzerland: World Health Organization; 2000.
22. Pew Charitable Trusts. US religious landscape survey. In: Pew Forum
on Religion and Public Life. Philadelphia, PA: Pew Charitable Trusts; 2008.
23. Gillman MW, Rifas-Shiman SL, Frazier AL, et al. Family dinner and
diet quality among older children and adolescents. Arch Fam Med. 2000;9:
235-240.
24. Fulkerson JA, Story M, Mellin A, Leffert N, Neumark-Sztainer D,
French SA. Family dinner meal frequency and adolescent development:
relationships with developmental assets and high-risk behaviors. J Adolesc
Health. 2006;39:337-345.
25. Whetstone WR, Hansson AM. Perceptions of self-care in Sweden: a
cross-cultural replication. J Adv Nursing. 1989;14:962-969.
26. Whetstone WR. Perceptions of self-care in East Germany: a crosscultural empirical-investigation. J Adv Nursing. 1987;12:167-176.
Are Oncologists Accountable Only to

Patients or Also to Their Families?
An International Perspective
By Antonella Surbone, MD, PhD, and Lea Baider, PhD
Overview: In most societies, health professionals traditionally carry responsibility only toward their patients. However,
this is not the case in all cultures. In the contemporary
practice of oncology in Western cultures, there is a shift
toward assuming broader responsibility for patients with cancer families during the illness course, the grieving stage, and
in cancer prevention and genetic counseling.
Traditional family, community, and religious values play a
central role in determining peoples perceptions and attitudes
toward life and death as well as toward caregiving for a sick
relative. The meaning of cancer illness within the family
culture is thus influenced not only by each individuals values
and beliefs but also by the familys makeup and dynamics, as

well as their taboos and secrets.
Global cancer care should therefore be directed at the
family as a unit, while respecting patient autonomy and
privacy. This reappraisal of our traditional understanding of
physicians duty as solely directed at the patient is reflected
in the recent US trend toward a patient- and family-centered
care approach. An additional challenge for oncology professionals is to integrate and tailor interventions toward the
needs of both care recipients and caregivers and relate it to
this dyad as the basic and enduring unit of care.
Patient- and Family-Centered Care in the United States
protective needs of individuals, especially children and the

elderly.5
The American Academy of Pediatrics and the American
College of Emergency Physicians support the model of
patient- and family-centered care (PFCC). The latter defines
the PFCC as an approach to health care that recognizes the
role of the family in providing medical care, encourages
collaboration between the patient, family, and health care
professionals, and honors individual and family strengths,
cultures, traditions, and expertise.6 This approach has
proven effective in pediatric hospital and home care and
emergency settings and is now being extended to adult
care.7
According to The Institute for Patient- and FamilyCentered Care, the patient- and family-centered approach
focuses on the partnerships between patients, families, and
providers at the clinical level. Patients and families are
engaged as advisors and leaders in developing new policies
on quality and safety. Practitioners recognize the vital role
of family members of patients of all ages and acknowledge
that emotional, social, and developmental support are integral components of health care. The definition of family, as
well as the degree of the familys involvement in health care,
is determined by the competent patient, who maintains
control of decisions concerning his or her own health care.8,9
In its expert report on cancer pain relief and palliative
care, the World Heath Organization states the importance of
enhancing the quality of life of both patients and family
members.10 Teno et al developed a conceptual model of
patient-focused, family-centered medicine in end-of-life care,
where the patient is at the center of care and the family,
defined as all peoplewhether or not they are blood relativeswith whom the patient has a very close or intimate
The Institute of Medicine has identified the concept of

patient-centered care as one of the keys to improving health
and quality of care in the United States, including efficiency
and costs.3 In patient-centered care, physicians clearly have
responsibilities toward both the patient and the family. The
American Academy of Family Physicians defines family as
a group of individuals with a continuing legal, genetic,
and/or emotional relationship. Society, the Academy states,
relies on the family group to provide for the economic and
From the Department of Medicine, New York University Medical School, New York, NY;
and the Hebrew University Medical School, Sharett Institute of Oncology, Hadassah
University Hospital, Jerusalem, Israel.
Address reprint requests Antonella Surbone, MD, PhD, Department of Medicine, New
York University Medical School, 550 First Ave., New York, NY 10016; email: antonella.
surbone@gmail.com.
1092-9118/10/1-10
ANCER HAS become a global health emergency because of the aging of the world population that is
occurring at unprecedented rates and portending an everincreasing demand for care. It is expected that by 2030, the
US elderly population conventionally defined as the population segment composed by people over age 65will double
and reach about 70 million people.1 As the population ages,
morbidity for chronic illness and disability increases, with
consequent escalating demands for informal care delivered
by family members or close friends. Oncologists are asked to
recognize the key role of family in each patients experience
of cancer and progressively assume more responsibilities
toward families, during the entire course of the illness, from
diagnosis to survivorship or end-of-life, as well as in the
grieving stages.2
A familys experience with cancer, including caregiving,
always occurs in a particular cultural milieu: the traditional
and religious values of each community play a central role in
determining peoples perceptions and attitudes toward illness and death, as well as toward caregiving to a sick
relative. The meaning of cancer is thus influenced not only
by each individuals values and beliefs but also by the
familys convergence and diversity of meaning, as well as
their taboos and secrets. Consequently, in many cultures
physicians have always included patients, families, and
communities in their professional responsibility.3
Although Western physicians professional duty has traditionally being only toward their patients, todays medicine
increasingly acknowledges the need for health professionals
to involve the family in patient-centered care and to develop
a functional system of collaboration and partnership with
patients and their families.2,4
e15
SURBONE AND BAIDER
relationship is also acknowledged. This approach is based

on sharing information with family members on what to
expect and the skills necessary to help care for the patient at
home, providing desired physical comfort and emotional
support, and promoting shared decision making focused on
the patient, while also attending to the needs of family
members.
Sidebar 1. Demographics of Care in the United

States
Role of the Family and Informal Caregiving in Oncology
Approximately 28.8 million adults in the United States

are now family caregivers.1,12 Sidebar 1 lists are some
relevant data on the demographics of care in the United
States, where unpaid caregiving is the largest source of
long-term care services (Sidebar 1).
Caring is fundamental to human survival.13 It is understood to imply a distinct way of being, believing, and acting
that calls for commitment, emotional involvement, knowledge, and new coping skills. It motivates families and gives
new meaning and structure to life.14 At the same time,
caregiving imposes a great burden on families at psychologic, social, and financial levels. As many as two-thirds of
caregivers report physical or mental health problems owing
to caregiving, and nearly 50% have a chronic health condition that requires ongoing medical care.1
Because of the progressive shift from hospital to ambulatory or home cancer care, informal family caregivers are
often required to replace skilled health care workers in the
delivery of complex care to their ill family members. They
KEY POINTS
e16
Although Western health professionals traditionally

carry responsibility only toward their patients, the
physicians duty in other cultures extends to families
and communities.
Todays Western medicine increasingly acknowledges
the need for health professionals to include the family
in patient-centered care and to develop a functional
system of collaboration and partnership with patients
and their families.
The American Academy of Family Physicians defines
family as a group of individuals with a continuing
legal, genetic, and/or emotional relationship and
acknowledges that society relies on the family group
to provide for the economic and protective needs of
individuals, especially children and the elderly.
As informal caregiving by family members is rapidly
increasing worldwide, oncologists are asked to assume broader responsibility toward their patients
families during the course of the patients illness and
in the grieving stage.
Cancer is a family illness that can only be understood in terms of the social, cultural, religious, and
family responses to the underlying disease, including
the ways in which the patient and family conceptualize and elaborate it. For this reason, the family is
increasingly recognized as the basic social and ethical
unit of care.
Fifty-two million informal/family caregivers provide care to an ill or disabled person 20 years and
older.
Thirty-four million adults (16% of population)
provide care to persons 50 years and older.
Between 6 to 7 million family caregivers provide
care to adults 65 years and older who need
assistance with everyday activities.
It is estimated there will be 37 million informal
caregivers by 2050, an increase of 85% from 2000.
are asked to perform multidimensional tasks, including

treatment monitoring; treatment-related symptom management; emotional, financial, and spiritual support; and assistance with personal and instrumental care.15,16
Family members of patients with cancer strive to make
sense of the unforeseen illness and its intrusion by enduring
and reinterpreting the meaning of each others experience
and finding a shared meaning within the private family
culture.17 Cultural expectations also regulate and influence
basic meanings in the course of the socialization process and
the roles that each member might assume. Even though
caregiving can be a deeply rewarding spiritual experience of
connection with a loved one in need, caregivers of patients
with cancer often perceive themselves as living a shrinking
life, being forced to take responsibilities beyond what they
feel comfortable doing, struggling to keep their home as a
home, and no longer feeling the joy of being together with
their family.18,19 Sidebar 2 lists some of the positive and
negative behavioral changes that can arise within a family
when a member has cancer.
To optimize the quality of caregivers lives and their
ability to support and enhance the quality of care delivered
to each patient with cancer, we must take into account the
extent to which providing care results in physical, emotional, spiritual, and financial burdens for them and the
ways in which these burdens can be addressed. Empirical
studies show that the burden placed on family caregivers
has negative effects on the quality of life of both the patients
and their caregivers, particularly during advanced stages of
cancer.20 Professional interventions should therefore be directed at the family as a unit and not aimed solely at making
caregiving easier. The challenge for oncology professionals is
to integrate and tailor interventions according to the specific
needs of the care recipient and caregiver and relate to this
Sidebar 2. Behavioral Changes in Response to

Cancer in a Family Member
New coping styles and strengths

Discovery of meaning
Increased cohesiveness and spirituality
Stress and crises following recovery
Complaints, anger, and withdrawal
ONCOLOGISTS DUTIES TO PATIENT AND FAMILIES
dyad as the basic and enduring unit of care, while respecting

and fostering the patients autonomy.19,21,22
Oncologists Responsibilities to Their Patients and
Families in Different Cultures
Within each society, there are divergent groups with

particular forms of family culture. Both the broad culture
and private family culture frame peoples attitudes toward
family roles, concepts of health, illness and suffering, and
decisions about life and death. Systems of belief and cultural
values are not static. Rather, cultures are dynamic and
evolve over time under the influence of many demographic,
social, and political factors.23
Culture has been powerfully described as a tapestry,
where each single thread cannot be understood outside of
the entire fabric, and different attitudes and beliefs cannot
be judged outside the entire context of any given culture.22
Applied to health care, this means that the same illness
under different social and familial conditions could elicit diverse sets of emotional behaviors, suggesting the need for
differential approaches to different patientfamily units.23,24
Amani was a 34-year-old religious Muslim. She, her husband, and their eight children lived with the husbands
extended family. She was diagnosed in 2011 with stage IV
metastatic breast cancer, along with brain and lung metastases. Shame and guilt prevented her from coming earlier for
an examination. Her male physician communicated only
with her family. During Amanis hospitalization, the word
cancer was never mentioned and she refused to have any
direct discussions with physicians and nurses. Her preferences made perfect sense to her based on traditional values,
but the Western medical team was frustrated and found it
difficult to withhold the truth and allow the family to choose
her place of death. Yet Amanis family took her home, where
she died a few days later, surrounded by the familys prayers
for Allahs protection and hopes for an afterlife (Lea Baider,
from a private diary).
Disclosing information to family members, rather than
patients themselves, and allowing families to make decisions, are among the best known examples of cultural
differences in physicians responsibility toward patients
and/or their families. They stem from a cultural belief in
protecting patients from bad news and painful decisions.
In contrast to the patient autonomy approach, considered
potentially harmful to patients with cancer, the prevalent
approach among most cultural groups is a guided medical
paternalism that is family-oriented. Attitudes and practices
regarding the most appropriate behavior toward seriously ill
patients influences the physicians duty in a manner that
privileges family ties and family involvement in decision
making over patient autonomy.25,26 The view that physicians have a duty toward their patients families and that,
together with them, they should protect patients from bad
news is still dominant in many cultures. Terminal illness is
usually disclosed to the family and not to the patient, who is
generally being cared for at home rather than in a hospital
or medical institution.
A 2011 review related to disclosure of terminal illness in
codes of medical ethics from 14 Islamic countries showed
that approximately one-third of the physicians do not give
any information about cancer prognosis and there is a
general tendency to favor a religiously paternalistic/utilitarian, family-centered approach.27 The Islamic religion not
only shapes the social and personal lives of Muslims but also
plays a significant role in guiding values and beliefs in every
life event or circumstance, such as illness and death.28
Muslim physicians have a code of medical ethics consistent
with cultural and religious norms, which is incorporated into
their daily practice.
A study of medical care in Lebanon reported that it is the
physicians duty to withhold the true nature of the diagnosis
and prognosis from the patient, even though no false hope
should be given to the family in the case of serious or
terminal illness.25 In Jordan, physicians and other health
care providers attending to women with breast cancer are
required to consider the influence of culture, religion,
and the financial and personal characteristics of each
patient before offering information and support, and they
often withhold them when asked to do so by the patients
family.30
In Nepal, physicians are duty bound not to discuss a
diagnosis of malignancy with their patients: communication
should be only with family members, who then may filter the
information to the patient.36 In Pakistan, although breast
cancer is the most common form of cancer among women,
the experiences of patients and the suffering they endure
routinely go unreported.32 The social stigma attached to a
diagnosis of breast cancer often prevents women from informing relatives and physicians of their illness as breast
cancer is viewed as a taboo that diminishes the opportunities for single women to get married and there is also a social
stigma to having a mastectomy. When discovering a lump
in their breast, many women seek advice from local homeopaths on medicines or search spiritual support rather than
medical advice. Often, they are afraid to seek medical help
because of their mistrust of Western physicians and breast
cancer is often diagnosed at advanced stages.32
In the Islamic Republic of Iran, breast cancer ranks first
(21.4%) among all malignancies diagnosed in women. One in
four women with breast cancer report for screening less
frequently than recommended. The screening process itself
is characterized by misinformation, lack of guidance and
referral, and concealment on the part of the physician
concerning the severity of the diagnosis and prognosis.33 In
a recent research study, several women who visited either a
private physician or one working in the public health sector
described how they felt when they learned that the physician had concealed the diagnosis of breast cancer. One
woman described the physician who had diagnosed her case
and referred her to the public hospital: He did not tell me
that I have breast cancer for month and a half. . . . All that
time, he did not spill out the word cancer. Several women
recruited from the public hospital interpreted this concealment as denial: Hethe physician did not want to terrify
us by saying, you have cancer. Notwithstanding the physicians intention, many women expressed the need to know
the truth about their medical condition.
These and other data show that, while often consonant
with cultural and religious values and norms, patients with
cancer are no longer satisfied with nondisclosure and wish to
be more respected in their autonomy and right to take an
active role in decision making about their care. Hence, it
appears that, in family- and community-centered cultures,
the duty of the oncologist is to involve his or her patients,
rather than the family only.34
e17
SURBONE AND BAIDER

Conclusion
In more individualistic cultures, such as Western ones,

family roles also change when a member has cancer and the
psychologic, social, and financial burdens of family caregiving increase. Consequently, the patient doctor relationship is shifting from its traditional dual partnership to a
more extended collaboration among professionals, patients,
and their families. This involves a parallel shift in the
culture of medicine in which each patients family is included in the team as part of a functional system.2,35 A
reappraisal of the notion of individual autonomy is necessary and there is a growing trend toward speaking of
relational autonomy, based on a renewed sense of connectedness among family members. Relational autonomy refers to an understanding of autonomy in light of family and
community ties and of individual, cultural, and socioeconomic factors that affect our decision-making processes.
The autonomy of each individual is no longer seen as an
abstract but rather as a contextual concept.34,36
As oncologists, we consider part of our professional responsibility to acknowledge our patients needs based on
religious beliefs, family traditions, and family roles and to

involve family members, especially caregivers, in the processes of information giving and decision making offering
them support and counseling in full awareness and acceptance of differences.
Family adaptation to cancer diagnosis is continuous
across different stagesfrom hope to rebellion to despair
and family resilience, recovery, and growth may emerge
from adversity.37,38 Before encountering a major disease, a
family may have never had to negotiate a situation perceived as threatening not only to the normal life of its
members but also to the integrity of the family system. As a
consequence of such complexities, part of the intimacy
between the oncologist and his or her patient may be, or
appear, lost. We still have to learn how to best keep our
direct connection with cancer patients, while including
among our responsibilities also those to their families and
caregivers. This additional challenge for all oncologists is
part of the novel paradigm of care where professional responsibilities are no longer directed only at patients but also
at their families.
Author
Antonella Surbone*
Lea Baider*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Spillman B, Black K. Staying the Course: Trends in Family Caregiving.
Washington, DC: AARP; 2005.
2. Baider L, Cooper CL, De-Nour K, eds. Cancer and the Family: Second
Edition. Sussex, England: J Wiley & Sons, Ltd; 2000.
3. Baider L, Surbone A. Cancer and the family: The silent words of truth.
J Clin Oncol. 2010;28:1269-1272.
4. Institute of Medicine, National Academy Press. Envisioning the National Healthcare Quality Report (2001). http://books.nap.edu/openbook.
php?record_id10073&page41. Accessed Matrch 12, 2012.
5. American Academy of Family Physicians. Family, Definition of. http:
//www.aafp.org/online/en/home/policy/policies/f/familydefinitionof.html. Accessed March 12, 2012.
6. American Academy of Pediatrics (AAP) and American College of Emergency Physicians (ACEP). Patient- and family-centered care and the role of
the emergency physician providing care to a child in the emergency department. Policy Statement. Pediatrics. 2006;118:2242-2244.
7. Committee on Hospital Care and American Academy of Pediatrics.
Family-centered care and the pediatricians role. Pediatrics. 2003;112:691696.
8. Institute for Family-Centered Care. Advancing the Practice of Patientand Family-Centered Care: How to Get Started. http://www.ipfcc.org/pdf/
getting_started.pdf. Accessed March 12, 2012.
9. Institute for Family-Centered Care. Hospitals and Communities Moving
Forward with Patient- and Family-Centered Care: Enhancing Quality and
Safety for Patients and Their Families http://www.familycenteredcare.org/
resources/multimedia/index.html. Accessed March 12, 2012.
10. World Health Organization. Cancer Pain Relief and Palliative Care:
Report of a WHO Expert Committee Technical Report Series No. 804. Geneva,
Switzerland: World Health Organization; 1990.
11. Teno JM, Casey VA, Welch LC, et al. Patient-focused, family-centered
end-of-life medical care. J Pain Sympt Manag. 2001;22:738-751.
12. National Alliance for Caregiving and AARP. Caregiving in the US.
Washington, DC: AARP and NAC; 2005.
13. Soothill K, Morris SM, Thomas C, et al. The universal, situational, and
personal needs of cancer patients and their main carers. Eur J Oncol Nurs.
2003;7:5-13.
e18
14. Frank A. Relations of caring: Demoralization and remoralization in the

clinic. Int J Human Caring. 2002;6:13-19.
15. Given B, Given CW, Kozachik S. Family support in advanced cancer.
CA Can J Clin. 2001;51:213-231.
16. Yabroff KR, Kim Y. Time costs associated with informal caregiving for
cancer survivors. Cancer. 2009;115:362-4373.
17. Fegg MJ, Brandstatter M, Kramer M, Kogler M, et al. Meaning in life
in palliative care patients. J Pain Symptom Manag. 2010;40:502-509.
18. Surbone A, Baider L, Balducci L. Caregiving to elderly patients:
clinical, ethical, and psychosocial challenges. In Govindan R (ed). ASCO
Educational Book. Alexandria, VA: American Society of Clinical Oncology,
2010.
19. Baider L, Goldzweig G. Exploration of Family Care: A Multicultural
Approach. In Clinical Psycho-Oncology: An International Perspective. L
Grassi, M Riba (eds.) Chichester, West Sussex, UK: J Wiley & Sons Ltd.;
2012;187-196.
20. Grov EK, Fossa SD, Sreb O, et al. Primary caregivers of cancer
patients in the palliative phase: A path analysis of variables influencing their
burden. Soc Sci Med. 2006;63:2429-2439.
21. Baider L. Unsteady Balance: The Constraints of Informal Care. J Pediatr Hematol Oncol. 2011;33:S108-S111.
22. Surbone A. The difficult task of family caregiving in oncology: Which
roles do autonomy and gender play? Supp Care Cancer. 2003;11:617-619.
23. Kagawa-Singer M, Valdez D, Yu M, et al. Cancer, culture and health
disparities: Time to chart a new course? CA Can J Clin. 2010;60:12-39.
24. Searight HR, Gafford J. Cultural diversity at the end of life. Am Fam
Phys. 2007;72:515-522.
25. Kamen BA. Cancer pain, suffering, and spirituality: A Middle Eastern
Cancer Consortium meeting. J Pediatr Hematol Oncol. 2010;32:341.
26. Mobeireek AF, Al-Kassimi F, Al-Zahrani K, et al. Information disclosure and decision-making: The Middle East versus the Far East. J Med
Ethics. 2009;34:225-229.
27. Abdulhameed HE, Hammami MM, Hameed Mohamed EA. Disclosure
of terminal illness to patients and families: Diversity of governing codes in 14
Islamic countries. J Med Ethics. 2011;37:472-475.
28. Bulow HH, Sprung CL, Reinhart K, et al. The worlds major religions
ONCOLOGISTS DUTIES TO PATIENT AND FAMILIES

points of view on end-of-life decisions in the intensive care unit. Intensive Care
Med. 2008;34:423-430.
29. Adib SA, Hamadeh GN. Attitudes of the Lebanese public regarding
disclosure of serious illness. J Med Ethics. 2003;25:399-403.
30. Alqaissi NM, Dickerson SS. Exploring common meanings of social
support as experienced by Jordanian women with breast cancer. Cancer Nurs.
2010;33:353-361.
31. Gongal R, Vaidya P, Jha R, et al. Informing patients about cancer in
Nepal: What do people prefer? Palliat Med. 2006;20:471-476.
32. Banning M, Hafeez H, Faisal S, et al. The impact of culture and
sociological and psychological issues on Muslim patients with breast cancer in
Pakistan. Cancer Nurs. 2009;32:317-324.
33. Lamyian M, Hydarnia A, Ahmadi F, et al. Barriers to and factors
facilitating breast cancer screening among Iranian women. East Mediterr
Health J. 2007;13:1160-1169.
34. Surbone A. Telling Truth to Patients with Cancer: What is the Truth?
Lancet Oncol. 2006;7:944-950.
35. Rolland JS. Cancer and the family: An integrative model. Cancer.
2005;104:2584-2595.
36. Sherwin S. A Relational Approach to Autonomy in Health Care. In S
Sherwin, The Feminist Health Care Ethics Research Network, The Politics of
Womens Health: Exploring Agency and Autonomy. Philadelphia: Temple
University Press; 1988;19-44.
37. Vilchinsky N, Dekel R, Leibowitz, et al. Dynamics of support perception among couples coping with cardiac illness. Health Psychol. 2011;30:411419.
38. Naaman S, Radwan K, Johnson S. Coping with early breast cancer:
Couple adjustment process. Psychiatr. 2009;72:321-342.
e19
The Oncologists Duty to Provide Hope:

Fact or Fiction?
By Simon Wein, MD
Overview: There are many sources of conflict in oncology.

Conflicts arise because there are numerous therapeutic options, each of which is imperfect, and these conflicts produce
ethical dilemmas. A recent American Medical Association
(AMA) publication outlined the principles of medical ethics for
managing conflicts. Common conflicts in oncology include
whether to resuscitate, to give more chemotherapy, and how
much truth to tell. These conflicts are magnified because of
the life and death scenario of advanced cancer. Denial,
avoidance, and hope are psychologic mechanisms that enable
adaptation to the life-threatening circumstances. Hope is
widely written about though poorly understood and defined.
Ethical statements regarding its virtue and importance to
preserve are frequently given. In an effort to progress the
HE ONCOLOGIST in clinical practice is conflicted in

several domains, including truth telling, resuscitation,
when to stop palliative chemotherapy, and changing goals of
care. The conflicts are magnified because cancer is a lifethreatening illness.
Conflicts underpin ethical dilemmas. The conflicts today
are in part because of the success of modern medical oncology, which has provided many therapeutic options. There
are different models of how to solve ethical dilemmas. The
AMA recently published the American College of Physicians
(ACP) Ethics Manual, which listed four principles of medical
ethics: beneficence, nonmaleficence, respect for patient autonomy, and justice.1 These principles assist in ethically
resolving clinical conflicts.
The ACP Ethics Manual also defines professionalism. It
declares that professionalism is a code of ethics and a duty
of service that put patient care above self-interest.1 This
duty and attendant humility is another principle that encourages ethical behavior.2 It was disappointing not to see
noted in the AMA document the importance of selfawareness (or self-knowledge) as a tool for physicians to
reflect on their motives in decision making and as a further
safeguard to ethical practice.
Are We Treating Patients with Cancer
Too Aggressively?
A recent Annals of Oncology editorial asked, Why are we

not ceasing chemotherapy when it is useless, toxic, logistically complex, and expensive? This was based on statistics
that show up to 25% of patients with solid tumors receive
chemotherapy in the last month of life.3 Unfortunately,
these statistics tell only part of the story. They cannot
include those patients who did respond to chemotherapy and
whose lives were prolonged and therefore did not die within
four weeks of initiating therapy. Furthermore, since we
cannot know with certainty how each individual will respond, it is likely that we will have to treat a certain
percentin retrospect, in futilityso others will benefit.
This holds true for many therapeutic scenarios in medicine.
Nevertheless, there are instances when we recognize in
advance that chemotherapy (and parenteral nutrition, radiation therapy, and antibiotics) should not be given. It can be
harrowingly difficult to balance respect for autonomy and
e20
understanding of hope, two critical features are defined:

(1) hope as a thought process only exists in the future, and
(2) hope is only ever associated with positive and good
thoughts. The future is unknown and uncertain; therefore,
hoping can be manipulated by presenting statistics in a way
to boost hoping. Thus a dilemma and specific ethical responsibility falls on oncologists when discussing conflicts. Furthermore, since hope is a subjective assessment of a
possibility that is considered good by the hoper, it cannot
be perceived as false. False hope is an erroneous assessment. Finally, this article introduces the concept that there
might be a role to stop hopingsince hope of the future is also
filled with doubt and fearand instead live in the present and
try to find joy and meaning today.
desire for beneficence against nonmaleficence. There is a

gray zone between ethical intent and clinical outcome.
Various factors influence the decision making of the oncologist and include fear of death, medico-legal, pity, philosophy,
fear of failure, and burn-out. Sometimes the oncologist is
aware of these influences and at other times not. The
psychologic mechanisms that influence these factors include, among others, denial, avoidance, and hope. This
article will focus on hope.
What Is Hope?
There is no agreed-on definition of what constitutes hope.

There is not even agreement on whether hope is an emotion
or a cognition. Some authors describe generalized and particularized hope, while others see hope built around faith,
love, meaning, dignity, peace, and spirituality, even using
the words interchangeably.4,5 More specifically, some of
these definitions mistake the mechanism (or process) of hope
for its content or goal.6
Hope is a factor in many of the conflicts in decision
making, particularly at the end of life: parenteral nutrition,
chemotherapy, resuscitation, and truth telling (e.g., the
hope that further treatment will prolong or save a life, or
the concern that telling the truth will destroy a patients
hope and shorten life).
Daneault noted that there is a growing awareness that
nurturing hope is one of the most important tasks in the
oncology clinic.7 Helft went further saying, all patients
with cancer have a vital and inalienable right to maintain
hope.8 In the absence of a clear definition or even a
consensus on a description of hope, and with no research
measuring hope in oncologic care, such important tasks
and inalienable . . . rights are not evidence based. So how
From the Pain and Palliative Care Service, Davidoff Cancer Center, Rabin Medical
Center, Petach Tikvah, Israel.
Address reprint requests Simon Wein, MD, Pain and Palliative Care Service, Room 36,
Davidoff Cancer Center, Rabin Medical Center, 39 Jabotinsky St., Petach Tikvah, Israel,
49100; email: simonwe@clalit.org.il.
1092-9118/10/1-10
DOCTORS DUTY TO PROVIDE
do we know when, how, and if to nurture hope in the

oncology clinic?
Definitions of Hope
Dufault and Martocchio define hope as a multidimensional dynamic life force characterized by a confident yet
uncertain expectation of achieving a future good, which, to
the hoping person, is realistically possible and personally
significant.4 This definition is broad and contains the core
features of hope (the future and subjective good). However, by including multiple dimensions and dynamic life
force in the definition, researchers tend to apply attributes
such as faith, relationships and dignity to the definition
itself. These life forces support hope but are not hope.
Snyder defines hope based on empirical research and a
validated (his own) hope scale.9 His work has not been
extensively applied to oncology or palliative care. Snyder
firmly holds the view that hope is a cognitive and not an
emotional process. He defines hope as the perceived capability to produce workable routes to desired goals and the
requisite motivation to use those routes.9 That is, there is a
goal, a plan to get there, and the will to do it. Hope is a
consciously planned process, with emotions as an epiphenomenon. Meaning, faith, and love are adjuvants or struts to
hope and influence Snyders motivation and will to but
are not themselves integral to the definition of hope. Although Snyders definition is logical and measurable, it is
not easy to apply at the bedside.
Is Hope Good or Evil?
The Greek myth of Pandora involves a woman who had a

jar and was told not to open it. Curiosity got the better of
her, and upon opening the jar, various evils escaped: plague
and disease. She quickly shut the lid, but one thing remained trapped inside: hope.10 The ancient Greeks interpreted hope as evil: Precisely because of its ability to keep
the unfortunate in continual suspense, the Greeks considered hope the evil of evils, the truly insidious evil.11 That is,
a person trapped in an unwinnable situation who still hopes
to escape, merely tortures himself in the futility of the task.
The parallel to incurable cancer and another round of
chemotherapy given in the hope is compelling.
However, later in history, Judeo-Christian and Western
philosophy associated life with an ultimate purpose and
goal. Hope became identified with good things. Menninger
said, Love, faith, hopein that order. The Greeks were
wrong. Of course hope is real, and of course it is not evil. It
is the enemy of evil, and an ally of love. Which is goodness.5
A Practical Description of Hope
The critical characteristic of hope is that it exists only in

the future. We cannot hope in the past or the present. The
outcome of hoping is subjectively assessed as a probability

but neither a certainty nor an impossibility. It is this
characteristic that is important to bedside decision making.
The second immutable characteristic is that hope is egosyntonic. That is, whenever we hope, it makes us feel good
and optimistic, not sad, distressed, or anxious. Even if we
hope something bad for ourselves (e.g., death), we note that
its primary motivation is good.
The key feature about the future is that it is unpredictable. Since hope exists only in the future, it too is unpredictable and unknowable. This is what Boris meant when he
said that we create ambiguity in order to preserve hope.12
That is, by creating uncertain situations, we also create an
opportunity to hope. Into uncertainty we can hope. This is
the principle of gambling. An uncertainty is deliberately
created and people visit casinos not for the fun but for the
possibility to hope and the good feeling it creates.
Similarly, when we offer chemotherapy and the outcome is
uncertain (as it must be), this ambiguity or uncertainty of a
future outcome allows the patient to hope. A physician may
be deliberately vague about the possible benefit of a new
treatment or about the result of a scan, thus manipulating
hope. The physician does not create hope, nor can the
physician prescribe hope as such. More precisely, the physician amplifies an ambiguity about the future to which the
patient may or may not respond.
Husebo, however, states that hope is not just something
which is limited to an expectation of the future or afterlife.
Hope is closely connected to a persons life story.13 Of course
hope is connected to a persons values and history. Snyder
noted this is what determines the content of the goals of
hope and influences the determination to succeed. However,
hopethe verb and the noun exists only in the future and
in the imagination. Hope is a figment of the imagination.
How can we understand hope in terms of good and evil?
The Greeks thought hope was evil, whereas Western society
sees hope as good. The appreciation that hope itself is good
exists, in part, because the goals of hope are always perceived as subjectively good. The ancient Greeks were concerned about this characteristic of goodness being
manipulated and abused.
A more accurate description was furbished by Kertesz in
his acceptance speech for the 2002 Nobel Prize for Literature. He described how after invading Hungary, the Russians controlled the people by manipulating hope and
similarly the Germans manipulated the inmates of the
concentration camps. Kertesz stated, I understood that
hope is an instrument of evil.14 Hope itself is not evil, since
hope is a path and goal chosen by each individual in the hope
of a good outcome. Neither, however, is hope necessarily
beneficial to the hoper.
False Hope
KEY POINTS
Clinical conflicts create ethical dilemmas.

Hope exists only in the future.
Hope may cause more harm than good.
False hope is a self-contradiction.
There is benefit in living without hope.
Strictly speaking, hope can never be false since it is a

subjective thought process whose primary function is to plan
and achieve goals. The future is unknown, therefore, the
subjective evaluation of a possibility can be contracted or
expanded depending on psychologic needs. Thus, a 5%
chance of response to chemotherapy may generate hope for
some but appear futile to others. Similarly, after being told
there is 0% chance of being cured, people who believe in
miracles will still be able to hope. A miracle may be false to
e21
SIMON WEIN
an external observer, but to the person hoping for a miracle,

it is true in all respects and realistically possible.4
A common understanding of false hope is a patient who
mis-states the probability of a goal and hence harbors false
hopes. It is common for patients to misunderstand statistics.
This process is better labelled as denial or on occasions as a
delusion. It is not uncommon for oncologists to be a little
vague and ambiguous in communicating prognosis and statistics in order to encourage hope, as Boris taught us.9
Furthermore since hope exists in the unknowable future it
cannot be disproven and therefore it cannot be shown to be
false. For hope to function, it does not matter whether the
statistics are true or false, or whether the patient misconstrued them. All that it matters is for the patient to believe
them. Hope will then thrive, for better or for worse.
Denial as an Elixir of Hope
Daneault observed that false hope, when it denies the

reality of imminent death, may lead to a neglect of final
arrangements and added burden for families, while Von
Roenn categorically stated that false hope leads to disappointment and disillusionment.7,15 It might be more pragmatic to approach such hoping as we would denial and not
to judge whether hope is false or true. If denial is not
causing harm and is functioning as a psychological defense,
then there is no need to intervene. Some people hope until
the very end and thereby preserve their emotional wellbeing. Others, by hoping against hope, create problems
evil as the Greeks would have itthat cause harm or
damage. It would be appropriate to consider counseling this
latter group, similar to those where denial is causing problems.
Denial is the gap between a subjective and objective
assessment of a medical situation. Into this gap the patient
can develop new goals and by reducing anxiety, can improve
motivation. Clayton et al performed a systematic review of
hope in terminally ill patients and showed that some patients and caregivers found hope in avoiding the facts, that
is, denial.16 The practical question to ask is: is hope functioning in a beneficial or maladaptive way?
Life Without Hope
Boris observed, Indeed, there is no torment quite like

hopelessness. But hopelessness marks the presence of
thwarted hope, not hopes absence. This fine distinction
points the way to a therapeutic paradigm. A hope that has
failed (i.e., hopelessness) causes frustration and pain. However, if one does not hope for anything, then one will not
experience the anguish of frustrated hope. Not engaging in
hoping means one cannot lose or, indeed, benefit from hope.
Fear is also an expectation of the future. Fear in contrast
to hope creates negative feelings and anxiety. It is in some
ways an opposite to hope. If one were to disregard the future
and just live in the here and now, then there would be no
hope or fear as Gravlee observed, The one with no hope also
does not fear. Fear about the uncertain future remains, so
long as hope about the future remains.17 We are anxious
that our hope our gamble in the futurewill fail. Huxley
said the same thing, though used drugs (soma) to stop
worrying about what might happen: Was and will make me
ill, I can take a gramme and only am.18 Without memories
there will be no anxiety about something going wrong in
e22
the future. Without a future there will be no hope or fear

that the hope might fail. This fear-hope dyad is reflected in
the ancient Greeks jaundiced view of hope. By not hoping
(clearly in some cases this is a form of denial), there will no
longer be the merry-go-round of hope and despair.
Hope in Oncology
In applying hope to clinical oncology a number of scenarios

arise:
Should we give antitumor treatments to boost hope?
How ethically bound are we to prescribe chemotherapy
because a patient hopes it will work?
How prepared are we to say no to further treatment at
the risk of the patient going elsewhere?
Is there a role for placebo to maintain hope?
Does telling the truth diminish hope?
Does loss of hope shorten life?
How do oncologists stop their own hopes (and fears) from
influencing the decision?
Three common strategies can be used when hope has
failed in advanced cancer: to live today, to exchange one
hope for another, and to reflect about the life lived. It is
sometimes useful to discuss the nature of hope with the
patient and how it influences decision making, especially if
hope is driving the patient to unrealistic decisions or is
causing psychologic symptoms of distress. A common approach is to suggest to the patient to live day to day and to
enjoy the moment. Leave tomorrow to the future. Of course
what we are doing is replacing the future (hope and fear)
with the present. If the process of hoping cannot be relinquishedif the perceived subjective good is too comforting to
give upyet is starting to cause harm then Snyders hope
model allows lost goals to be mourned and replaced with new
ones. Instead of hoping for a cure, the goal might be switched
to hoping to see a daughter married. Thus hoping in the
future comforts the living in the present and pushes acceptance aside.
Finally, a simple and effective strategy, the life narrative, is to talk to the patient about their lives and loves. This
starts the process of replacing hope (in the future) with
acceptance (in the present). Roy valued the physician who
know(s) how to drop the professional role mask and relate to
others simply and richly as a human being.19
Conclusion
Our professional duty demands not only an adherence to

the AMA ethical guidelines but also an effort to be aware of
the psychologic processes involved in decision making both
the patients and ours. Denial, avoidance, and hope are
psychologic mechanisms used to avoid conflict. Hope has two
core characteristics: it is a thought process that can exist
only in the future and it is hardwired to make us feel good.
Hope is neither inherently good nor bad. The outcome of
hope is neither necessarily good nor bad.
False hope does not exist in the sense that each time we
hope, we do so subjectively for our own psychologic benefit. If
an external observer dismisses the likelihood of my hope
materializing, this does not make my hope any less realistic
or significant for me. Nor does it matter if the patient
him/herself misunderstands the prognosis and build their
hopes on a false premise. This is a form of denial and if it
helps the patient overcome the stress of the illness without
causing harm, then it is not false hope.
DOCTORS DUTY TO PROVIDE
The value of studying hope in oncology and palliative care

is to understand the role it plays in psychologic health, in
making decisions about treatment, and in establishing goals
of care.20
Talking about death and hope from the day of diagnosis
means that when the disease has run its course, the discussion will be the continuation of a conversation, not a spinechilling revelation. The responsibility for presenting medical
information in a way that enables the patient to balance
acceptance against hope lies with the treating oncologist.
However, the oncologist is not obligated to prescribe hope

and the oncologist can do better than collude or be disingenuous to reinforce hope.
There are other values in lifefamily, work, hobbies,
dignity, achievementsthat the oncologist can explore a
little at each visit. Discussion about these values may (or
may not) enable acceptance of death. However, as Breitbart
noted, The paradox of the end of life dynamic is that
through acceptance of the life one has lived comes acceptance of death.21
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Simon Wein*
REFERENCES
1. Snyder L. American College of Physicians Ethics Manual, 6th ed. Ann
Intern Med. 2012;156:73-104.
2. Pellegrino ED. Professionalism, Profession and the Virtues of the Good
Physician. Mt Sinai J Med. 2002;69:378-384.
3. Braga S. Why do our patients get chemotherapy until the end of life?
Annal Oncol. 2011;22:2345-2348.
4. Dufault K, Martocchio BC. Hope: Its spheres and dimensions. Nurs Clin
North Am. 1985;20:379-391.
5. Menninger K. The Academic Lecture: Hope. Am J Psychiatry. 1959;116:
481-491.
6. Nekolaichuk CL, Bruera E. On the nature of hope in palliative care.
J Palliat Care. 1998;14:36-42.
7. Daneault S, Dion D, Sicotte C, et al. Hope and noncurative chemotherapies: Which affects the other? J Clin Oncol. 2010;28:2310-2313.
8. Helft PR. Necessary collusion: Prognostic communication with advanced
cancer patients. J Clin Oncol. 2005;23:3146-3150.
9. Gum A, Snyder CR. Coping with terminal illness: The role of hopeful
thinking. J Palliat Med. 2002;5:883-894.
10. Wikipedia. Pandora. http://en.wikipedia.org/wiki/Pandora. Accessed
January 31, 2012.
11. Hollingdale RJ (ed). The Nietzsche Reader. London: Penguin Classics,

1977;277.
12. Boris HN. On hope: Its nature and psychotherapy. Int R Psycho-Anal.
1976;3:139-150.
13. Husebo S. Is there hope, doctor? J Palliat Care. 1988;14:43-48.
14. Kertesz I. Nobel Lecture, Nobel Prize in Literature 2002. http://
www.nobelprize.org/nobel_prizes/literature/laureates/2002/kertesz-lecture-e.
html. Accessed January 12, 2012.
15. Von Roenn JH, von Gunten CF. Setting goals to maintain hope. J Clin
Oncol. 2003;21:570-574.
16. Clayton J, Hancock K, Parker S, et al. Sustaining hope when communicating with terminally ill patients and their families: A systematic review.
Psychooncology. 2008;17:641-659.
17. Gravlee GS. Aristotle on Hope. J Hist Philos. 2000;38:461-477.
18. Huxley A. Brave New World. New York: Perennial Classics, 1998.
19. Roy D. Dying with dignity. J Palliat Care. 1986;2:3-4.
20. Wein S. Hope: Concerning structure and function. Palliat Support
Care. 2004;2:229-230.
21. Breitbart W. Thoughts on the goals of psychosocial palliative care.
Palliat Support Care. 2008;6:211-212.
e23
MEDICAL ERRORS IN CANCER CARE: PREVENTION,

DISCLOSURE, AND PATIENT AND
FAMILY MEMBER RESPONSES
CHAIR
New York University
New York, NY
SPEAKERS
Itzhak Brook, MD, MSc
Georgetown University
Washington, DC
Michael Rowe, PhD
Yale School of Medicine, Program Recovery and Community Health
New Haven, CT
O n c l o g i s t s D i f fi c u l t i e s i n F a c i n g a n d
Disclosing Medical Errors: Suggestions for
the Clinic
By Antonella Surbone, MD, PhD
Overview: Along with improved safety measures and changes

in the culture of medicine, communication is key to reducing
the effect of medical errors and to easing the medical,
psychologic, and existential burdens they impose on all parties. Disclosure demonstrates respect for patients autonomy
and promotes patients involvement in informed decision
making about ways to correct or alleviate the effects of the
error. It also enhances oncologists integrity and helps restore
trust in the patient-doctor relationship.
Because of the complexity of cancer treatments and the
uncertainty regarding outcomes in oncology, oncologists may
rationalize nondisclosure as a way to avoid adding to the
physical and existential suffering of their patients. Although
there is broad agreement among professional and regulatory
bodies, as well as medical ethicists, that physicians should
disclose errors to patientsand physicians largely support
N RECENT years, research on the incidence and causes

of error in medicine has led to calls for change in policy
and practice.1-2 Increasing attention has been paid to the
role that appropriate communication plays in preventing
and managing errors, not only at the medical level, but also
with regard to the emotional and psychologic aftermath of
error for patients, family members and physicians. The
limited research that has been conducted on the psychologic
consequences of medical errors on physicians shows that
they experience a wide range of emotions and thoughts
from guilt, feelings of inadequacy, anguish, shame (even to
the point of leaving the profession) to excessive caution
toward other patients following the medical error.3 Most of
what we know on this subject, however, comes from narrative accounts of physicians who have been brave enough to
share with readers their experiences of committing or witnessing a medical errorthe first starting with Dr. David
Hilfiker in The New England Journal of Medicine in 1984.4
Communication about medical errors is key to easing the
burden of errors on all parties with respect to all medical,
psychologic, and existential dimensions. Given the lifethreatening nature of their illnesses and their dependence
on their oncologists, patients with cancer can be especially
hurt both physically and emotionally when errors or suspected errors occur. Oncologists who carry the emotional
and psychologic burdens of caring for seriously ill patients
over long periods of time can also be deeply affected by their
own mistakes.5 Proper disclosure, including a sincere apology, is the first step to restore trust in the patient-doctor
relationship. It should be part of the immediate management of errors and of dealing with their long-term aftermaths.
Disclosure of Medical Errors and the
Patient-Doctor Relationship
There is broad agreement among professional and regulatory bodies and medical ethicists that physicians should
disclose errors to their patients.2,6,7 Empirical studies show
that most patients want to be informed in detail about
e24
disclosure of error to patientsstudies show discrepancy

between physicians responses to hypothetical clinical scenarios of truth telling about medical errors and actual practices of withholding or tempering the truth. Among common
reasons for avoiding disclosure are risk of malpractice lawsuits, fear of being exposed as incompetent, and feeling
shame before patients and colleagues.
Proper disclosure, however, including a sincere apology,
should be part of the management of errors and of their
long-term aftermaths. In disclosing medical errors, it is essential for oncologists to pay equal attention to the medical
and the emotional aspects of the information they are giving
and the reaction that it elicits in patients and families. Specific
communication skills regarding disclosure of medical errors
can be learned.
medical errors that occur during their care, even when they
do not lead to a negative outcome.8 Disclosure demonstrates
respect for patients autonomy and promotes their informed
decision making about ways to correct or alleviate the effects
of the error.5,7
Physicians largely support disclosure of error to patients,
but there is discrepancy between responses to hypothetical
clinical scenarios of medical errors, in which doctors support
disclosure, and practice, in which they often do not provide
full disclosure.9 Among common reasons for avoiding
disclosure are fear of malpractice lawsuits or of being
exposed as incompetent, as well as feelings of shame before
colleagues.9-13 A pivotal study comparing 1,404 surgeons
and general physicians practicing in Canada to 1,233 practicing in the United States showed that, even in a different
medical-malpractice environment, a misleading culture of
medicine that privileges technical perfectionism and success
rates over the humanist aspect of care in medical school,
rather than fear of being sued, is the prime deterrent to full
disclosure.14,15 These attitudes toward the medical profession are mostly transmitted through hidden curricula, which
push doctors away from talking about their mistakes for fear
of peer judgment or of punishment by senior staff.
Retrospective studies suggest that patients and family
members often sue because of doctors silence following
adverse events and the accompanying sense that they have
no other recourse for gaining respect, acknowledgment, and
satisfaction.16 By contrast, patients and families reasons
not to sue after medical errors are far less known.
From the Department of Medicine, New York University Medical School, New York, NY.
Address reprint requests Antonella Surbone, MD, PhD, New York University Medical
School, 5550 First Avenue, RBCD516, New York, NY 10016; e-mail: antonella.surbone@
gmail.com.
1092-9118/10/1-10
DISCLOSING MEDICAL ERRORS
Disclosure in oncology, as in other fields of medicine,

follows directly from the fiduciary nature of the relationship.5,11,17,18 The relationship between patients with cancer
and their oncologists occurs in a particular context of uncertainty about the evolution of the illness and prospects for
treatment effectiveness that affects both parties and their
reciprocal communication.18,19 This uncertainty also contributes to burnout among oncologists.20 In addition, the
close relationships they may develop with their patients can
make committing and disclosing error more difficult for
them to bear. At the same time, severe illness and imminent
mortality among their patients may tempt them to withhold
information about error or adverse events.5
Even if barriers to disclosure could be reduced, the use of
experimental anticancer protocols and multiple medications often carrying a high toxicityalong with the multidisciplinary nature of most cancer treatments, can render
more difficult the oncologists assessment of whether an
adverse event is a side effect of treatment or the result of
medical error. Given all these factors, oncologists may be
tempted to rationalize away some errors and question the
need to disclose them to patients.5
Questions of error, disclosure, and living with error in
oncology should be viewed in the context of truth telling to
patients with cancer. On one hand, because of the seriousness of their illnesses, patients with cancer may be particularly vulnerable to the inherent power asymmetry between
them and their oncologists. On the other hand, oncologists
often must deliver devastating news to their patients, and
the manner and content of their delivery can mean the
difference between mere fact giving and truth telling.18
They may hide behind a torrent of medical information,
which, although factually correct, overwhelms and confuses
their patients. On the other hand, however, oncologists can
provide accurate and complete information to their patients,
while still supporting their sense of hope through assurance
that they will be there for them through the difficult paths
toward cure or palliative care.
KEY POINTS
The two most effective ways to reduce the effect of

medical errors are system management and improved communication.
Oncologists may experience psychologic and emotional difficulties, such us anguish, guilt, and shame,
associated with acknowledging and disclosing a
harmful medical error.
Errors should be communicated openly and clearly,
and a sincere apology should be offered to patients
and their loved ones.
An apology is essential to restoring trust in the
patient-doctor relationship and initiating a healing
process for all parties in the aftermath of a medical
error.
Making a shift toward increased disclosure requires
humility, accountability and openness among oncologists, as well as communication skills that can be
learned.
The reality of each patients illness evolves in unique ways

under the influence of many interrelated, contextual variables, including the patient-doctor relationship.18 In disclosing to patients the seriousness of their cancer diagnosis and
formulating a prognosis, oncologists can draw on the creative power of the reciprocal trust involved in their relationship, as much as focusing on the facts of the case. Patients
with cancer and their families can thus know that honest
dialogue and true cooperation will take place throughout the
course of the illness and feel reassured that they will not be
left alone to cope with the hardest truths, including those
about medical errors.21
Disclosure of Medical Errors in Oncology
The two most effective ways to reduce the effect of medical

errors are system management and improved communication.22 Telling the truth to patients about medical errors
involves psychologic-emotional elements, in addition to professional and legal ones. For patients, the physical effect
of error and the shock of learning about it may be compounded by grief, loss, and a sense of isolation. Furthermore,
patients may be forced to live with the possible long-term
physical effects of error and the stigma of reduced family,
social, and occupational roles sometimes associated with
illness.5 Finally, the effect on grieving families of knowing or
suspecting that a loved one with an incurable illness has
suffered or died from a medical error is potentially devastating, and thus requires special consideration in regard to
disclosure.21
As in all clinical communication, content, setting, and
emotional support are important.23 The late Dr. Robert
Buchman, a major expert in communication with patients
with cancer summarized in the acronym CONES the essential elements of a discussion involving a medical error: C for
choosing the appropriate context for carrying a difficult
conversation; O for the opening statement that alerts the
patient to a difficult conversation; N for a proper narrative
to describe the error, starting with plain words such as I
found out that; E for acknowledging and addressing emotions in a direct empathic way by stating upfront, it is very
upsetting for you and it is awful for me too; and S for
strategy and summary.24
In disclosing medical errors it is essential for oncologists
to attend to both medical and emotional aspects of the
information provided and the reactions they may elicit from
patients and families. The physicians response must be
clear at the clinical level and also address the emotional
needs of patients and family members. In addition, in
disclosing medical errors and discussing their causes and
actual or potential consequences, oncologists should also
share their own emotions. Four different levels of response
direct, escalationary, exploratory and empathic have been
identified. Although the first three predominantly address
the factual aspects of the situation, only an empathic response has the potential to result in effective communication
about something so painful and upsetting as a medical
error.24
Culled from practical guidelines developed by institutional bodies, published research, and clinical experience,
Table 1 lists suggestions for oncologists facing the difficult
task of revealing to patients or family members that a
medical error has occurred.25-27
e25
ANTONELLA SURBONE
Table 1. Key Steps in Disclosing Medical Errors to Patients
and Relatives
Initiate the conversation in a proper setting.

Start by saying that the conversation will be difficult and then give an explicit
statement about what happened.
Talk and respond with humility and sorrow, not with defensiveness.
Offer a sincere clear apology, including admission of the error and remorse.
If you are still in training, ask for the presence of a senior staff member.
Refrain from pointing to someone else or to the system.
Disclose relevant data and take responsibility.
Promise that a thorough investigation of the error will be conducted.
Make sure that precise and timely answers are provided.
Guarantee constant monitoring of the patient as your priority.
While keeping an open attitude about possible good outcomes, avoid making
promises you cant keep, yet state that you hope for the best outcome.
Set up a follow-up meeting with family members, including anyone else they wish
to involve.
Offer the help of a patient advocate and legal consultant to assist patient and
relatives.
Disclosure of Medical Errors: Individual and

Cultural Aspects
According to the Institute of Medicine (IOM), the standard

definition of medical errors is the failure of a planned action
to be completed as intended or the use of a wrong plan to
achieve an aim.1 Most patients, however, have a broader
perspective on medical errors than physicians or researchers.5 For example, they tend to include among medical
errors their physicians failure to communicate effectively
with them before and after an error has occurred. Individual
physician or institutional arrogance has been described by
some patients with cancer as contributing to their perception of medical errors and to the possibility of repairing a
breach in trust.28,29 Furthermore, the absence of physician
empathy and honesty in the aftermath of harmful medical
error not only exacerbates patients and family members
suffering but may be seen as part of the error itself.
Disclosure of medical errors is now an ethical requirement
of the medical profession in the Western world. It cannot be
ignored, however, that disclosing an error is inevitably
related to the different attitudes and practices of providing
information to patients with cancer in different countries
and cultural groups. Clearly, it is considered more difficult,
and it is less common to disclose medical errors in contexts
where the truth about their overall condition and prognosis
is withheld from patients with cancer.
A recent study of surgeon-patient disclosure practices in
Southwestern Nigeria asked 102 surgeons the question: To
whom should surgical errors be disclosed? Only 15.7%
replied that the information should be conveyed to the
patient, whereas 30% indicated that it should be delivered to
a family member. Others suggested that errors should be
referred to the hospital ethics committee (22.5%), to hospital
management staff (21.5%), to a colleague (21.5%), or to other
staff members during a clinical meeting (2%).30
Overall, the authors of the study reported that surgeons
were unsure about whether they should disclose medical
errors.30 Notably though, participating surgeons valued the
potential benefits of disclosure in a similar way to that of
Western physicians, listing among those benefits: learning
e26
from errors; prevention of further complications; promotion

of honesty, openness, trust and confidence; reduction of risk
of litigation; relieving the doctors conscience; and overall
practice improvement. Among the potential risks of disclosing medical errors they noted litigation and prosecution,
negative effect on practice, loss of trust and confidence, loss
of medical license, undue patient anxiety, and even risk of
physical assault.30
Conclusion
Oncologists failure to disclose medical errors to their

patients betrays the fiduciary nature of the patient-doctor
relationship and diminishes the integrity of our profession
by silencing patients and family members through neglect of
their stories and experiences, as well as by obfuscating the
effect that errors have on ourselves.5 Silence may increase
the risk of malpractice suits and is always morally wrong,
for it inevitably adds to the pain of patients and their loved
ones.21
Further research on the incidence and types of error, near
errors, and perceived errors in oncology, using both qualitative and quantitative methods, and development of clear
policies and practices regarding disclosure and other posterror interactions with patients and family members, are
needed. Understanding that, like truth-telling about cancer
diagnosis and prognosis, attitudes toward and practices of
disclosure of medical errors are subject to individual and
cultural variability and should not prevent us from learning
about sensitive communication about medical errors and
from meeting adequate standards of ethical clinical practice
worldwide. Analysis of institutional standards, policies, procedures, and training regarding disclosure of medical errors
with a focus on how oncologists and other team members
understand and apply these specific communication elements will guide researchers and practitioners in efforts to
build on current practice.
Mortality and Morbidity Conferences are often a proper
setting for physicians to disclose and discuss their errors,
but they do not provide the setting for expressing emotions
and finding support. Didactic and experiential education
and training regarding responsibility and accountability for
optimal ongoing communication throughout the course of
the patients illness, including disclosure of medical errors,
should be enhanced at all levels, from medical school
through internship, to residency and oncology fellowship.
Emphasis should be placed on the redemptive value of
apology and forgiveness for patients and their families and
for oncologists, and on provision of emotional support for all
parties affected by medical errors.
Many compelling physicians narratives reporting the experience of medical errors are available to teach us that
communication of medical errors is not only about disclosure
to patients and families. Rather, it is also about creating a
culture of medicine in which the patients interest and
well-being always has priority over physician or institutional self-interest. This principle and goal can be achieved
through establishing a serene atmosphere of collaboration,
rather than competition, and open communication among all
members of the staff, one in which a medical error is not a
reason to fear punishment or loss of esteem from peers, but
an incentive for all to share, understand, and improve.
DISCLOSING MEDICAL ERRORS
Author
Antonella Surbone*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Kohn LT, Corrigan J, Donaldson MS. To err is human: building a safer
health system. Institute of Medicine (IOM) Washington, D.C.: National
Academy Press; 2000.
2. Ethics manual, American College of Physicians. 6th Edition. http://
www.acponline.org/running_practice/ethics/manual.htm. Accessed February
15, 2012.
3. Rowe M. Doctors responses to medical errors. Critical Reviews in
Oncology/Hematology. 2004;52:147-163.
4. Hilfiker D. Facing our mistakes. N Engl J Med. 1984;310:318-322.
5. Surbone A, Rowe M, Gallagher TH. Confronting medical errors in
oncology and disclosing them to cancer patients. J Clin Oncol. 2007;25:14631467.
6. Finkelstein D, Wu A, Holtzman N, et al. When a physician harms a
patient by a medical error: ethical, legal, and risk-management considerations. J Clin Ethics. 1997;8:330-335.
7. Gallagher TH, Suddert D, Levinson L. Disclosing harmful errors to
patients. N Engl J Med. 2007;356:2713-2719.
8. Duclos CW, Eichler M, Taylor L, et al. Patient perspectives of patientprovider communication after adverse events. Intl J Qual Health Care.
2005;17:479-486.
9. Gallagher TH, Waterman AD, Ebers AG, et al. Patients and physicians
attitudes regarding the disclosure of medical errors. JAMA. 2003;289:10011007.
10. Vincent C, Stanhope N, Crowley-Murphy M. Reasons for not reporting
adverse events: an empirical study. J Eval Clin Pract. 1999;5:13-21.
11. Beckman HB, Markakis KM, Suchman AL, et al. The doctor-patient
relationship and malpractice. Lessons from plaintiffs depositions. Arch Intern Med. 1994;154:1365-1370.
12. Hickson GB, Federspiel CF, Pichert JW, et al. Patient complaints and
malpractice risk. JAMA. 2002;287:2951-2957.
13. Mazor KM, Simon SR, Gurwitz JH. Communicating with patients
about medical errors: a review of the literature. Arch Intern Med 2004;164:
1690-1697.
14. Gallagher TH, Garbutt JM, Waterman AD, et al. Choosing your words
carefully: how physicians would disclose harmful medical errors to patients.
Arch Intern Med. 2006;166:1585-1593.
15. Mahhod SC. Medical education. beware the hidden curriculum. Can
Fam Phys. 2011;57:983-985.
16. Vidmar N. Medical malpractice lawsuits: an essay on patient interests,
the contingency fee system, juries, and social policy. Loy L. Law Rev. 2005; 38;
1217-1266.
17. Pellegrino ED, Thomasma DC. For the Patients Good. The Restoration
of Beneficence in Health Care. London: Oxford University Press; 1988.
18. Surbone A. Telling the truth to patients with cancer: what is the truth?
Lancet Oncol. 2006;7:944-950.
19. Schapira L. Shared uncertainty. J Support Oncol. 2004;2:14-18.
20. Penson RT, Dignan FL, Canellos GP, et al. Burnout: caring for the
caregivers. Oncologist. 2000;5:425-434.
21. Rowe M. The rest is silence: hospitals and doctors should beware of
what can fill the space of their silence after a loved ones death. Health Aff.
2002;21:232-236.
22. Penson RT, Svendsen SS, Chabenr BA, et al. Medical mistakes: a
workshop on personal perspectives. Oncologist. 2001;6:92-99.
23. Parker PA, Baile WF, de Moor C, et al. Breaking bad news about
cancer: patients preferences for communication. J Clin Oncol. 2001;19:20492056.
24. Buchman R. Interpersonal Communication And Relationship Enhancement: I*CARE. http://www.mdanderson.org/education-and-research/
resources-for-professionals/professional-educational-resources/i-care/index.
html. Accessed February 15, 2012.
25. Fallowfield L. Communication with patients after errors. J Heath Serv
Res. 2010;15:56-59.
26. Evans SB,Decker R. Disclosing medical errors: a practical guide and
discussion of radiation oncology-specific controversies. Int J Radiat Oncol Biol
Phys. 2011;5:1285-1288.
27. OConnell D, Kemp White M, Platt FW. Disclosing Unanticipated
Outcomes and Medical Errors. JCOM. 2003;10:25-29.
28. Ingelfinker FJ: Arrogance. N Engl J Med. 1980;303:1507-1511.
29. Brook I. A physicians experience as a cancer of the neck patient. Surg
Oncol. 2010;19:188-192.
30. Ogundiran TO, Adebamowo CA. Surgeon-Patient information disclosure practices in Southwestern Nigeria. Med Princ Pract. Epub 2011 Nov 23.
e27
Preventing Errors in Oncology: Perspective

of a Physician Who Is Also a Cancer Patient
By Itzhak Brook, MD, MSc
Overview: This article presents my personal experiences as a

physician who underwent laryngectomy for hypopharyngeal
squamous cell carcinoma. I describe the numerous medical
and surgical errors that occurred during my hospitalization at
RRORS IN patient care are very common, occurring in

up to 40% of patients who undergo surgery. These
errors can cause complications in up to 18% of these patients.1,2 These errors often lead to malpractice lawsuits,
increase expense of medical care, extend hospital stays, and
can lead to greater morbidity and mortality.3
I have been a physician specializing in infectious diseases
in the hospital setting for 40 years. I was not aware of how
frequent medical errors were until I became a patient with
cancer myself following the diagnosis of throat malignancy
(hypopharyngeal carcinoma). For the first time, I had to deal
with medical and surgical errors in my care as a patient and
not as a physician.
My primary small hypopharyngeal cancer (T1, N0, M0)
was surgically removed, and I received local radiation;
however, after 20 months, a local recurrence at a different
location (pyriform sinus) was found (T2, N0, M0). Even
though they tried three times, the surgeons failed to completely remove the cancer using laser. I then underwent
complete pharynolaryngectomy with flap reconstruction at a
different medical center that had greater experience with
this type of tumor. The cancer was completely removed, and
no local or systemic recurrence was noted.
I was treated at three large academic medical centers (two
military and one civilian) located at a major metropolitan
center. Even though the overall medical care I received at all
of these hospitals was adequate, I very quickly learned that
medical errors occurred at all levels of my care. Despite
these errors, I am grateful to the medical staff, including
physicians, nurses, and other health care providers who
cared for me.
This article describes the medical errors I experienced
during my hospitalizations at three hospitals and their
impact on me. My inability to speak following my laryngectomy made it difficult for me to prevent many of these errors.
Fortunately, I was able to prevent many of these mistakes,
but not all of them.
Delay in Diagnosing Cancer Recurrence
My surgeons did not discover the recurrence of my tumor

in a timely fashion, although they endoscopically examined
me on a monthly basis following my initial surgery. This
happened despite the fact that I had been complaining of
sharp and persistent pain in the right side of my throat for
7 months. My surgeons reassured me that the pain was most
likely caused by irritation because of the reflux of stomach
acid. My physician increased the acid-reducing medication I
was taking, but the pain did not disappear.
The return of cancer was finally discovered by an astute
resident who was the first examiner to perform an endoscopic examination using a Valsalva maneuver (exhaling
and holding the air). This technique allows visualization of
e28
three different medical centers. It is my hope that my presentation will contribute to the reduction of such errors and lead
to a safer hospital environment for patients.
the pyriform sinus where the tumor was found. In retrospect, I was wonder why the experienced head and neck
surgeons who always examined me failed to perform such a
basic procedure before. If they had done it earlier, the cancer
recurrence (which was 4 by 2 cm) would have most likely
been seen and removed at an earlier time, thus reducing the
chance for the spread and need for laryngectomy.
Only three weeks earlier, I was also examined endoscopically by a radiation oncologist who had seen no abnormality. This physician confessed to me later that he did not look
down into the area where the new cancer was located
because his instrument broke during the procedure.
Failure to Excise the Recurrent Tumor Using Laser
The first mistake during my admission for surgery was

when my surgeons, using laser technology, mistakenly removed scar tissue instead of the cancerous lesion. The tumor
was located further down in the pyriform sinus. It took a
week before they realized that an error was made when the
pathological studies showed that the excised material was
scar tissue and not the lesion. This mistake could have been
prevented if frozen sections of the suspected lesion, rather
than just the margins of the pyriform sinus, had been
studied in the operating room. As a result of this mistake, I
had to undergo an additional surgical procedure with laser
10 days later to remove the cancerous lesion.
Following the second surgery, my surgeons informed me
they were able to remove the cancer in its entirety by using
the laser, and all the margins of the removed area were clear
of cancer. I was therefore spared from a more extensive
surgery, which would have included removal of tissues in
the right side of my neck, requiring their replacement by
tissue transplanted from my thighs or shoulder areas. I felt
great relief when I heard the news and felt very fortunate.
Even though the bulk of the tumor was finally removed, the
surgeons could not get all of it out. I had to undergo two
additional attempts to remove the cancer.
I later realized that the failed attempt to remove the
cancer on the first surgery made any repeated attempt more
difficult because each surgery induces extensive local swelling and inflammation. This was especially true in my case
because my cancer was located at a very narrow and inaccessible location. This made any follow-up interventions
From the Department of Pediatrics, Georgetown University School of Medicine, Washington, DC.
Address reprint requests to Itzhak Brook, MD, MSc, 4431 Albemarle St. NW, Washington
DC 20016; email: ib6@georgetown.edu.
1092-9118/10/1-10
PREVENTING MEDICAL ERRORS
more difficult because insertion of an endoscope and visualization of the area were more difficult.
I learned the hard way that experience is of primary
importance with this kind of surgery. Since the frequency of
throat cancer has declined in the United States, there are
fewer patients with this cancer and surgeons have less
experience removing it. With fewer patients, the expertise in
its removal and care is concentrated in fewer institutions.
Even though my surgeons had very little experience using
laser to remove my type of cancer, they still offered to
remove it using this technique. One of my surgeons summarized his philosophy about learning new procedures as: See
one, do one, teach one. I believe the best approach should
be: See a hundred, do 200, teach one.
Although the error made by my surgeons was very regrettable, they admitted their responsibility for making it. This
made it easier for me to accept the error and allow them to
try to remove the cancer again, even though they suggested
that I could seek care in another center.
Failure of the Surgical Intensive Care Unit Staff to
Respond to Breathing Difficulties
I had experienced multiple unsafe situations because of

errors made by nurses. A serious situation occurred one day
after my laryngectomy while I was still in the Surgical
Intensive Care Unit (SICU). I suddenly felt an obstruction of
my airway and reached for the call button. I could not find it
because it had fallen to the floor. The nurse who cared for me
also cared for two more patients and was absent when I
needed her help. I was unable to move because I was
connected to multiple tubes and lines. I attempted to get the
attention of the staff by disconnecting my heart- and oxygensaturation monitors, and even though I was a few feet away
from the nursing station, I was ignored until my wife arrived
about 10 minutes later and called for help. I was helpless in
getting assistance without a voice and was desperately in
need of air while medical personnel passed me by.
When my wife complained about what happened to the
nursing supervisor and SICU attending physician, she was
rudely rebuffed. When I informed my surgeon about the
incident, he just shrugged his shoulders and told me that he
had little influence on what happens in the SICU, but he
assured me that things would be much better for me when I
was moved to the otolaryngology floor. The lack of willingness of my surgeon to act upon my complaint was very
disappointing. Instead of addressing the problem in the
SICU that cares for his patients when they are critically ill,
KEY POINTS:
Medical, surgical, and nursing errors are common in

patient care.
Some medical errors may be life threatening and can
lead to increased morbidity and mortality.
Preventing such errors is of utmost importance.
A dedicated patient advocate, such as a family member or a friend, is very much needed for all hospitalized patients.
Open discussion of such errors may decrease them
and eventually lead to better care.
he comforted me by promising better care when I will be less

needy for it.
Failure to Respond to Breathing Difficulties in the
Otolaryngology Ward
A similar incident also occurred in the otolaryngology

ward a week later when a nurse failed to respond to my
urgent call to suction my trachea. It happened after I
suddenly experienced difficulty in breathing because mucus
was blocking my airway. Even though I pressed the call
button, no one came to my assistance. I was finally able to
get the attention of a nurse assistant who informed me that
my nurse was on a break. The nurse assistant was not
trained in suctioning airways, but she promised to look for a
nurse. It took my nurse 15 minutes to come to suction my
trachea. I later learned that she was busy ordering supplies
on the phone during that time.
I was very distressed as I was struggling to breathe in.
Present in the otolaryngology ward during this time were
two resident physicians and several nurse assistants, yet no
one came to my help. It is clear that even on a ward
dedicated to helping people with breathing difficulties, there
were distractions that prevented physicians and nurses
from paying attention to their patients urgent needs.
Even though I brought the incident to the attention of the
nurse supervisor and the head surgeon, I never received any
feedback from them about what was to be done to prevent
such incidents in the future. The lack of response was
inappropriate and contributed to my frustration and anxiety.
Prematurely Feeding by Mouth
One of the most serious errors in my care was feeding me

by mouth with food a week too early. Early mouth-feeding
following laryngectomy with free-flap reconstruction can
lead to its failure to integrate. The oral feeding continued for
over 16 hours. The premature feeding was stopped only
because I continued questioning this practice and brought it
to the attention of the attending surgeon. I wonder what
would have occurred if I would not have not persistently
questioned the feeding. Would the error been eventually
discovered?
I repeatedly requested an explanation for this error, but
the staff avoided responding to my inquiries. I finally
learned only by looking in my medical records that this
mistake occurred because a verbal order to start oral feeding
intended for another patient was erroneously written in my
chart.
The handling of this incident was an example of a lack of
communication by the physicians and me. They failed to
explain and apologize for the mistake. Accepting responsibility for the mistake and outlining what steps should be
taken to prevent such mishaps in the future would have
been the most appropriate way of handling the incident.
Errors in Nursing Care
Some of the errors by nursing and other staff included the

following: not cleaning or washing their hands, not using
gloves when indicated, taking an oral temperature without
placing the thermometer in a plastic sheath, using an
inappropriately sized blood pressure cuff (thus getting
alarming readings), attempting to administer medications
e29
ITZHAK BROOK
by mouth that were intended to be given by nasogatric tube,

dissolving pills in hot water and administering them
through the feeding tube (which caused burning in the
esophagus), delivering an incorrect dose of a medication,
connecting a suction machine directly to the suction port in
the wall without a bottle of water, forgetting to rinse away
the hydrogen peroxide used for cleaning the tracheal breathing tube (thus causing severe irritation), forgetting to connect the call button when I was bedridden and unable to
speak, and forgetting to write down verbal orders.
I always informed the nurse supervisors and whenever I
could the resident and/or attending physicians about the
mistakes that were made; however, I was never informed
what action was taken to prevent similar errors in the
future.
Conclusion
One of the key recommendations to prevent medical errors

is that patients should choose, whenever possible, experienced places that deal with their kind of disease regularly.
Such familiarity and experience can reduce the occurrence of
errors and increase the safety of the patient. The way a
patient can contribute to the prevention of medical errors is
to be proactive and take these steps: be informed and not
hesitate to challenge and ask for explanations; become
knowledgeable about their medical condition; have family or
friends serve as an advocate in the hospital; get a second
opinion when making an important decision, such as deciding on the course of treatment; and educate the medical
caregivers about their condition and needs (prior to and
after surgery).
All of the errors made in my care make me wonder what
happens to individuals without medical background who
cannot recognize and prevent many errors. Fortunately,
despite the errors made in my care, I did not suffer any
long-term consequences. However, to prevent medical errors, I had to be continuously on guard and vigilant, which
was a very exhausting chore, especially during the difficult

recovery period.
My family members were instrumental in preventing
many errors, highlighting the value of a dedicated patient
advocate. My experiences taught me that it is essential that
medical personnel openly discuss with their patients the
mistakes that were made in their care. Since errors in
patient care weaken patients trust in their caregivers,
admission and acceptance of responsibility by the care
providers can rebuild trust and re-establish the lost confidence. The establishment of a dialogue among patients,
physicians, and other staff facilitates the discovery of the
circumstances leading to the mistake, which assists in
preventing similar ones in the future. Open discussion can
reassure the patient that their care givers are taking the
matter seriously and taking steps to make hospital stays
safer for patients.
Avoiding the discussion of the errors with the patient and
their family only increases anxiety, frustration, and anger.
This can interfere with the patients recovery and contribute
to malpractice law suits.
Medical mistakes should be prevented as much as humanly possible. Ignoring them can only lead to their repetition. Medical errors usually involve problems in caring for
and treating patients and are improved if the processes are
improved and a blame-free reporting mechanism is instituted. The recent development of a mandatory bedside
checklist is a simple, cost-effective method to prevent many
medical errors.4
I am sharing my personal experiences in the hope that
they will encourage better medical training, contribute to
greater diligence in care, and increase supervision and
communication between health care providers and their
patients. It is my hope that this presentation and article will
contribute to the reduction of medical errors and create a
safer environment in the hospital setting. It is also my hope
that if mistakes do happen, medical caregivers will openly
discuss them with their patients.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Itzhak Brook*
REFERENCES
1. Tezak B, Anderson C, Down A, et al. Looking ahead: the use of
prospective analysis to improve the quality and safety of care. Healthc Q.
2009;12:80-84.
2. Griffen FD, Turnage RH. Reviews of liability claims against surgeons:
what have they revealed? Adv Surg. 2009;43:199-209.
3. Studdert DM, Mello MM, Gawande AA, et al. Claims, errors, and
e30
compensation payments in medical malpractice litigation. N Engl J Med.

2006;354:2024-2033.
4. Byrnes MC, Schuerer DJ, Schallom ME, et al. Implementation of a
mandatory checklist of protocols and objectives improves compliance with a
wide range of evidence-based intensive care unit practices. Crit Care Med.
2009;37:2775-2781.
PERSONALIZED ONCOLOGY FOR COLORECTAL

CANCER: READY FOR PRIME TIME OR
STOP THE TRAIN?
CHAIR
Alan P. Venook, MD
San Francisco, CA
SPEAKERS
The Sarah Cannon Research Institute
Nashville, TN
Robert S. Warren, MD
San Francisco, CA
Is There Currently an Established Role for

the Use of Predictive or Prognostic Molecular
Markers in the Management of Colorectal
Cancer? A Point/Counterpoint
By Alan P. Venook, MD, Johanna C. Bendell, MD, and Robert S. Warren, MD
Overview: The term personalized oncology means different

things to the oncologist than to the patient. But fundamentally,
the phrase creates the expectation that decisions can be
informed by the unique features of the patient and patients
cancer. Much like determining antibiotic sensitivities in urinary tract infections, the oncologist is expected to choose the
right treatment(s), for each individual patient.
Numerous methods can be used to personalize management decisions, although truly useful biomarkers continue to
escape our grasp. Positron Emission Tomography in patients
with GI stromal tumors or genotyping of c-kit in chronic
myelogenous leukemia cells can guide the use of imatinib,
these scenarios represent a minority of patients. The promise
of individualized therapy, however, has led to the commercialization of numerous assays to probe patients genetic
make-up and that of the tumor. Breast cancer management
HE TERM personalized oncology means different

things to the oncologist than to the patient. But fundamentally, the phrase creates the expectation that decision
making can be informed by the unique features of the
patient and of that patients cancer. Much like culturing the
urine in a urinary tract infection and determining antibiotic
sensitivities, it implies that oncologists can pick just the
right treatment, or combination of treatments, for each
specific patient.
There are numerous tools that can be used to personalize management decisions, although for much of the last
decade the truly useful biomarkers were few and far between. Although positron emission tomography imaging in a
patient with a GI stromal tumor patient or mutational
analysis of chronic myelogenous leukemia cells could direct
the proper decision about whether to use imatinib, these
scenarios represented a distinct minority of patients. The
promise, however, led to the commercialization of numerous
assays to probe a patients genetic make-up and/or the
cancer. We now know that gene recurrence scores can
personalize breast cancer management in some patients,
and more recently we have capitalized on mutations in
non-small cell lung cancers and melanomas to subdivide
these patients into subsets of diseases. We have shown
recently that some of these infrequent mutations can be
targeted successfully by small-molecule inhibitors.
Despite the high incidence of colorectal cancer and our
relatively long-standing grasp of the molecular pathways in
colorectal cancerthe polyp-to-malignancy transformationthe management of disease in patients with colorectal
cancer for years is mostly empiric and movement toward
personalization has been incremental. Now, however, the
availability of newer imaging modalities and commercial
assays for genetic mutational analysis of tumor specimens
has put the oncologist and oncologic surgeon in the crossfire
with patients and families who believe the era of personalization is here.
Now the debate: The medical oncologist (Johanna Bendell)
has benefitted from the analysis of gene recurrence scores.

More recently the analysis of germline or tumor-associated
mutations in non-small cell lung cancer and melanoma has led
to clinically meaningful molecular subsets of these diseases,
guiding the successful targeting of such cancers with smallmolecule inhibitors.
Despite the high incidence of colorectal cancer and our
relatively long-standing grasp of the molecular pathways in
colorectal carcinogenesis, the management of these patients
remains mostly empiric and movement toward personalization has been slow and incremental. Now, however, molecular imaging and commercial assays for genetic makeup of
tumor specimens has put the oncologist and oncologic surgeon in the crossfire with patients and families who believe
the era of personalization is here.
wants the surgeon (Robert Warren) to carefully collect

tumor tissue of the patient undergoing surgery. Always a
willing collaborator, the surgeonmindful of cost and the
pathologists timeasks what makes the medical oncologist
think these tests will be helpful. The patient expects a
compromise that will help personalize care.
Johanna Bendell, MD: Ready for Prime Time
As part of the referral to the surgeon, the medical oncologist lists a number of specific tissue acquisition needs and
handling instructions to the surgeon. The message: these
are the biomarkers we will use to decide how to treat this
patient and here are the data to support our decision.
KRAS
Many of the biomarkers currently in use or under active

investigation are based around research done with epidermal growth factor receptor (EGFR) pathway inhibitors.
These include KRAS, BRAF, and rash. The data from the
CRYSTAL trial, where patients were randomly assigned in
the first-line setting to receive fluorouracil (FU), leucovorin
(LV), and irinotecan (IFL; FOLFIRI) with or without cetuximab, is well known.1 This trial showed that the addition of
cetuximab improved response rate (49% vs. 39%; p 0.0038)
and progression-free survival (8.9 vs. 8.0 months; p 0.05).
For the patients with wild-type KRAS status, who comprised
37% of the patients, the benefit of cetuximab over no cetuximab was greater for response rate (57.3% vs. 39.7%; p
From the Department of Medicine (Hematology/Oncology); Department of Surgical

Oncology, University of California San Francisco Helen Diller Comprehensive Cancer
Center, San Francisco, CA; Gastrointestinal Oncology Research, Sarah Cannon Research
Institute, Nashville, TN.
Address reprint requests to Robert S. Warren, MD, Professor of Surgery, Chief, Surgical
Oncology, University of California, San Francisco, 1600 Divisadero St., Room A-723, San
Francisco, CA 94115; email: robert.warren@ucsfmedctr.org.
1092-9118/10/1-10
193
VENOOK, BENDELL, AND WARREN
0.001), progression-free survival (9.9 vs. 8.4 months; p

0.0012), and overall survival (OS; 23.5 vs. 20.0 months; p
0.0093). Patients who had KRAS mutations derived no
benefit from the addition of cetuximab treatment. Multiple
other large trials, including OPUS (infusional FU, LV, and
oxaliplatin [FOLFOX] with or without cetuximab), COIN
(FOLFOX/capecitabine plus oxaliplatin with or without cetuximab), NORDIC VII (infusional FU/folinic acid/oxaliplatin with or without cetuximab), PRIME (FOLFOX with or
without panitumumab), PACCE (FOLFOX-bevacizumab
with or without panitumumab), 181 (FOLFIRI with or
without panitumumab), NCIC CO.17 (cetuximab vs. best
supportive care), and panitumumab with or without best
supportive care, have all consistently shown that patients
with KRAS mutations do not have any improvement in
outcomes with EGFR-based treatments.
Determination of KRAS status is now standard of care
and required before initiation of anti-EGFR therapy for
patients with colorectal cancer. Although KRAS status does
not define which patients will benefit, it can define patients
KEY POINTS
194
Current staging methods for early colorectal cancer

are inadequate to accurately define the prognosis of
stage 2 or stage 3 disease.
Similarly, there are very few biomarkers that have
been proven to guide the choice adjuvant chemotherapy or biological therapy, with some notable
exceptions.
A large body of evidence exists that argues for the use
of genotyping the ras and raf oncogenes when the
application of anti epidermal growth factor receptor
(EGFR) antibodies is being considered, or the determination of microsatellite stability when considering
adjuvant treatment with fluorouracil and leucovorin.
At least three gene expression arrays have been
commercialized for determining the prognosis of either stage 2 or stage 3 colon cancer. Each signature
differs from the others in terms of genes that are
assayed, and none has been found to predict response
to any specific colon cancer therapy. And none of the
available arrays has been approved for routine use
under the National Comprehensive Cancer Network
(NCCN) guidelines for colorectal cancer.
The road forward will require quality control over
biospecimens; developing reference labs with stateof-the-art analytical performance for DNA, RNA, and
protein assays; using the best trial design and informatics that represent the consensus of leaders in the
field; and working with the U.S. Food and Drug
Administration (FDA) on the complex regulatory issues attendant to biomarker studies.
There is a surplus of possible prognostic and predictive biomarkers for colon cancer, but how best to
reliably develop these to the point that they benefit
our patients in individualized cancer care remains a
huge challenge.
who will not benefit and spare them the toxicities, loss of
time, and loss of opportunity from receiving an ineffective
therapy.
KRAS-Specific Mutations: It Is Not As Simple As We Once
Thought It Was
What was first thought to be black or whiteKRAS wild

type or mutantturns out to be shades of gray as we are
learning that not all mutations are the same. Most mutations of KRAS are located in codon 12, but approximately
20% of patients with KRAS mutations have a mutation in
codon 13, the G13D mutation.2 In a retrospective analysis,
patients with tumors harboring G13D mutations had
equivalent benefit from treatment with cetuximab therapy
as did KRAS wild type patients. This was confirmed in a
pooled analysis of patients in the CRYSTAL and OPUS
studies where the progression-free survival hazard ratios
(HRs) were similar for patients with wild-type (0.66; 95% CI,
0.55 to 0.80) and G13D mutant (0.60; 95% CI, 0.32 to 1.12)
KRAS.3
Although this observation needs to be validated in a
prospective randomized study, it suggests that the specific
KRAS mutation is an important detail and that tumor
behavior may vary over codons 12 and 13 as well as other
mutations that are lumped together.
BRAF
BRAF sits downstream of KRAS in the cell-signaling

pathway, so it had been assumed that mutations in BRAF
would have the same ramifications as KRAS mutations. The
inactivity of EGFR-targeted monotherapy either cetuximab or panitumumabin patients who had BRAF mutations was confirmed in a retrospective report by
DeNicolantonio.4 However, when data from two first-line
chemotherapy plus cetuximab trials (CRYSTAL and OPUS)
were pooled and analyzed, it seemed that patients with
BRAF mutations had much poorer prognoses than the larger
number of patients without BRAF mutations, but that there
was still some benefit to the EGFR antibody (OS BRAF
mutant, 14.1 months with cetuximab vs. 10.3 months without cetuximab).1
This prognostic value makes knowledge of BRAF mutational status an important factor in management. Although
it is a rare mutation (10% of colorectal cancer), patients with
BRAF mutations have average OS of approximately 1 year,
which is approximately half that over the overall metastatic
colorectal cancer population. For these patients, that knowledge may be important when oncologist and patient consider
issues related to treatment holidays given the aggressiveness of their disease. It also may lead to consideration of
directing these patients toward specific clinical trials. There
are a number of inhibitors of the Ras/Raf/MEK pathway
currently in development. A recent phase II study of the
BRAF inhibitor PLX4032 (vemurafenib) in 21 patients with
BRAF mutant colorectal cancer showed a response rate of
5%, with a median progression-free survival of 3.7 months.5
Though not as impressive as the data in patients with BRAF
mutant melanoma, this agent does appear to warrant further evaluation. Trials are also ongoing looking at MEK
inhibitors both as monotherapy and in combination (MEK is
a downstream target) and combinations of BRAF and KRAS
inhibitors. These agents already show promising data in
BRAF mutated melanoma.6
BIOMARKERS AND COLORECTAL CANCER
Determination of BRAF status is important at the outset

for guidance in terms of prognosis. As of today, do not order
BRAF until the tumor is known to be KRAS wild type; by
convention, KRAS mutant cancers do not have BRAF mutations.
And I reserve the right to change the above recommendation. The BRAF story teaches us that we should not rush to
judgment on the utility of biomarkers until the correct,
adequately powered trials have been performed. To determine whether a biomarker is predictive, where the biomarker status can be used to determine whether a patient
will or will not benefit from treatment, a four-arm trial must
be done. The arms of the trial will include patients who do
and do not receive active therapy and patients who do and do
not have the presence of the biomarker. Trials that include
only the presence or absence of a biomarker can tell us only
whether that biomarker is prognostic, when the biomarker
is associated with better or worse outcomes, no matter what
the treatment.
Rash: Acneiform Skin Change to EGFR Inhibitors
Biomarkers not only include tests on tumor and blood, but

also the reaction an individual patient may have to therapy.
Skin rash reaction to EGFR inhibitors is an example. It has
been well established that worsened skin reaction to EGFR
inhibitors is associated with improved outcome to treatment. Data from the PRIME study, in which patients were
randomly assigned to receive FOLFOX with or without
panitumumab found that wild-type KRAS patients receiving
panitumumab who had a grade 2 to 4 skin reaction compared with patients who had a grade 0 to 1 skin reaction had
improved progression-free survival (11.1 vs. 6.0 months; p
0.001) and OS (28.3 vs. 11.5 months; p 0.0001).7 Given the
importance of skin rash reaction to outcomes, the EVEREST
study attempted to increase the dose of cetuximab therapy
on the basis of rash. Patients received IFL and cetuximab,
and patients who developed only grade 0 to 1 rash were
randomly assigned to receive high-dose versus low-dose
cetuximab.8 Though the response rate was improved in the
high dose patients (30% vs. 16%), there was no difference in
OS. At the moment, we can say that the individual patient
biomarker of skin reaction to EGFR inhibitors can help us
predict improved outcomes with higher-grade rash.
When you see our patient in follow-up, take note of the
extent of acne. If it is quite severe, make sure the patient
knows that is a good sign.
Node-Negative Colon Cancer: Mismatch Repair
The use of adjuvant chemotherapy for patients with stage

II colon cancer has been the subject of much debate, and the
ability to identify patients who would be more or less likely
to benefit for adjuvant chemotherapy is needed. The status
of DNA mismatch repair (MMR) seems to be a biomarker to
guide the choice to use adjuvant chemotherapy for these
patients. It is known that patients with tumors that are
deficient in MMR (dMMR) have a better overall prognosis
than those who have proficient MMR (pMMR). A pooled
analysis of 457 patients and then 1,027 patients with stage
II or III colon cancer found that patients with pMMR status
had a benefit from adjuvant chemotherapy with a diseasefree survival HR of 0.67 (95% CI, 0.48 to 0.93), whereas
patients with dMMR status derived no benefit from FU-
based chemotherapy (HR 1.10; 95% CI, 0.42 to 2.91).9

Further, for the stage II patients with dMMR there seemed
to be a worse OS with the use of FU adjuvant chemotherapy
(HR 2.95; 95% CI, 1.02 to 8.54).
When you refer the patient to the medical oncologist for
discussion of adjuvant therapy for stage II disease, please
also ask the pathologist to determine MMR status. If the
tumor is dMMR, the use of adjuvant chemotherapy will be
discouraged.
Recurrence Score
Although there are numerous gene signatures in development, the assay with the most supporting data is a 12-gene
recurrence score. The recurrence score was derived from
tumor samples from 1,851 stage II or III patients who
received FU adjuvant chemotherapy or surgery alone, identifying a set of 12 genes that appeared to be associated with
risk of disease recurrence.10 This recurrence score was then
analyzed in 1,436 tumor samples from patients with stage II
cancer treated with or without FU chemotherapy on the
QUASAR trial.11 Patients with a recurrence score greater
than 40 had a 25% rate of 3-year disease recurrence compared with patients with a recurrence score less than 30 who
had a 3-year disease recurrence rate of 8%. However, although the recurrence score seems to identify patients with
an increased risk of disease recurrence, when the recurrence
score was applied to look for patients who would then have
benefit from adjuvant chemotherapy, no conclusion could be
drawn. Larger prospective trials are needed.
I expect the recurrence score to be helpful in some patients, but let me discuss the ramifications of the score with
the patient before we send the specimen, since the importance of the score is a function of the philosophy of the
patient and oncologist. In general, though, we should get as
much information as we can.
Up-and-Coming Biomarkers
EGFR Ligands
Amphiregulin and epiregulin are ligands of EGFR. A

study of examining DNA, RNA, and protein expression
profile in 110 patients who received cetuximab found that
patients with increased expression of amphiregulin and
epiregulin had an increased rate of disease control.12 Further, a retrospective analysis of IFL-refractory patients
treated with cetuximab and IFL for metastatic colorectal
cancer found that patients with high epiregulin levels plus
KRAS wild-type status showed a response rate of 58%
compared with 40% for KRAS wild type alone, and a median
progression-free survival of 36 weeks compared with 24
weeks.13 Another analysis for patients treated on the NCIC
CO.17 study of cetuximab versus best supportive care
showed that patients who had KRAS wild-type and high
epiregulin had a much improved HR for OS benefit compared with the KRAS wild-type only or total population
(HR 0.46 vs. 0.55 vs. 0.70, respectively).14 The combination of KRAS and EGFR ligand status appears promising to
select patients who will have the most benefit from antiEGFR therapy.
Phosphoinositide 3-Kinase and Phosphatase and Tensin Homolog
The phosphoinositide 3-kinase (PI3K) pathway is likely

the most commonly activated pathway in cancers. Approxi-
195
mately 40% of patients with colorectal cancer have alterations in the PI3K pathway including mutations and loss of
phosphatase and tensin homolog (PTEN) activity.15,16 Multiple agents that affect this pathway are in development,
including PI3K inhibitors, mammalian target of rapamycin
(mTOR) inhibitors, and AKT inhibitors. For patients with
colorectal cancer, studies have been reported looking at
mTOR inhibitors17 and AKT inhibitors.18 It remains to be
seen whether there is any association with outcomes with
these agents in patients who have documented abnormalities in this pathway, but biomarker studies are underway.
c-MET
Overexpression of the c-met receptor in colorectal cancers

is associated with a poor prognosis.19 c-MET activation is
associated with colon cancer tumorigenesis and metastasis,
and in preclinical studies inhibition of c-met decreases colon
tumor spread.20 c-MET is also involved in signal transduction when growth factor receptors such as EGFR and vascular EGFR (VEGFR) are activated.21 The combination of
c-MET and EGFR inhibition has shown in a randomized
phase II study improved time to progression and OS in
patients with non-small cell lung cancer with c-met overexpression.22 c-MET inhibitors are currently under investigation for patients with colorectal cancer. A phase I study of
IFL, cetuximab, and c-MET inhibitor ARQ-197 showed very
promising results in pretreated disease.23 First- and secondline studies are ongoing with inhibitors of the c-MET pathway. We await further data on the correlation in patients
with colorectal cancer between c-MET expression and outcomes with treatment with c-MET pathway inhibitors.
Ready for Prime Time: Conclusion
As you can see, the era of personalized medicine has

arrived for patients with colorectal cancer. We are already
using markers to help guide clinical decisions. The markers
we have will only continue to grow from here. We need to
continue to find new markers, which can be done only if
correlative biomarker studies are included in clinical trials.
We are at only the beginning of fully realizing the potential
of personalized biomarker status to deliver the best treatment to our patients.
Robert S. Warren, MD: Personalized Oncology for
Colorectal Cancer: Stop the Train
The application of molecular biomarkers to prognosis, and

markers that are predictive of benefit from specific chemotherapeutic regimens, is receiving tremendous interest on
the part of patients, clinicians, and the pharmaceutical
industry. Patients wishes, on the one hand, successful
therapeutics, and on the other hand, avoiding potentially
toxic therapy if the chance of benefit is remote; clinicians
search for guidance in the care of their patients who have
exhausted most standard therapy options, and the keen
interest of the pharmaceutical and biotechnology companies
in developing drugs that are effective, even if only in a
subset of patients with colorectal cancer, are all driving the
interest in the application of prognostic and predictive
biomarkers. Identifying biomarkers that would permit personalized therapies and more successfully targeted drug
development may save lives and require many fewer millions of dollars in drug development and in standard clinical
196
trials. If fact, a review of PubMed shows papers describing

more than 60 new biomarkers for cancer just in 2011. Most
are retrospective studies, small and inadequately powered
to perform multivariable analysis in combination with more
thoroughly examined potential markers.
As a measure of the increasing interest in this pursuit, it
may be noteworthy that in PubMed, searching under colon
cancer biomarkers, 560 papers were published in 2008,
whereas 1,228 were published in 2011. Under personalized
oncology, 43 papers were published in 2008, whereas 161
papers were published in 2011.
The questions that Dr. Bendell so nicely discussed focus
on a few of these: KRAS/BRAF mutations and the activity of
EGFR antibodies (MoAby), the utility of gene expression
arrays in determining prognosis (generally in advanced
colorectal cancer), and the impact of microsatellite instability on prognosis and response to chemotherapy. Despite a
plethora of research publications, the U.S. Food and Drug
Administration (FDA), the National Comprehensive Cancer
Network (NCCN), and ASCO24,25,26 have recommended the
standard use of very few markers. The important question
here is why? Are there guidelines for the development,
verification, and validation of markers to hasten their application to the clinic? Following is a summary of definitions
and a process to permit moving forward, part of a paper that
gives a good perspective on the challenges we face in this
pursuit26: biomarkera characteristic that is objectively
measured and evaluated as an indicator of normal biologic
processes, pathogenic processes, or pharmacologic responses
to a therapeutic intervention; diagnostic biomarkers early
detection biomarkers and disease classification; predictive
biomarkerspredict patients likely to have an adverse
event to a specific agent and predict patients likely to
respond to a specific agent; outcome biomarkersforecast
response, progression, and recurrence; assay validation
the process of assessing the assay and its performance
characteristics and determining the optimal conditions that
will generate a reliable, reproducible, and accurate biomarker assay for the intended application; clinical qualificationlinking a biomarker (using data obtained by a
biomarker assay) with meaningful biologic or clinical outcomes; high-quality biospecimens (collection methods, quality assessment), accurate detection of markers (reference
standards, analytic performance and methods), useful patient annotation, design of new bioinformatics approaches to
the study of biomarkers, and the application of optimized
retrospective and prospective study designs (adaptive clinical trials).
The goal of this discovery process is to elucidate the
physiologic, toxicological, and pharmacologic, or clinical significance of the test results.26 The tools that we need to
query our patients genomes and their tumors have developed rapidly. Analysis of DNA copy number by comparative
genomic hybridization (CGH), mutation detection, epigenetic profiling, gene expression profiling, determining splice
variants of significant genes and functional proteomics and
metabolomics are becoming routine on frozen or paraffinfixed tumors. High-density single nucleotide polymorphism
(SNP) arrays (containing 500,000 to 1,000,000 individual
SNPs) permit an in-depth analysis to the host genome.
Clearly, these methods are not rate limiting in terms of
progress toward individualized cancer care. Even the most
robust of assays associated with therapeutic response or
Fig. 1. Signalling pathways that are activated upon

ligation of epidermal growth factor with its receptor
(EGFR).
Abbreviations: CRC, colorectal cancer; EGFR, epidermal
growth factor receptor; PI3K, phosphoinositide 3-kinase;
PTEN, phosphatase and tensin homolog.
prognosis have various weak links. A few examples of this

are discussed in the following sections.
EGFR Antibodies and Mutations in KRAS and BRAF
Dr. Bendell accurately points out that the majority of

retrospective clinical trials support the notion that patients
whose tumors harbor mutations in codon 12 of KRAS show
very little or no response to EGFR MoAby, whereas tumors
with mutations in codon 13 seem to respond as well as Ras
wild-type tumors.27 But Dr. Bendell has omitted the observation from the COIN, NORDIC, PACCE, and CAIRO-2
suggest that there is a detrimental effect of adding cetuximab to chemotherapy in patients with tumors containing
mutant KRAS compared with chemotherapy alone.28-31 The
assumption that approximately 40% of colorectal cancers harbor ras mutation is misleading. The focus most likely should
be on the estimated 75% to 80% KRAS codon 12 mutations.
Further, the story of Braf is more complicated. Dr. Bendells point is well taken that tumors with BRAF mutations
(10% of colon cancer) have a poorer prognosis than wild-type
tumors, but they may respond equally well to EGFR inhibition as do wild-type tumors. But, there are other abnormalities that can mimic (phenocopy) KRAS mutation vis a` vis
EGFR inhibition that merit more emphasis than discussed
by Dr. Bendell. PTEN blocks signaling downstream of
growth factor receptors; PTEN especially manifests this
effect by impairing signaling emanating from PI3K step27
(Fig. 1). Although the assay methodology for PTEN has not
Fig. 2. Prevalence of epidermal growth factor (EGFR)

pathway deregulations and response to monoclonal antibodies targeting EGFR in chemotherapy-refractory advanced colorectal cancer.
Abbreviations: PI3K, phosphoinositide 3-kinase; PTEN,
phosphatase and tensin homolog.
197

Table 1. Tumor Marker Utility Grading System Levels of
Evidence
Level
Definition
Prospective, marker primary objective

Well-powered or meta-analysis
Prospective, marker the secondary objective
Retrospective, outcomes, multivariate analysis (most currently published
marker studies are level of evidence III)
Retrospective, outcomes, univariate analysis
Retrospective, correlation with other marker, no outcomes
II
III
IV
V
Adapted with permission from Hayes DF, Bast RC, Desch CE, et al. Tumor
marker utility grading system: a framework to evaluate clinical utility of tumor
markers. J Natl Cancer Inst. 1996;88:1464.
been well established, the 20% to 40% of colon cancers

lacking PTEN expression as a result of promoter methylation would be expected not to respond to EGFR inhibition.
The same applies to tumors with activated PI3K, which
again is downstream of ligand signaling through the EGFR
and could explain why so few patients with wild-type KRAS
and BRAF fail to benefit from EGFR inhibition. Figure 2
(from Dienstmann, Vilar, and Tabernero27) describes the
possible abnormalities that can account for resistance to
EGFR antibodies.
Gene Expression Arrays
There are numerous papers describing the correlation

between gene expression and prognosis in colorectal cancer.
Although Oncotype Dx (Genomic Health, Inc., Redwood
City, CA) and Coloprint (Agendia, Irvine, CA) have been
commercialized and are available for purchase, a number of
other investigators have reported, tested, and subsequently
validated mRNA arrays as prognostic tools in colon cancer.
In 2004, Wang and colleagues32 reported a 23-gene signature based on the Affymetrix (Santa Clara, CA) Gene Chip
that accurately predicted relapse in stage II colon cancers.
Eschrich and colleagues as well as Arango33,34 reported in
2005 that a 43-gene signature predicted recurrence in stage
II and III colon cancer. Barrier and colleagues in 2006
identified a 30-gene signature using mRNA microarrays
that predicted outcome in stage II colon cancer35 In 2007,
Gray and colleagues36 reported the first experience using a
set of 761 genes assayed by reverse transcriptase polymerase chain reaction (RT-PCR) to develop a signature for
colon cancer recurrence, and this evolved in Oncotype Dx
into a validation study in 2011.37,38 The diagnostics company Almac (Souderton, PA) used a 634-probe set signature
(Col Dx)39 to identify patients with colon cancer with a high
risk of recurrence in stage II colon cancers that is undergoing multiple independent validation studies. And in 2011,
Salazar and colleagues demonstrated and validated the
utility of an 18-gene signature, originally described in 2007,
and is now commercially available40,41 and sold as Coloprint
for colon cancer prognosis. Febbo and colleagues identified

levels of evidence for marker studies and NCCN categories
of evidence.24 Level 1 is used to characterize markers that
were evaluated prospectively where the marker was the
primary objective of the study. Most of the markers in the
literature fall under Level IV, in which evaluation was
retrospective and outcomes were determined in univariate
analysis. Most of the array-based and RT-PCR based
genomic signatures fall under this category, including ColoPrint and Oncotype Dx. Interestingly, none of these arraybased signatures was informative in terms of response to
chemotherapy. Also, there is very little overlap in the set of
genes chosen to create a prognostic signature in colon
cancer, even though it seems that risk assessment is comparable regardless of the platform used. Choosing the best
platform involves consideration of cost, rapidity of turnover,
reproducibility, and whether any of these gene expression
signatures have been assessed in multivariate analysis with
other potential markers of outcome (p53 mutational status,
p27 loss, thymidylate synthase expression, microsatellite
instability, and others). Consequently, although available to
the clinician and the patient, the exact utility of gene
expression signatures remains a subject of debate. Neither
Oncotype Dx nor ColoPrint is included in the NCCN guidelines for colon cancer.
p53: When Genotype Interacts with Gender and
Chemotherapeutic Regimen
The p53 tumor suppressor is frequently mutated in colon

cancer, but the influence of such mutations on survival is
still controversial42 p53 mutations have been inferred by
positive staining using IHC since the wild-type protein is
degraded very rapidly, and is not demonstrable in the
nucleus by immunohistochemistry, although the gold standard for p53 mutation analysis is Sanger sequencing. We
investigated whether DNA-binding domain-specific mutations in p53 are predictive of survival in stage III colon
cancer. p53 was evaluated in an intergroup trial, CALGB
89803,43 in patients with stage III colon cancer who were
randomly assigned to receive adjuvant FU/LV or FU/LV
with IFL, and various molecular markers were correlated
with outcomes. p53 was genotyped in 607 patient tumors:
p53 mutations were identified in 274 tumors, divided
equally between zinc-binding and nonzinc-binding regions
of the DNA-binding domain. Overall, p53 status was not
predictive of benefit from either adjuvant regimen. Unexpectedly, however, the 5-year OS of women with tumors
harboring nonzinc-binding mutations treated with FU/LV
was 97% compared with OS of 72% for women with p53
wild-type tumors (p 0.004). Adding IFL to FU/LV negated
this survival benefit (5-year OS of 81% vs. 72%). Conversely,
5-year OS of women harboring tumors with zinc-binding
Table 2. Pair-Wise Comparisons of Survival by p53 Mutational Status

Wild-Type vs. Zinc Binding
Wild-Type vs. NonZinc Binding
Zinc Binding vs. NonZinc Binding
Patient Subset
P OS
DFS
P OS
DFS
P OS
DFS
Men
Women, FU/LV
Women, IFL
Men, FU/LV
Men, IFL
0.58
0.04
0.71
0.48
0.12
0.28
0.24
0.79
0.66
0.05
0.19
0.004
0.18
0.80
0.10
0.18
0.002
0.53
0.48
0.24
0.48
0.001
0.49
0.35
0.85
0.7
0.001
0.83
0.29
0.63
Abbreviations: DFS, disease-free survival; FU, fluorouracil; IFL, irinotecan; LV, leucovorin; OS, overall survival.
198
mutations who received FU/LV reversed the poor survival of

women with tumors harboring zinc-binding mutations and
improved 5-year OS (50% vs. 73%; p 0.1). No difference in
OS was observed for men in either treatment arm or when
genotype was considered. We conclude that CALGB 89803
demonstrated a lack of survival benefit for patients with
stage III colon cancer when IFL was added to FU/LV. We
now show that in the setting of a large clinical trial, refined
stratification of women, based on domain-specific mutations
of p53 identifies subsets of patients likely to benefit from, or
respond poorly to, adjuvant FU/LV. The interaction of p53
genotype, gender, and adjuvant therapy regimen has the
potential to be paradigm changing in the treatment of colon
cancer, and possibly other malignancies (Table 2).
These data, if validated, suggest that evaluation of p53
genotype and gender may guide clinicians to make rational
choices of adjuvant therapy. Clearly, prospective clinical
trials must be sufficiently large and adequately powered to
ferret out previously undescribed interactions between oncogenes, tumor suppressor genes, and gender.
In summary, although many studies purporting to demonstrate either prognostic or predictive value in a variety of
markers (DNA based, protein based, and metabolite based)
have been published, only a few have acquired sufficient
evidence to warrant possible inclusion in clinical decision
making in colon cancer.46-48 Further, a variety of regulatory
hurdles must be addressed during biomarker development
in the future.47
The Road Ahead: Companion Diagnostics in
Clinical Trials
As use and interest in laboratory-developed tests continues to grow, it will be important to work with the FDA to
define the oversight framework that takes into account
higher-risk tests that require more complex validation,
equipment, and software. Regulatory clarity and predictability will be important to ensure that high-quality tests
reach the market expeditiously.
Laboratory tests have been critical to recent advances in

oncology because they can be developed rapidly and targeted
locally. Currently, no controls have been placed on the
marketing claims of theses tests, and limited control of test
development exists. Likewise, there are limited premarket
independent review or postmarket reporting requirements.
These requirements are dependent on state regulations
and/or the laboratory accreditation process. Currently, some
of these functions for laboratories performing molecular
testing, such as postmarket performance, are conducted
voluntarily.
Some sectors are concerned that additional regulation of
molecular testing may harm innovation because associated
regulatory barriers may hinder the pace of progress.
The product lifecycles of molecular testing assays often
differ from those of drugs. This could present issues related
to innovation with new therapies and their companion
diagnostics.
Most molecular testing is performed in laboratories that
meet or exceed the laboratory quality standards currently
set by the Centers for Medicare & Medicaid Services (CMS,
Rockville, MD) under the Clinical Laboratory Improvement
Amendments (CLIA). However, CLIA regulations do not
measure the clinical validity or clinical utility of individual
molecular tests. The FDA does have the authority to ensure
the safety and effectiveness of molecular tests, which includes an evaluation of clinical validity.
New therapies that are found to be effective in patients
with a specific biomarker profile may require a companion
diagnostic test to be approved by the FDA if these tests are
considered essential for the safe and effective use of a
therapy. Label changes to already approved treatments can
also be required if a companion diagnostic is shown to
improve safety. It will be important to understand how
clinical trials for drugs and their companion diagnostics
should be designed and carried out, validated, and brought
to the clinic.49,50
Author
Alan P. Venook
Employment or
Leadership
Positions
Consultant or
Advisory Role
Abbott
Laboratories (U);
Bristol-Myers
Squibb (U);
Chugai Pharma
(U)
Stock
Ownership
Honoraria
Research
Funding
Amgen; Bayer;
Genentech;
ImClone
Systems;
Novartis; Pfizer
Expert
Testimony
Other
Remuneration
Johanna C. Bendell*
Robert S. Warren*
REFERENCES
1. Van Cutsem E, Kohne CH, Lang I, et al. Cetuximab plus irinotecan,
fluorouracil, and leucovorin as first-line treatment for metastatic colorectal
cancer: updated analysis of overall survival according to tumor KRAS and
BRAF mutation status. J Clin Oncol. 2011;29:2011-2019.
2. De Roock W, Jonker DJ, Di Nicolantonio F, et al. Association of KRAS
p.G13D mutation with outcome in patients with chemotherapy-refractory
metastatic colorectal cancer treated with cetuximab. JAMA. 2010;304:18121820.
3. Tejpar S, Bokemeyer C, Celik I, et al. The role of the KRAS G13D
mutation in patients with metastatic colorectal cancer (mCRC) treated with
first-line chemotherapy plus cetuximab. J Clin Oncol. 2011;29 (suppl; abstr
630).
4. Di Nicolantonio F, Martini M, Molinari F, et al. Wild-type BRAF is

required for response to panitumumab or cetuximab in metastatic colorectal
5. Kopetz, S, Desai J, Chan E, et al. PLX4032 in metastatic colorectal
cancer patients with mutant BRAF tumors. J Clin Oncol. 2010;28 (suppl;
abstr 3534).
6. Infante JR, Falchook GS, Lawrence DP, et al. Phase I/II study to assess
safety, pharmacokinetics, and efficacy of the oral MEK 1/2 inhibitor
GSK1120212 (GSK212) dosed in combination with the oral BRAF inhibitor
GSK2118436 (GSK436). J Clin Oncol. 2011;29 (suppl; abstr 630).
7. Douillard JY, Cassidy J, Jassem F, et al. Randomized, open-label, phase
3 study of panitumumab With FOLFOX4 vs FOLFOX4 alone as 1st-line
199

treatment (tx) for metastatic colorectal cancer (mCRC): efficacy by skin
toxicity. J Clin Oncol. 2010;28 (suppl; abstr 3528).
8. Tejpar, S, et al. Phase I/II study of cetuximab dose-escalation in patients
with metastatic colorectal cancer (mCRC) with no or slight skin reactions on
cetuximab standard dose treatment (EVEREST): Pharmacokinetic (PK),
Pharmacodynamic (PD) and efficacy data. J Clin Oncol. 2007;25 (suppl; abstr
4037).
9. Sargent DJ, Marsoni S, Monges G, et al. Defective mismatch repair as a
predictive marker for lack of efficacy of fluorouracil-based adjuvant therapy in
colon cancer. J Clin Oncol. 2010;28:3219-3226.
10. OConnell MJ, Lavery I, Yothers G, et al. Relationship between tumor
gene expression and recurrence in four independent studies of patients with
stage II/III colon cancer treated with surgery alone or surgery plus adjuvant
fluorouracil plus leucovorin. J Clin Oncol. 2010;28:3937-3944.
11. Gray RG, Quirke P, Handley K, et al. Validation study of a quantitative
multigene reverse transcriptase-polymerase chain reaction assay for assessment of recurrence risk in patients with stage II colon cancer. J Clin Oncol.
2011;29:4611-4619.
12. Khambata-Ford S, Garrett CR, Meropol NJ, et al. Expression of
epiregulin and amphiregulin and K-ras mutation status predict disease
control in metastatic colorectal cancer patients treated with cetuximab. J Clin
Oncol. 2007;25:3230-3237.
13. Tejpar S, De Roock W, Biesmans B, et al. High amphiregulin and
epiregulin expression in KRAS wild-type colorectal primaries predicts response and survival benefit after treatment with cetuximab and irinotecan for
metastatic disease. Presented at: 2008 Gastrointestinal Cancers Symposium;
January 2010; Orlando, FL.
14. Jonker DH, Karapetis C, Harbison C, et al. High epiregulin (EREG)
gene expression plus K-ras wild-type (WT) status aspredictors of cetuximab
benefit in the treatment of advanced colorectal cancer (ACRC): results from
NCIC CTG CO. 17a phase III trial of cetuximab versus best supportive care
(BSC). J Clin Oncol. 2009;27 (suppl; abstr 4016).
15. Brugge J, Hung MC, Mills GB. A new mutational AKTivation in the
PI3K pathway. Cancer Cell. 2007;12:104-107.
16. Baldus SE, Schaefer KL, Engers R, et al. Prevalence and heterogeneity
of KRAS, BRAF, and PIK3CA mutations in primary colorectal adenocarcinomas and their corresponding metastases. Clin Cancer Res. 2010;16:790-799.
17. Altomare I, Russell KB, Uronis HE, et al. Phase II trial of bevacizumab
(B) plus everolimus (E) for refractory metastatic colorectal cancer (mCRC).
J Clin Oncol. 2010;28 (suppl; abstr 3535).
18. Bendell JC, Nemunaitis J, Vukelja SJ, et al. Randomized placebocontrolled phase II trial of perifosine plus capecitabine as second- or third-line
therapy in patients with metastatic colorectal cancer. J Clin Oncol. 2011;29:
4394-4400.
19. Zeng ZS, Weiser MR, Kuntz E, et al. c-Met gene amplification is
associated with advanced stage colorectal cancer and liver metastases. Cancer
Lett. 2008;265:258-269.
20. Takeuchi H, Kim J, Fujimoto A, et al. HGF activation in colorectal
cancer via c-met receptor regulates IAP proteins. Presented at: 98th Annual
Meeting of the American Association for Cancer Research; April 2007; Los
Angeles, CA.
21. Ding S, Merkulova-Rainon T, Han ZC, et al. HGF receptor upregulation contributes to the angiogenic phenotype of human endothelial cells
and promotes angiogenesis in vitro. Blood. 2003;101:4816-4822.
22. Spigel DR, Ervin TJ, Ramlau R, et al. Final efficacy results from
OAM4558g, a randomized phase II study evaluating MetMAb or placebo in
combination with erlotinib in advanced NSCLC. J Clin Oncol. 2011;29 (suppl;
abstr 7505).
23. Eng C, Eng C, Bendell JC, et al. Phase I results of the randomized,
placebo controlled, phase I/II study of the novel oral c-Met inhibitor, ARQ 197,
irinotecan (CPT-11), and cetuximab (C) in patients (pts) with wild-type (WT)
KRAS metastatic colorectal cancer (mCRC) who have received front-line
systemic therapy. Presented at: Gastrointestinal Cancers Symposium; January 2011; San Francisco, CA.
24. Febbo PG, Ladanyi M, Aldape KD, et al. NCCN Task Force report:
evaluating the clinical utility of tumor markers in oncology. J Natl Compr
Canc Netw. 2011;5:S1-S32.
25. Tejpar S, Bertagnolli M, Bosman F, et al. Prognostic and predictive
biomarkers in resected colon cancer: Current status and future perspectives
for integrating genomics into biomarker discovery. Oncologist. 2010;15:390404.
26. Khleif, SN, Doroshow JH, Hait WN, et al. AACR-FDA-NCI Cancer
Biomarkers Collaborative consensus report: advancing the use of biomarkers
in cancer drug development. Clinical Cancer Res. 2010;16:3299-3318.
200
27. Dienstmann R, Vilar E, Tabernero J. Molecular predictors of response

to chemotherapy in colorectal cancer. Cancer J. 2011;17:114-126.
28. Maughan TS, Adams R, Smith CG, et al. Identification of potentially
responsive subsets when cetuximab is added to oxaliplatin-fluoropyrimidine
chemotherapy (CT) in first-line advanced colorectal cancer (aCRC): mature
results of the MRC COIN trial. J Clin Oncol. 2010;28 (suppl; abstr 3502).
29. Tveit K, Guren B, Glimelius P, et al. Randomized phase III study of
5-fluorouracil/folinate/oxaliplatin given continuously or intermittently with
or without cetuximab, as first-line treatment of metastatic colorectal cancer:
the Nordic VII study (NCT00145314), by the Nordic Colorectal Cancer
Biomodulation Group. Ann. Oncol. 2010;21:viii9 (suppl; abstr LBA.20).
30. Hecht JR, Mitchell E, Chidiac T, et al. A randomized phase IIIB trial of
chemotherapy, bevacizumab, and panitumumab compared with chemotherapy and bevacizumab alone for metastatic colorectal cancer. J Clin Oncol.
2009;27:672-680.
31. Tol J, Koopman. M, Cats M, et al. Chemotherapy, bevacizumab, and
cetuximab in metastatic colorectal cancer. N Engl J Med. 2009;360:563-572.
32. Wang Y, Jatkoe T, Zhang Y, et al. Gene expression profiles and
molecular markers to predict recurrence of Dukes B colon cancer. J Clin
Oncol. 2004;22:1564-1571.
33. Eschrich S, Yang I, Bloom G, et al. Molecular staging for survival
prediction of colorectal cancer patients. J Clin Oncol. 2005;23:3526-3535.
34. Arango D, Laiho P, Kokko A, et al. Gene-expression profiling predicts
recurrence in Dukes C colorectal cancer. Gastroenterology, 2005;129:874884.
35. Barrier A, Boelle PV, Roser R, et al. Stage II colon cancer prognosis
prediction by tumor gene expression profiling. J Clin Oncol. 2006;24:46854691.
36. Gray RG, Quirke P, Handley K, et al. Validation study of a quantitative
multigene reverse transcriptase-polymerase chain reaction assay for assessment of recurrence risk in patients with stage II colon cancer. J Clin Oncol.
2011;29:4611-4619.
37. Clark-Langone KM, Sangli C, Krishnakumar J, et al. Translating
tumor biology into personalized treatment planning: Analytical performance
characteristics of the Oncotype DX Colon Cancer Assay. BMC Genomics.
2007;8:279.
38. OConnell MJ, Lavery I, Yothers G, et al. Relationship between tumor
gene expression and recurrence in four independent studies of patients with
stage II/III colon cancer treated with surgery alone or surgery plus adjuvant
fluorouracil plus leucovorin. J Clin Oncol. 2010;28:3937-3944.
39. Kennedy RD, Bylesjo M, Kerr P, et al. Development and independent
validation of a prognostic assay for stage II colon cancer using formalin-fixed
paraffin-embedded tissue. J Clin Oncol. 2011;29:4620-4626.
40. Salazar R, Roepman P, Capella G, et al. Gene expression signature to
improve prognosis prediction of stage II and III colorectal cancer. J Clin
Oncol. 2011;29:17-24.
41. Tan IB, Tan P. Genetics: an 18-gene signature (ColoPrint) for colon
cancer prognosis. Nat Rev Clin Oncol. 2011;8:131-133.
42. Robles A, Harris CC. Clinical outcomes and correlates of TP53 mutations and cancer. Cold Spring Harb Perspect Biol. 2010;2:a001016.
43. Saltz LB, Niedzwiecki D, Hollis D, et al. Irinotecan fluorouracil plus
leucovorin is not superior to fluorouracil plus leucovorin alone as adjuvant
treatment for stage III colon cancer: results of CALGB 89803. J Clin Oncol.
2007;25:3456-3461.
44. Van Schaeybroeck S, Allen W, Turkington, RC, et al. Implementing
prognostic and predictive biomarkers in CRC clinical trials. Nat Rev Clin
Oncol. 2011;8:222-232.
45. Tabernero J, Baselga J. Multigene assays to improve assessment of
recurrence risk and benefit from chemotherapy in early-stage colon cancer:
has the time finally arrived, or are we still stage locked? J Clin Oncol.
2010;28:3904-3907.
46. Arteaga CL, Baselga J. Impact of genomics on personalized cancer
medicine. Clin Cancer Res. 2012;18:612-618.
47. Engstrom PF, Bloom MG, Demetri GD, et al. NCCN molecular testing
white paper: effectiveness, efficiency, and reimbursement. J Natl Compr Canc
Netw. 2011;6:S1-S16.
48. Wistuba II, Gelovani JG, Jacoby JJ, et al. Methodological and practical
challenges for personalized cancer therapies. Nat Rev Clin Oncol 2011;8:135141.
49. Wu W, Shi Q, Sargent DJ. Statistical considerations for the next
generation of clinical trials. Semin Oncol 2011;38:598-604.
50. McClellan M, Benner J, Schilsky R, et al. An accelerated pathway for
targeted cancer therapies. Nat Rev Drug Discov. 2011;10:79-80.
CURATIVE-INTENT TREATMENT OF COLORECTAL

CANCER METASTASES
CHAIR
Leonard Saltz, MD
New York, NY
SPEAKERS
Stephen B. Solomon, MD
New York, NY
Steven A. Curley, MD
Houston, TX
The Interventional Radiologist Role

in Treating Liver Metastases for
Colorectal Cancer
By Stephen B. Solomon, MD, and Constantinos T. Sofocleous, MD, PhD
Overview: Interventional radiologists (IRs) have an expanding role in the treatment of liver metastases from colorectal
cancer. Increasing data on the ability to treat liver metastases
with locoregional therapies has solidified this position. Ablative approaches, such as radiofrequency ablation and microwave ablation, have shown durable eradication of tumors.
Catheter-directed therapiessuch as transarterial chemoembolization (TACE), drug-eluting beads (DEB), Y90 radioembo-
NTERVENTIONAL RADIOLOGISTS have an expanding

role in the treatment of liver metastases from colorectal
cancer. Understanding the various locoregional treatment
options and their integration into the care of the metastatic
patient is important for good oncologic care. This review will
describe some of these treatment options and their existing
supportive data.
Ablative Therapies
Patients undergoing liver resection of their colorectal metastases have prolonged survival.1 However, only approximately 20% of patients with liver metastases are surgical
candidates.2 Ablation technologies cause focal destruction of
tissue, and when coupled with imaging guidance, this targeted destruction can be aimed at a particular metastasis.
Similar to surgical resection, the goal of ablation is to create
a margin of destruction around the targeted tumor to prevent recurrence. There are a number of ablative tools availableincluding radiofrequency ablation (RFA), microwave
ablation, cryoablation, laser ablation, and focused ultrasound ablationthat rely on extreme temperature conditions of heat or cold to exact the tissue damage.3 A new,
nonthermal technique called irreversible electroporation
uses electric fields to cause cell death without apparently
harming tissue protein architecture that makes up structures such as bile ducts and vessels. This technique may
open up new ablative opportunities near critical structures
that were previously risky using thermal ablation tools,
which could potentially damage these critical structures.
Also, this nonthermal technique may be more effective than
thermal ablation techniques near blood vessels where thermal techniques suffer because of a heat sink effect that
limits how hot the tissue adjacent to a vessel can get.4 More
research is needed to better understand this modality. The
technical differences among all of these techniques is beyond
the scope of this review, suffice it to say, that RFA has been
the most commonly used technique with the most extensive
literature for treating liver metastases.
Ablation techniques are less invasive and consequently
less morbid than surgical resection. Patients can generally
be treated as outpatients with a rapid recovery to normal
activities. Percutaneous ablation procedures can be repeated
if necessary and can be used to salvage recurrences after
resection.5 Although chemotherapy routines are frequently
interrupted by surgical resection for 6 weeks, this same
requirement is not present with percutaneous ablation techniques.6
202
lization, intra-arterial chemotherapy ports, and isolated

hepatic perfusion (IHP)are potential techniques for managing patients with unresectable liver metastases. Understanding the timing and role of these techniques in the
multidisciplinary care of the patient is critical. Implementation
of the IR clinic for consultation has enabled better integration
of these therapies into the patients overall care and has
facilitated improved opportunities for clinical studies.
In a review of nine published articles for patients treated

for unresectable colorectal liver metastases, the 5-year survival rate varied between 14% and 55% (median 30%). For
the subgroup of patients with metastases smaller than or
equal to 4 cm, this 5-year survival rate was better at 18% to
56% (median 34%).3 These are improved survival rates
compared with chemotherapy alone and comparable to patients for whom metastases could be resected.7 There has
not been a randomized study for resectable liver metastases
comparing resection with RFA.8
Test of Time Approach
The test of time approach using percutaneous ablation

was proposed by Livraghi et al in 2003 to allow the biology
of the disease to express itself and to maintain quality of life
in patients with liver metastases from colorectal cancer.9
The concept is that many patients may be able to avoid
unnecessary surgery by undergoing ablation of liver metastases in the interval from the time of diagnosis of liver
metastases to the time of hepatic metastatectomy. The
theory is that RFA can completely treat many metastases
with a small local recurrence rate. Those without recurrence
will have avoided resection. In the delayed period from
diagnosis to surgery, the patient who develops innumerable
metastases would also be able to avoid unnecessary, unbeneficial surgery. For patients with local recurrence after
ablation, there would still be an opportunity to repeat
ablation or perform resection. In summary, this approach
would allow the patient with a successful ablation and the
patient for whom surgery would not have helped because of
an early explosion of metastases to benefit by avoiding
resection. This theory has not been tested in a randomized,
controlled study but, nonetheless, offers an interesting management concept.
Arterial Therapies
Arterial therapies for colorectal liver cancer metastases

can be performed to complement or salvage the effects of
systemic therapy. The application of a locoregional therapy

Address reprint requests to Stephen B. Solomon, MD, Memorial Sloan-Kettering Cancer
Center, 1275 York Ave., H-118, New York, NY; email: solomons@mskcc.org.
1092-9118/10/1-10
RADIOLOGY AND COLORECTAL LIVER METASTASES
can occasionally downstage the patient with inoperable liver

metastases to an operable status. As a whole, intra-arterial
therapies rely on the fact that liver cancers derive their
blood supply predominantly from hepatic arteries, whereas
normal liver parenchyma has a predominantly portal vein
source of blood supply.10
Intra-arterial Hepatic Chemotherapy
Intra-arterial hepatic chemotherapy (IAHC) aims to increase the drug concentration in liver metastases and
thereby improve response rates.11 This approach can be best
applied with drugs having a high first pass effect. Floxuridine with a first pass extraction rate of 95% can increase the
liver dose by 100 to 300 times higher than the systemic
perfusion. Historically, repeated or continuous IAHC has
been delivered by a catheter and pump system requiring
laparotomy. More recently, IAHC can be delivered through
an interventional radiology approach with a subcutaneous
port placed.12 In one study of 36 patients with extensive
nonresectable liver metastases (i.e., 4 metastases in 86%
and bilobar in 91%), IAHC was used with oxaliplatin (100
mg/m2 in 2 hours) plus intravenous 5-fluorouracil (5-FU)
and leucovorin (leucovorin 400 mg/m2 in 2 hours; 5-FU 400
mg/m2 bolus then 2,500 mg/m2 in 46 hours) and cetuximab
(400 mg/m2 then 250 mg/m2/week or 500 mg/m2 every 2
weeks) as first-line treatment. Overall response rate (ORR)
was 90% and disease control rate was 100%. Forty-eight
percent of patients were downstaged enough to undergo an
R0 resection and/or RFA.13
TACE
There are several different regimens used to deliver

TACE. One group using the regimen of cisplatin, doxorubicin, mitomycin C, ethiodol, and polyvinyl alcohol has shown
an ORR of 43%. In this study, the median survival of 33
months from initial diagnosis, 27 months from the time of
liver metastases, and 9 months from the start of chemoembolization suggests a possible improvement over reported
survival time for systemic therapies alone.14 Another group
using mitomycin C alone (52.5%), mitomycin C with gemcitabine (33%), or mitomycin C and irinotecan (14.5%) has
shown an ORR of 63%.15
KEY POINTS
Interventional radiologists play an important role in

the multidisciplinary care of patients with colorectal
cancer with liver metastases.
Ablation therapy can focally destroy liver metastases,
providing long-term survival in select cases.
The test of time approach with ablation before
hepatic metastatectomy can help select which patients will benefit from surgery and which patients
biology will lead to innumerable metastases making
them ultimately not a good surgical candidate.
Intra-arterial therapies with chemoembolization, radioembolization, drug-eluting beads, intra-arterial
chemotherapy, and isolated hepatic perfusion can
play a role in treating unresectable liver metastases
and liver-dominant disease.
Recently, DEB have been developed that allow drug release after the bead has been embolized into the tumor
microcirculation. The advantage of the beads is a reduced
systemic delivery of chemotherapy. One of the drugs that
has been loaded on these beads is irinotecan, which had a
75% reduced systemic plasma level compared with intraarterial irinotecan alone.16
In a randomized study of two courses of DEB with irinotecan (36 patients) compared with eight courses of intravenous irinotecan, 5-FU, and leucovorin (FOLFIRI; 36
patients) for 72 patients who failed at least two lines of
chemotherapy, the response rates were 70% for the DEB
group compared with 30% for the systemic FOLFIRI
group.17 Similarly the 2-year overall survival (OS) was 38%
compared with 18%, and the median OS was 690 days
compared with 482 days. These both favored the DEB arm.
Improvement in quality of life was 60% for the DEB group
compared with 22% for the FOLFIRI group. Finally, overall
cost was lower for the DEB treatment arm.
In a multicenter, single-arm study of 55 patients who
underwent DEB with irinotecan after failing systemic chemotherapy, response rates were 66% at 6 months and 75% at
12 months with an OS of 19 months and a progression-free
survival (PFS) of 11 months.18
Radioembolization
External beam radiation therapy is challenged by the

sensitivity of normal liver to radiation. The dose to treat a
liver tumor is estimated at 70 Gy, while the liver tolerance
dose is 35 Gy. Radioembolization refers to the targeted
intra-arterial delivery of yttrium-90 (90Y) permanently
bound to microspheres. Selective intra-arterial delivery enables doses over 120 Gy to target the tumor without reaching the liver toxicity threshold. An Italian multicenter,
phase II study examined 50 patients with liver-only or
liver-dominant colorectal metastases who failed at least
three lines of systemic chemotherapy with at least one
oxaliplatin and one irinotecan regimen and who underwent
radioembolization. The ORR was 24%, the PFS was 3.7
months with a median OS of 12.6 months, and the 1- and
2-year survival rates were 50.4% and 19.6%, repectively.19
In a randomized study by Van Hazel et al comparing
radioembolization plus systemic chemotherapy with chemotherapy alone as first-line therapy for colorectal liver metastases, the authors found an improved median survival of
29.4 months compared with 11.8 months in the
chemotherapy-alone group.20 In a Belgian multicenter
phase III study, patients were randomly selected to receive
Y90 microspheres in addition to intravenous 5-FU infusion
compared with intravenous 5-FU alone. Median time to
tumor progression was 4.5 compared with 2.1 months,
respectively (hazard ratio 0.51; 95% CI, 0.28 to 0.94; p
0.03).21 There are at least 10 prospective clinical trials of
radioembolization that are open to accrual.9
IHP
IHP allows high-dose chemotherapy to be delivered

through the hepatic artery to the liver without reaching the
systemic circulation. The venous outflow from the liver is
circulated through extracorporeal filters to remove the drug
before the blood is returned to the patients circulation. This
technique opens the possibility of chemotherapy agents to
those that do not necessarily have a high liver first pass
203
SOLOMON AND SOFOCLEOUS
effect. The most common drug that has been investigated

has been melphalan at 200 mg doses. In a phase II study
of 154 patients, the authors found a 50% ORR and a median
PFS and OS of 7.4 and 24.8 months, respectively.22 In
another report comparing IHP with a comparable non-IHP
cohort, the median OS for IHP was 25.0 months compared
with 21.7 months for the control, which was not significantly different. Although the technique historically has
been a surgical one, new angiographic tools have been
developed, making interventional, image-guided implantation feasible.
Conclusion
IRs are playing an increasing role in the management of

patients with colorectal liver metastases. The increasing clinical role of the IR as a consultantseeing patients in clinic
has increased their role as an important contributor to the

multidisciplinary management and the customization of treatment for these patients. The IR armamentarium includes
ablative tools that can focally destroy small numbers of liver
metastases. The use of ablation in a test of time paradigm
may limit unnecessary and morbid resections significantly
contributing to the preservation of patient quality of life.
Locoregional arterially directed therapies for liver-dominant
metastases also allow the physician to manage the unresectable patient to extend disease-free periods and OS and, hopefully, convert him or her to a resection candidate for a potential
cure. As all of these tools and techniques have become available and perfected over the past decade, it will become important for them to be investigated in clinical trials to best
determine the appropriate use in the care of the patient with
colorectal cancer with liver metastases.
Author
Stephen B. Solomon
Employment or
Leadership
Positions
Consultant or
Advisory Role
Althera;
AngioDynamics
(U); Covidien (U);
GE Healthcare;
Johnson &
Johnson
Stock
Ownership
AngioDynamics;
Johnson &
Johnson
Honoraria
Research
Funding
AngioDynamics;
GE Healthcare
Expert
Testimony
Other
Remuneration
Constantinos T. Sofocleous*
REFERENCES
1. House MG, Ito H, Gonen M, et al. Survival after hepatic resection for
metastatic colorectal cancer: Trends in outcomes for 1,600 patients during
two decades at a single institution. J Am Coll Surg. 2010;210:744-752.
2. Adam R, Vinet E. Regional treatment of metastasis: Surgery of colorectal
liver metastases. Ann Oncol. 2004;15:103-106.
3. Pathak S, Jones R, Tang JMF, et al. Ablative therapies for colorectal
liver metastases: A systemic review. Colorectal Dis. 2011;13:252-265.
4. Ben-David E, Appelbaum L, Sosna J, et al. Characterization of irreversible electroporation ablation in in vivo porcine liver. AJR Am J Roentgenol.
2012;198:62-68.
5. Sofocleous CT, Petre EN, Gonen M, et al. CT-guided radiofrequency
ablation as a salvage treatment of colorectal cancer hepatic metastases
developing after hepatectomy. J Vasc Interv Radiol. 2011;22:755-761.
6. Erinjeri JP, Fong AJ, Kemeny NE, et al. Timing of administration of
bevacizumab chemotherapy affects wound healing after chest wall port
placement. Cancer. 2011;117:1296-1301.
7. Machi J, Oishi AJ, Sumida K, et al. Long-term outcome of radiofrequency ablation for unresectable liver metastases from colorectal cancer:
evaluation of prognostic factors and effectiveness in first- and second-line
management. Cancer J. 2006;12:318-326.
8. Mulier S, Ni Y, Jamart J, et al. Radiofrequency ablation versus resection
for resectable colorectal liver metastases: time for a randomized trial? Ann
Surg Oncol. 2008;15:144-157.
9. Livraghi T, Solbiati L, Meloni F, et al. Percutaneous radiofrequency
ablation of liver metastases in potential candidate for resection: the test-oftime approach. Cancer. 2003;97:3027-3035.
10. de Baere T, Deschamps F. Arterial therapies of colorectal cancer
metastases to the liver. Abdom Imaging. 2011;36:661-670.
11. Kemeny NE, Melendez FD, Capanu M, et al. Conversion to resectability
using hepatic artery infusion plus systemic chemotherapy for the treatment of
unresectable liver metastases from colorectal carcinoma. J Clin Oncol.
2009;20:3465-3471.
12. Ganeshan A, Upponi S, Hon L-Q, et al. Hepatic arterial infusion of
chemotherapy: the role of diagnostic and interventional radiology. Ann Oncol.
2008;19:847-851.
13. Malka D, Paris E, Caramella C, et al. Hepatic arterial infusion (HAI) of
204
oxaliplatin plus intravenous (iv) fluorouracil (FU), leucovorin (LV), and

cetuximab for first-line treatment of unresectable colorectal liver metastases
(CRLM) (CHOICE): A multicenter phase II study. J Clin Oncol. 2012;28:15s
14. Albert M, Kiefer MV, Sun W, et al. Chemoembolization of colorectal
liver metastases with cisplatin, doxorubicin, mitomycin C, ethiodol, and
polyvinyl alcohol. Cancer. 2011;117:343-352.
15. Vogl TJ, Gruber T, Balzer JO, et al. Repeated transarterial chemoembolization in the treatement of liver metastases of colorectal cancer: Prospective study. Radiology. 2009;250:281-289.
16. Taylor RR, Tang Y, Gonzalez MV, et al. Irinotecan drug eluting beads
for use in chemoembolization: in vitro and in vivo evaluation of drug release
properties. Eur J Pharm Sci. 2007;30:7-14.
17. Fiorentini G, Aliberti C, Montagnani F, et al. Transarterial chemoembolization of metastatic colorectal carcinoma to the liver adopting polyvinyl
alcohol microspheres loaded with irinotecan compared with FOLFIRI: Evaluation at two years of a phase III clinical trial. ESMO. 2010 (abstr 588).
18. Martin RCG, Joshi J, Robbins K, et al. Hepatic intra-arterial injection
of drug-eluting bead, irinotecan (DEBIRI) in unresectable colorectal liver
metastases refractory to systemic chemotherapy: results of multiinstitutional study. Ann Surg Oncol. 2011;18:192-198.
19. Cosmielli M, Golfieri R, Cagol PP, et al. Multi-centre phase II clinical
trial of yttrium-90 resin microspheres alone in unresectable, chemotherapy
refractory colorectal liver metastases. Br J Cancer. 2010;103:324-331.
20. Van hazel G, Blackwell A, Anderson J, et al. Randomised phased 2 trial
of SIR-Spheres plus fluorouracil/leucovorin chemotherapy versus fluorouracil/
leucovorin chemotherapy alone in advanced colorectal cancer. J Surg Oncol.
2004;88:78-85.
21. Hendlisz A, Van Den Eynde M, Peeters M, et al. Phase III trial
comparing protracted intravenous fluororacil infusion alone or with
yttrium-90 resin microspheres radioembolization for liver-limited metastatic
colorectal cancer refractory to standard chemotherapy. J Clin Oncol. 2010;
28:3687-3694.
22. Van Iersel LBJ, Gelderblom H, Vahrmeijer AL, et al. Isolated hepatic
melphalan perfusion of colorectal liver metastases: outcome and prognostic
factors in 154 patients. Ann Oncol. 2008;19:1127-1134.
Curative-Intent Treatment for Colorectal

Liver Metastases: A Medical Oncologists
Perspective
By Leonard B. Saltz, MD
Overview: Resection or ablation of CRC liver metastases can

be offered with curative intent in some, but not all patients
in whom resection is technically possible. Chemotherapy
can improve the potential for cure to some degree, either in
the adjuvant or neoadjuvant setting, or, in relatively rare
SUBSTANTIAL NUMBER of colorectal cancer (CRC)

patients with oligometastatic disease confined to the
liver are potentially curable by surgical resection and/or
ablation. The degree to which chemotherapy contributes to
the curability of such patients has not been extensively
studied. This article will discuss some of the relevant issues
in patient selection and choice of management strategies
and will review a number of studies that inform discussion
on this topic. Ultimately, in the absence of specific, definitive
data, conclusions must be drawn that are influenced by the
available data but are ultimately based on subjective interpretation of those data. In this context, this article will
present a number of opinions, all of which I believe to be well
supported by data, but all of which are open to discussion
and debate.
Neoadjuvant Compared with Conversion

to Resectability
Chemotherapy in a curative-intent treatment strategy for

CRC can be given in one of three modes; adjuvant, neoadjuvant, or conversion to resectability. Adjuvant chemotherapy
is treatment given after an R0 resection in the hopes of
eradicating residual microscopic disease. There is an important difference between neoadjuvant and conversion chemotherapy, each of which is given before surgical intervention.
Neoadjuvant chemotherapy refers to chemotherapy given
before a planned resection of fully resectable disease. By
definition, the disease that is treated with neoadjuvant
chemotherapy could be resected with no chemotherapy at
all, and no response is needed for the resection to be
accomplished. If tumor shrinkage or regression is required
for resection, then the chemotherapy is not correctly referred to as neoadjuvant; such chemotherapy is referred to
as conversion chemotherapy, designed to convert an unresectable patient to a resectable one.
Conversion to resectability is a complex and important
concept, since some patients who might previously have
been thought to be incurable might have a realistic chance
through this multimodality approach. However, the numbers may be far smaller than are widely appreciated. For
example, Adam et al reported in 2009 that of 184 patients
with initially unresectable CRC metastases who were assessed as having become resectable and then were resected
with an R0 resection, 24 were effectively cured (free of
disease off therapy for 5 or more years).1 So was the
chemotherapy given with curative intent? If we take the
denominator as the 184 patients who underwent an R0
resection, then this gives a cure rate of 16%, as reported.
However, a presumably larger cohort went to the operating
circumstances, by converting truly unresectable disease into

resectable. Careful and realistic patient selection, with an
individualized and realistic assessment of curative potential,
is key to providing each patient with the means to make
realistic treatment choices.
room, some of whom were found at operation to have tumors

that were unresectable; this larger denominator would
shrink the curative percentages somewhat. Consider further
that the 184 patients were culled from the experiences of
14 years of treating patients with metastatic colorectal
cancer at a busy tertiary referral center. Thus, the addition
of chemotherapy to resection resulted in a cure of an otherwise incurable patient in 24 metastatic colorectal patients at
this center over 14 years. It is a remarkable and laudable,
albeit relatively rare, achievement. It is also notable that
Adams group identified favorable risk factors within the
16% of cured patients. Chances of being cured were higher in
patients with three or fewer metastases, patients in whom
the largest tumor before chemotherapy was no greater than
3.0 cm, and in those patients who achieved a pathologic
complete response. Thus, patients having complete resections but with bulky tumors, larger numbers of liver lesions,
or absence of a complete pathologic response had a cure
rate that was substantially smaller than 16%, a point that
needs to be considered when discussing prognosis with such
patients.
It should be noted that most patients who are candidates
for conversion to resectability are those who have disease
confined to the liver, with lesions that are abutting critical
structures, such as major vessels, whereby a resection is not
possible, but in whom tumor shrinkage will create a plane to
render resection possible. Since most CRC metastases that
achieve a clinical complete response are not pathologic
complete responses and are destined to grow back, the
likelihood of resection of residual disease being curative is
small if multiple areas of previous disease are left in.2
Adams group noted that resection of liver metastases in
patients with resected celiac or retroperitoneal lymph nodes,
even if those lymph nodes responded to chemotherapy, had
a poor prognosis with no cures.3
Defining Curative Intent
Terms such as survival, progression-free survival, response, and, clinical benefit, while useful, are often mistaken to equate to more than they actually mean. The words
cure and curative are pure; cure means the cancer is
From the Department of Medicine, Memorial Sloan Kettering Cancer Center, New York,
NY.
Address reprint requests to Leonard B. Saltz, MD, Memorial Sloan Kettering Cancer
Center, 300 East 66th Street, Room 1049, New York, NY 10065; email: saltzl@mskcc.org.
1092-9118/10/1-10
205
LEONARD B. SALTZ
gone forever, no further therapy is needed. Curative, when

referring to a treatment strategy be it medical, surgical, or
combined modalitymeans that the regimen will result, or
has resulted, in a cure. However, the definition of the term
curative intent is a bit more difficult to pin down. First, we
must recognize that long-term survival is not the same as
cure. A patient who is rendered free of known disease but
who then recurs a year later and is alive 5 years after that
has a good long-term survival but has not been cured. It is
true that favorable long-term survival is a highly desirable
outcome, but the degree to which surgery, ablation, and
chemotherapy might or might not improve long-term survival is a discussion for another day; if the tumor is present,
the patient has not been cured, and if a treatment strategy
does not result in cure in some reasonable percentage of
patients, then we cannot realistically consider that we are
using it with curative intent.
There is no agreement on what degree of likelihood of
success constitutes curative intent. If an approach will
result in a cure more than 90% of the time, is that a
treatment given with curative intent? Most of us would say
it is. What about approximately 75% of the time? Fifty
percent? Twenty-five percent? What if a treatment strategy
is curative only 10% or 5% of time? What about 3% or 1%? To
provide some context for this discussion, consider the use of
systemic chemotherapy in patients with surgically unresectable disease. Dy et al reported the long-term outcome of
patients on intergroup protocol N9741.4 A total of 1,508
patients with metastatic CRC received one of three first-line
regimens. Of these, 62 patients achieved a clinical complete
response. However, with a median follow-up of 50 months,
only 10 remained disease-free and could be considered
cured. Thus, at first look, systemic chemotherapy would
appear to have a cure rate of 10 out of 1,508, or 0.66%.
However, if we take the group on N9741 who were treated
with folinic acid, fluorouracil, and oxaliplatin (FOLFOX), 43
of 679, or 6.3%, had a clinical complete response, and if we
assume the same 16% cure rate for those achieving a clinical
complete response, then the cure rate for FOLFOX is 1.0%.
But the chance of achieving a complete response (a necessary first step toward a cure) was increased by having low
tumor burden, a single site of disease, and measurable
rather than evaluable disease, all of which are characteristics that would be present in the vast majority of patients
who are candidates for resection or ablative therapy. Thus,
for the population that might be considered for curative
resection or ablation, greater than 1% of patients would be
expected to be cured by FOLFOX alone. Clearly we would
KEY POINTS
206
Cure is easy to define; curative intent is not.

Some, but not all, patients who can undergo resection
or ablation have a realistic chance at cure.
Adjuvant or neoadjuvant FOLFOX can increase the
likelihood of cure in FOLFOX-naive patients.
Therapies that are inactive in stage III adjuvant are
unlikely to be active in stage IV adjuvant.
Conversion to resectability chemotherapy and neoadjuvant chemotherapy are not the same thing.
not regard FOLFOX as being given with curative intent to

an unresectable patient. So the possibility of a rare cure
is not sufficient to allow us to say a regimen has curative
intent.
We are probably more inclined to perceive a regimen as
being given with curative intent if we think of it as curing
10%, rather than if we consider that it fails to cure 90% of
patients. If more than four out of five or more than nine out
of 10 patients treated by a strategy are destined to not
achieve cure, can we say we are treating with curative
intent? The answer to this question will vary from reader to
reader, and this article is not proposing to set a definition;
however, consideration of these points will be useful to the
reader in making his or her own determination.
Defining Curative Resection
The reason that resection/ablation is an accepted standard

practice without randomized data is that the cure rate with
this approach appears to be substantially higher than what
could realistically be expected from systemic therapy alone.
It should be emphasized that it is the cure rate, and no other
factor, that has led to the universal acceptance of this
approach without a randomized trial. What is the cure rate,
and to what degree does chemotherapy add to that? The
European Organisation for Research and Treatment of Cancer (EORTC) has conducted one of the few large-scale
randomized trials involving liver resection of hepatic metastases of CRC.5 EORTC 40983 was limited to patients with no
more than four liver-only metastases. As it turned out, more
than one-half of the patients had only a solitary liver
metastasis, 14% had three liver lesions, and only 7% had
four liver lesions. Extrapolation to patients with more extensive disease must be done with caution. The three-year
progression-free survival (effectively recurrence-free survival in this resected population) was the primary endpoint.
Since the overwhelming majority of recurrences happen
within the first 3 years, 3-year recurrence-free survival is a
reasonable surrogate for cure. In this EORTC trial, of the
patients actually undergoing resection, 3-year disease-free
survival was 33% in patients treated with surgery only
and 42% in patients treated with surgery plus perioperative FOLFOX (3 months preoperative and 3 months postoperative). Thus, of the patients with CRC with one to four
(mostly one to two) liver metastases who were able to
successfully undergo resection, approximately one-third
were potentially cured. The administration of perioperative
FOLFOX improved that potential cure rate by an absolute
increase of 9%, or a relative increase of 27% (because 9%
is 27% of the baseline of 33%). This 9% absolute increase
was relatively disappointing, considering that FOLFOX increased disease-free survival by 7.2% in patients with stage
III A and B disease and 11.5% in patients with stage IIIC
disease in the MOSAIC trial.6 The reason for this somewhat
disappointing performance by FOLFOX has not been definitively identified. One possible contributing factor is that
although no patient on the EORTC 40983 trial had received
prior oxaliplatin, 42% had received prior fluoropyrimidinebased adjuvant chemotherapy for their primary cancer,
thereby selecting for a relatively fluoropyrimidine-resistant
population.
The fact that the clinical trial was done with the FOLFOX
split into 3 months before and 3 months after surgery should
in no way be construed as creating an obligation to follow
TREATMENT FOR COLORECTAL LIVER METASTASES
that same timing in clinical practice. Delivery of a full 6

months of FOLFOX preoperatively, however, runs the risk
of creating a higher-risk resection because of chemotherapyassociated steatohepatitis. For this reason, for clearly resectable patients, our group often favors up-front resection
followed by postoperative FOLFOX for 6 months. No specific
data exist, however, to settle this question.
Resections with Four or More Liver Lesions
The Memorial Sloan-Kettering Cancer Center hepatobiliary surgical service reviewed the outcome of all patients
with four or more liver lesions resected over a 4-year period
between 1998 and 2002.7 It should be noted that oxaliplatin
was largely unavailable during that time period, so the
contribution of optimal therapies that are inactive in stage
III adjuvant setting are unlikely to be active in stage IV
adjuvant setting. Of 548 patients who underwent successful
resection, 98 had four or more liver metastases. As is the
case with many such surgical reports, the denominator of
how many patients were taken to the operating room with
the intent of full resection but who would end up having
an R1 or R2 resection is not available. Nevertheless, the
long-term follow-up on these patients does provide useful
insights, especially into the curative compared with noncurative outcomes and the important difference between overall survival and cure. Actuarial survival of these 98 patients
was 33% at 5 years. This number was markedly different for
those with four or five lesions (39% actuarial survival at 5
years) compared with 19% 5-year actuarial survival for
those with six or more lesions). These results should be
interpreted within the context of modern imaging and modern chemotherapy, which make older historic comparisons
moot. It is no longer correct or reasonable to say that the
median survival with systemic therapy is 1 year and that
most patients are dead within 2 years, as was the case 20
years ago. In the N9741 trial of systemic therapy, the 5-year
overall survival in patients treated with frontline FOLFOX
was 9.8% (note that 10% of these 5-year survivors did
ultimately undergo metastasectomy, while 90% did not).8
Five-year overall survival would be expected to be higher
with systemic treatment for those patients with good performance status, relatively low volume of disease, and one site
of metastases, so that the population that were candidates
for liver resection would be expected to have a 5-year overall
survival that would be somewhat higher than 10% with
systemic therapy. Thus, the argument that surgery for four
or more lesions improves long-term survival is compelling,
although not airtight. When we look at the cure rate,
however, a different picture emerges. The median diseasefree survival for these 98 patients was 12 months, with a
range of 10 to 15 months, and the actuarial disease-free
survival was 50% at 1 year, 12% at 3 years, and 0% at 5
years. The intriguing and important question from a medical oncology perspective is whether or not active systemic
oxaliplatin-containing adjuvant, neoadjuvant, or conversion
chemotherapy could have had a substantial effect on the
likelihood of cure in these patients.
Role of Non-FOLFOX Regimens
It is important to bear in mind that the role of adjuvant (or

neoadjuvant) chemotherapy is to eradicate micrometastases. This is true for treatment of stage II, stage III, or fully
resected stage IV. As such, there is no compelling argument

for why a micrometastasis in a patient with stage IV disease
should behave any differently from a micrometastasis in
a patient with stage III disease. Numerous studies in the
adjuvant setting of stage III and stage II colon cancer have
demonstrated that irinotecan, bevacizumab, and cetuximab
(and by reasonable extrapolation, panitumumab) are inactive in the adjuvant setting in stage III, with numerous
large-scale trials showing failure to improve the 3-year or
5-year recurrence-free survival rates or the overall survival
rates.9-14 Since it is well demonstrated that all of these
agents have activity against macrometastatic disease, it is
clear that we cannot extrapolate this activity to the micrometastatic setting. On the basis of these data, there is no
compelling argument for why any agents that have been
shown to be inactive in the adjuvant treatment of stage III
colon cancer should be active in the adjuvant or neoadjuvant
treatment of resected stage IV CRC. The only trial that
directly addresses this question is the randomized ACCORD
3 trial of 5-fluorouracil and leucorovin with or without
irinotecan in the adjuvant treatment of resected liver metastases, which showed no benefit for irinotecan.15 Note that
this statement regarding irinotecan, bevacizumab, and antiepidermal growth factor receptor (EGFR) monoclonal antibodies does not apply to conversion chemotherapy; if the
goal is to shrink a tumor you can see, then irinotecan,
bevacizumab, and the anti-EGFR monoclonal antibodies
are all reasonable considerations. However, if the goal is to
eradicate micrometastatic disease and increase the cure
rate, then in my opinion, they are not.
Role of FOLFOX in Patients with Prior FOLFOX
Adjuvant Treatment
When metachronous liver metastases develop after adjuvant chemotherapy, it must be recalled that the cells that
gave rise to those metastases were present during that
adjuvant chemotherapy and were therefore, by definition,
resistant to it. It is thus exceedingly unlikely that this same
chemotherapy that failed to eradicate micrometastases in
the first treatment will have activity on residual micrometastases this second time around after hepatic resection.
Remember, you are not treating the resected metastases;
you are treating residual micrometastases. Thus, adjuvant
or neoadjuvant FOLFOX does not appear to be a reasonable
maneuver in a patient who had previously received adjuvant
FOLFOX and who, by virtue of the development of metastases, failed that adjuvant therapy. If a patient has received
either no prior therapy or fluorpyrimidine adjuvant only,
then adjuvant or neoadjuvant FOLFOX would be indicated.
It is noteworthy that the EOTRC 40983 trial excluded
patients with prior oxaliplatin adjuvant chemotherapy. In a
patient who has had prior adjuvant FOLFOX (or capecitabine/oxaliplatin [Cape/Ox]), there unfortunately is not an
active adjuvant or neoadjuvant systemic treatment to offer,
and therefore I do not believe that any adjuvant or neoadjuvant systemic chemotherapy is indicated in such patients.
Note that FOLFOX and Cape/Ox can be used interchangeably. There is no evidence to support that one is superior to
the other, that one can rescue the other, or that capecitabine
has activity after failure of an infusional 5-fluorouracilcontaining regimen.) The fact that FOLFIRI (fluorouracil,
leucovorin, irinotecan) with or without either bevacizumab
or an anti EGFR monoclonal might be able to shrink a tumor
207
LEONARD B. SALTZ
is irrelevant if it is already resectable (and so neoadjuvant

and not conversion chemotherapy) since we have no reason
to believe that these agents or regimens will have activity
against the residual micrometastases that we are targeting.
Conclusion
Resection or ablation of CRC liver metastases can be

offered with curative intent in some, but not all patients in
whom resection is technically possible. Chemotherapy can
improve the potential for cure to some degree, either in the
adjuvant or neoadjuvant setting, or, in relatively rare circumstances, by converting truly unresectable disease into
resectable. Careful and realistic patient selection, with an
individualized and realistic assessment of curative potential, is key to providing each patient with the means to make
realistic treatment choices. Ultimately it is anticipated that
molecular and immunologic assessments of individuals and
their tumors will help guide rational selection of strategies
to increase the curative potential of combined modality
management of CRC liver metastases.
Author
Leonard B. Saltz
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Bristol-Myers
Squibb (U);
Genentech;
ImClone
Systems; Merck;
Novartis; Pfizer;
Roche
Research
Funding
Expert
Testimony
Other
Remuneration
Amgen; Bayer;
Bristol-Myers
Squibb;
Genentech;
ImClone
Systems; Merck;
Pfizer; Roche;
Taiho
Pharmaceutical
REFERENCES
1. Adam R, Wicherts DA, de Haas RJ, et al. Patients with initially
unresectable colorectal liver metastases: Is there a possibility of cure? J Clin
Oncol. 2009;27:1829-1835.
2. Benoist S, Brouquet A, Penna C, et al. Complete response of colorectal
liver metastases after chemotherapy: Does it mean cure? J Clin Oncol.
2006;24:3939-3945.
3. Adam R, de Haas RJ, Wicherts DA, et al. Is hepatic resection justified
after chemotherapy in patients with colorectal liver metastases and lymph
node involvement? J Clin Oncol. 2008;26:3672-3680.
4. Dy GK, Krook JE, Green EM, et al. Impact of complete response to
chemotherapy on overall survival in advanced colorectal cancer: Results from
Intergroup N9741. J Clin Oncol. 2007;25:3469-3474.
5. Nordlinger B, Sorbye H, Glimelius B, et al. Perioperative chemotherapy
with FOLFOX4 and surgery versus surgery alone for resectable liver metastases from colorectal cancer (EORTC Intergroup trial 40983): a randomised
controlled trial. Lancet. 2008;22:1007-1016.
6. Andre T, Boni C, Navarro M, et al. Improved overall survival with
oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or
III colon cancer in the MOSAIC trial. J Clin Oncol. 2009;27:3109-3116.
7. Kornprat P, Jarnagin WR, Gonen M, et al. Outcome after hepatectomy
for multiple (four or more) colorectal metastases in the era of effective
chemotherapy. Ann Surg Oncol. 2007;14:1151-1160.
8. Sanoff HK, Sargent DJ, Campbell ME, et al. Five-year data and
prognostic factor analysis of oxaliplatin and irinotecan combinations for
advanced colorectal cancer: N9741. J Clin Oncol. 2008;26:5721-5727.
208

treatment for stage III colon cancer: Results of CALGB 89803. J Clin Oncol.
2007;25:3456-3461.
10. Ychou M, Raoul J-L, Douillard J-Y, et al. A phase III randomised trial
of LV5FU2 irinotecan versus LV5FU2 alone in adjuvant high-risk colon
cancer (FNCLCC Accord02/FFCD9802). Ann Oncol. 2009;20:674-680.
11. Van Cutsem E, Labianca R, Bodoky G, et al. Randomized phase III trial
comparing biweekly infusional fluorouracil/leucovorin alone or with irinotecan in the adjuvant treatment of stage III colon cancer: PETACC-3. J Clin
Oncol. 2009;27:3117-3125.
12. Allegra CJ, Yothers G, OConnell MJ, et al. Phase III trial assessing
bevacizumab in stages II and III carcinoma of the colon: Results of NSABP
protocol C-08. J Clin Oncol. 2011;29:11-16.
13. De Gramont A, Van Cutsem E, Tabernero J, et al. AVANT: Results
from a randomized, three-arm multinational phase III study to investigate
bevacizumab with either XELOX or FOLFOX4 versus FOLFOX4 alone as
adjuvant treatment for colon cancer. J Clin Oncol. 2011;29 (suppl 4; abstr
362).
14. Alberts SR, Sargent DJ, Smyrk CJ, et al. Adjuvant mFOLFOX6 with or
without cetuxiumab (Cmab) in KRAS wild-type (WT) patients (pts) with
resected stage III colon cancer (CC): Results from NCCTG Intergroup Phase
III Trial N0147. J Clin Oncol. 2010;28 (suppl; abstr CRA3507).
15. Ychou M, Hohenberger W, Thezenas S, et al. A randomized phase III
study comparing adjuvant 5-fluorouracil/folinic acid with FOLFIRI in patients following complete resection of liver metastases from colorectal cancer.
Ann Oncol. 2009;20:1964-1970.
Surgical Treatment of Colorectal Cancer

Liver Metastases
By Steven A. Curley, MD
Overview: Treatment strategies for patients with stage IV

colorectal cancer have changed markedly in the last decade.
Patients with colorectal cancer metastases to the liver have
always been a fascinating group to consider biologically and
for local-regional treatment strategies. In the late 1980s
through the 1990s, resection was performed for a select
subset of patients who had resectable disease. However, a
high proportion of patients had bilobar unresectable disease
and were treated with either 5-fluorouracil based systemic
HE LIVER, second only to lymph nodes as the most

common site of metastasis from other solid tumors, is a
common site of colorectal cancer (CRC) metastasis. A subset
of patients with metastatic CRC have liver-only disease or
liver and resectable extrahepatic disease and are candidates
for surgical treatment. Resection of liver metastases is
associated with minimal mortality rates in these patients,
and resection can provide long-term survival benefit for a
substantial proportion of patients. Beliefs regarding which
patients with hepatic metastases are candidates for resection or other surgical treatment must be reevaluated
through careful examination of extant data and also by
considering the effect of improved diagnostic imaging techniques, new cytotoxic and targeted molecular therapies, and
improved patient selection and treatment criteria.
Multiple Liver Metastases
Recent studies have demonstrated that resection for four

or more hepatic metastases from CRC can yield long-term
survival for some patients. For example, Minagawa and
colleagues1 reported that although survival was improved
for patients with a solitary metastasis compared with those
with multiple metastases, 10-year survival rates in excess of
20% were achieved for people who had resection of as many
as 20 lesions. This study is the only report, to our knowledge,
of a 10-year survival rate after resection of four or more CRC
liver metastases. Our group has examined the prospective
database at The University of Texas M. D. Anderson Cancer
Center, which includes more than 1,600 patients who had
resection of CRC liver metastases since 1995. Among 159
patients with four to 12 liver metastases that were treated
with surgery (resection or resection plus radiofrequency
ablation [RFA]), the 5-year disease-free survival rate was
22% and the 5-year overall survival rate was 51%.2 The
median overall survival for these 159 patients was 62.1
months. The favorable long-term survival data in our study
relate to the fact that these patients were carefully selected.
No patient had radiographic or intraoperative evidence of
extrahepatic disease at the time of surgical treatment of the
hepatic metastases, most (89.9%) received neoadjuvant chemotherapy, most (72.7%) had reduction in tumor volume
after preoperative chemotherapy, and all patients underwent thorough intraoperative ultrasonography to detect
additional small lesions not evident on preoperative imaging
studies. On multivariate analysis of predictive factors, only
response to neoadjuvant chemotherapy remained an independent predictor of overall survival. Our data suggest that
chemotherapy or implanted hepatic arterial infusion pumps.

The advent of the new millennium was associated with the
availability of several new cytotoxic and biologic agents active
in colorectal cancer. These agents have completely changed
the approach to the treatment of patients with colorectal
cancer liver metastases and thus have increased the complexity of the decision-making process for treatment of these
patients.
tumor biology (response to cytotoxic therapy) rather than

morphologic criteria (tumor number or size) determines
long-term prognosis.
Unresectable Liver Metastases or
Extrahepatic Disease
The presence of extrahepatic disease has clearly been

demonstrated to be a negative prognostic factor in retrospective studies of CRC liver metastases treated with surgery
alone.3-6 The surgical dogma to not perform resection in
these patients has been challenged in the modern era of
multimodality management for CRC metastatic to the liver.
Adam and colleagues7 reported a single-institution experience of 1,439 consecutive patients with CRC liver metastases at the time of diagnosis. At initial presentation, 1,104
patients (77%) were deemed to have unresectable disease
and were treated by chemotherapy, whereas 335 (23%) had
resectable disease and underwent surgical treatment. There
were two interesting findings from this study. First, among
the 1,104 patients with nonresectable disease, 138 (12.5%)
had a substantial reduction in the tumor volume with
chemotherapy and became candidates for a potentially curative surgical procedure. These 138 patients were treated
with either resection alone or resection combined with
thermal ablation. The overall 5- and 10-year survival rates
for these 138 patients were 33% and 23%, respectively, with
disease-free survival rates at the same intervals of 22% and
17%, respectively. The second important finding was that
extrahepatic tumor was present at the time of surgical
treatment in 52 (38%) of these 138 patients. All of these
patients underwent surgical treatment of the liver metastases along with resection of lung, lymph node, peritoneal, or
other metastatic disease. When it was possible to treat all
metastatic disease surgically, there was no marked decrease
in the long-term survival probability compared with that for
patients who underwent treatment of liver metastases
alone.
From the Department of Surgical Oncology, University of Texas M. D. Anderson Cancer

Address reprint requests to: Steven A. Curley, MD, Division of Surgery, University of
Texas M. D. Anderson Cancer Center, 1400 Pressler St., Unit 1447, Houston, TX 77030;
email: scurley@mdanderson.org.
1092-9118/10/1-10
209
STEVEN A. CURLEY
Fig. 1. Treatment of 418 consecutive

patients with liver-only CRC metastases
at M. D. Anderson Cancer Center. A total
of 348 patients (83.3%) were treated
for cure with hepatic resection only, RFA
plus resection, or RFA only. Seventy patients were found to have disease too
extensive for curative therapy and underwent chemotherapy (systemic, intraarterial chemotherapy via a hepatic
artery infusion pump placed at the index laparotomy or intra-arterial plus
systemic chemotherapy).
Abbreviations: CRC, colorectal cancer;
RFA, radiofrequency ablation.
Resection and Thermal Ablation of Liver Metastases
We have reported the results of a prospective study of

outcomes after surgical treatment of CRC liver metastases
using hepatic resection only, RFA plus resection for multiple
tumors, or RFA only when patients had tumor in an unresectable site in the liver. The study group consisted of 348
consecutive patients with CRC liver metastases and no
extrahepatic disease who were treated for cure with hepatic
resection with or without RFA and 70 patients who were
found at laparotomy to have liver-only disease (but were not
candidates for curative treatment based on involvement of
too many liver segments) as outlined in Fig. 1.8 An example
of a case in which a combined approach of resection and
ablation was used is illustrated in Fig. 2.
The 5-year overall survival rate among the 348 patients
treated with curative intent was 44%. Survival with resection alone (58% at 5 years) was significantly greater than
that with either resection plus ablation (28%) or RFA
alone (19%, p 0.001), with no marked difference among
the approaches, including RFA. The overall 4-year survival rates after resection, RFA plus resection, and RFA
KEY POINTS
210
Management of colorectal cancer liver metastases

requires a multidisciplinary team approach.
Patients should be considered candidates for resection whenever possible, regardless of the number of
metastatic lesions.
Modern systemic chemotherapy can produce response sufficient to convert unresectable liver metastases to resectable disease in some patients.
Chemotherapy can be hepatotoxic, so careful selection of neoadjuvant therapy agents and duration
must be considered.
Resection provides better long-term survival probability compared with thermal tumor ablation
techniques.
alone were 65%, 36%, and 22%, respectively (p 0.0001);

however, 4-year survival rates with resection plus RFA
and RFA only were still substantial and significantly improved over that achieved with chemotherapy alone (5%) in
the 70 patients not treated with resection or RFA (p
0.0017).
Systemic Chemotherapy
Systemic chemotherapy can convert unresectable CRC

liver metastases to resectable disease in some patients, and
response to neoadjuvant chemotherapy in patients with
resectable disease is an important prognostic indicator of
improved probability of long-term survival.2,9,10 The use of
neoadjuvant chemotherapy in patients with resectable CRC
liver metastases must be carefully considered in multidisciplinary treatment planning involving medical oncologists
and hepatobiliary surgical oncologists. Neoadjuvant chemotherapy regimens that are based on oxaliplatin (Eloxatin;
Sanofi Aventis) or irinotecan (Camptosar; Pfizer) can produce specific types of liver injury. Oxaliplatin-based chemotherapy was reported to cause sinusoidal obstruction, venoocclusive lesions in the microvasculature of nontumoral
liver, and perisinusoidal fibrosis in more than half of the
patients receiving neoadjuvant therapy.11 Both irinotecan
and oxaliplatin treatment have been found to produce nonalcoholic steatohepatitis (NASH), and the severity of NASH
is greater in patients who are also obese.12 Hepatic steatosis
or steatohepatitis was identified as an independent variable
that predicted increased perioperative morbidity and mortality rates in a retrospective study of 135 patients undergoing resection of CRC liver metastases.13 Data suggest that
steatosis impairs hepatic regeneration after resection, and
the rates of liver insufficiency and hyperbilirubinemia are
higher for patients with hepatic steatosis than for patients
with normal livers.
A multi-institution review of 406 patients who underwent
resection of CRC liver metastases and had the severity of
NASH and nonalcoholic fatty liver disease scored on a
standardized system was completed recently.14,15 Preoperative chemotherapy was administered to 241 patients (61%),
and 158 patients (39%) received no neoadjuvant treatment.
SURGICAL TREATMENT OF COLORECTAL LIVER METASTASES
Fig. 2. The patient had a large tumor in the left

lateral segment (image on the left); also visible is one
of four smaller tumors in the right lobe. The large
tumor was resected, as was one of the smaller tumors near the dome (image on the right). The other
tumors were ablated, with ablation including a surrounding zone of the parenchyma to ensure a negative margin.
Patients who received oxaliplatin were found to have sinusoidal dilation at a much higher frequency compared with
patients who did not receive chemotherapy (p 0.001). In
contrast, the incidence of NASH was approximately 20%
among patients who received irinotecan, which was significantly greater than the 4% rate among patients who received no chemotherapy (p 0.001). The most striking
finding of this study was that only NASH was associated
with an increased 90-day mortality rate (14.7%) after liver
resection, whereas other types of liver injury or normal liver
were associated with a postoperative mortality rate of 1.6%
(p 0.001). Clearly, careful consideration must be given to
the type and duration of neoadjuvant chemotherapy to be
delivered to patients. In patients who have preoperative
imaging or intraoperative evidence of either nonalcoholic
fatty liver disease or NASH, consideration should be given to
obtaining a core liver biopsy to assess the severity of steatohepatitis. The presence of severe NASH should lead the
surgeon to question the safety of a major liver resection and
lower the threshold to consider preoperative portal vein
embolization in an attempt to reduce postoperative morbidity and liver failure rates.
A subset of patients have borderline resectable or unresectable liver metastases that progress during first-line
systemic chemotherapy. In addition, some patients who
receive neoadjuvant chemotherapy before a planned liver
resection have disease progression demonstrated on subsequent imaging examinations. Frequently, these patients
will receive second-line chemotherapy. We recently evaluated 60 patients who underwent resection of CRC liver
metastases after receiving two or more different systemic
chemotherapy regimens.16 These patients tended to have
more advanced CRC liver metastases as evidenced by a
mean SD number of tumors of 4 3.5 and a mean SD
maximum size of 5 3.2 cm. All the patients received
irinotecan or oxaliplatin and frequently had been treated
sequentially with regimens that contained both of these
agents. Despite receiving a mean SD of 17 8 cycles
of chemotherapy, the postoperative morbidity rate was
33%, and the 90-day mortality rate was 3%. The 1-, 3-, and
5-year overall survival rates were 83%, 41%, and 22%,
respectively. Thus, long-term survival after resection of CRC
liver metastases in patients who have received second-line
chemotherapy can produce a modest rate of long-term survival.
Synchronous CRC Liver Metastases
As many as 25% of patients with primary CRC will

present with synchronous liver metastases. Historically,
synchronous liver metastases from CRC has portended a
poor prognosis. The traditional strategy for managing
these patients was resection of the primary tumor followed

by systemic chemotherapy and then consideration of liver
resection if disease was considered to be resectable. This
strategy arose in the 20th century before the advent of
effective modern cytotoxic and biologic agents for advanced
CRC. Our group recently evaluated 156 consecutive patients with synchronous resectable liver metastases and
intact primary tumor from CRC.17 The patients in this
series had nonobstructing primary tumors and received
preoperative systemic chemotherapy. A subset of these
patients had a notable antitumor response in the primary
tumor, and the decision was made to perform a reverse
strategy by resecting the liver metastases before surgical
treatment of the primary tumor. When considering the
entire group, this reverse approach was used in 19% of
the patients, whereas the remaining patients underwent
either a classic approach of resection of the primary
tumor followed by later resection of the CRC liver metastases or a combined resection of the primary tumor and
liver metastases in one operation. The median number of
liver metastases was considerably higher in the reverse
strategy group, but no substantial difference was found
in postoperative morbidity or mortality rates when the
reverse strategy was compared with either the classic or
combined surgical approach groups. The 5-year survival
rates for the classic, combined, and reverse strategy approach groups were 48%, 55%, and 39%, respectively. These
data further underscore the importance of multidisciplinary
team planning and management of patients with CRC liver
metastases.
Conclusion
The criterion standard in the treatment of CRC liver

metastases remains complete resection. Complete resection
can be performed safely with low morbidity and mortality
rates. Local hepatic tumor ablation techniques have a role in
the treatment of patients with unresectable disease. However, it needs to be emphasized that no data currently exist
to suggest that either RFA or RFA combined with resection
provides survival comparable to that with complete resection and that long-term survival rates for patients treated
with thermal ablation techniques are still being established.
Aggressive surgical management combined with neoadjuvant and adjuvant chemotherapy regimens has yielded
5-year survival rates that exceed 50%. All patients with CRC
liver metastases should be evaluated by a multidisciplinary
team consisting of hepatobiliary surgical oncologists, medical oncologists, pathologists, and diagnostic imaging specialists to optimize the management of patients with stage IV
disease.
211
STEVEN A. CURLEY
Author
Steven A. Curley*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Minagawa M, Makuuchi M, Torzilli G, et al. Extension of the frontiers of
surgical indications in the treatment of liver metastases from colorectal
cancer: long-term results. Ann Surg. 2000;231:487-499.
2. Pawlik TM, Abdalla EK, Ellis LM, et al. Debunking dogma: surgery for
four or more colorectal liver metastases is justified. J Gastrointest Surg.
2006;10:240-248.
3. Hughes KS, Simon R, Songhorabodi S, et al. Resection of the liver for
colorectal carcinoma metastases: a multi-institutional study of patterns of
recurrence. Surgery. 1986;100:278-284.
4. Nordlinger B, Guiguet M, Vaillant JC, et al. Surgical resection of
colorectal carcinoma metastases to the liver. A prognostic scoring system to
improve case selection, based on 1568 patients. Association Francaise de
Chirurgie. Cancer. 1996;77:1254-1262.
5. Fong Y, Fortner J, Sun RL, et al. Clinical score for predicting recurrence
after hepatic resection for metastatic colorectal cancer: analysis of 1001
consecutive cases. Ann Surg. 1999;230:309-318; discussion 318-321.
6. Scheele J, Altendorf-Hofmann A, Grube T, et al. Resection of colorectal
liver metastases. What prognostic factors determine patient selection? [in
German]. Chirurg. 2001;72:547-560.
7. Adam R, Delvart V, Pascal G, et al. Rescue surgery for unresectable
colorectal liver metastases downstaged by chemotherapy: a model to predict
long-term survival. Ann Surg. 2004;240:644-657; discussion 657-658.
8. Abdalla EK, Vauthey JN, Ellis LM, et al. Recurrence and outcomes
following hepatic resection, radiofrequency ablation, and combined resection/
ablation for colorectal liver metastases. Ann Surg. 2004;239:818-825.
9. Tanaka K, Adam R, Shimada H, et al. Role of neoadjuvant chemotherapy
212
in the treatment of multiple colorectal metastases to the liver. Br J Surg.

2003;90:963-969.
10. Allen PJ, Kemeny N, Jarnagin W, et al. Importance of response to
neoadjuvant chemotherapy in patients undergoing resection of synchronous
colorectal liver metastases. J Gastrointest Surg. 2003;7:109-115
11. Rubbia-Brandt L, Audard V, Sartoretti P, et al. Severe hepatic sinusoidal obstruction associated with oxaliplatin-based chemotherapy in patients
with metastatic colorectal cancer. Ann Oncol. 2004;15:460-466.
12. Fernandez FG, Ritter J, Goodwin JW, et al. Effect of steatohepatitis
associated with irinotecan or oxaliplatin pretreatment on resectability of
hepatic colorectal metastases. J Am Coll Surg. 2005;200:845-853.
13. Behrns KE, Tsiotos GG, DeSouza NF, et al. Hepatic steatosis as a
potential risk factor for major hepatic resection. J Gastrointest Surg. 1998;2:
292-298.
14. Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a
histological scoring system for nonalcoholic fatty liver disease. Hepatology.
2005;41:1313-1321.
15. Vauthey JN, Pawlik TM, Ribero D, et al. Chemotherapy regimen
predicts steatohepatitis and an increase in 90-day mortality after surgery for
hepatic colorectal metastases. J Clin Oncol. 2006;24:2065-2072.
16. Brouquet A, Overman MJ, Kopetz S. Is resection of colorectal liver
metastases after a second-line chemotherapy regimen justified? Cancer.
2011;117:4484-4492.
17. Brouquet A, Mortenson MM, Vauthey J-N. Surgical strategies for
synchronous colorectal liver metastases in 156 consecutive patients: classic,
combined or reverse strategy? J Am Coll Surg. 2010;201:934-941.
RECTAL CANCER: NEW PARADIGMS BEYOND

STANDARD CHEMOTHERAPY AND RADIATION
CHAIR
David P. Ryan, MD
Massachusetts General Hospital Cancer Center
Boston, MA
SPEAKERS
Martin R. Weiser, MD
New York, NY
Karyn A. Goodman, MD
New York, NY
Minimally Invasive Surgery of Rectal Cancer:

Current Evidence and Options
By Atthaphorn Trakarnsanga, MD, and Martin R. Weiser, MD
Overview: Minimally invasive surgery (MIS) of colorectal

cancer has become more popular in the past two decades.
Laparoscopic colectomy has been accepted as an alternative
standard approach in colon cancer, with comparable oncologic outcomes and several better short-term outcomes com-
pared to open surgery. Unlike the treatment for colon cancer,

however, the minimally invasive approach in rectal cancer
has not been established. In this article, we summarize the
current status of MIS for rectal cancer and explore the various
technical options.
standard treatment in the United States.10 Based on evidence presented in European trials, however, several countries in Europe and Asia have endorsed MIS as an
alternative to standard open surgery in rectal cancer.
IS HAS become the standard procedure in some

operations, such as cholecystectomy and appendectomy. Several short-term benefits are associated with MIS,
including reduced postoperative pain and decreased need for
analgesic drugs, more cosmetically pleasing incisions,
shorter length of hospital stay, and earlier return to functionality. These are the major reasons for the increasing
popularity of the minimally invasive surgical approach.
Jacobs and colleagues1 first reported a case series of
successful laparoscopic colectomies in 1990. In the beginning, its application was limited to treatment of benign
lesions. However, the use of MIS in colon cancer raised
concerns about the possibility of port site metastases and
inadequate oncologic resection. Early reports indicated that
the incidence of port site metastasis was approximately 4%
in the laparoscopic group compared with 1% in the open
group.2,3 This contentious question was resolved by a large
meta-analysis of several controlled studies, which concluded
that the incidence of port site metastasis was 0.87% in the
laparoscopic group compared with 0.34% in the open group
(p 0.16).4 Correspondingly, concerns regarding oncologic
outcome were dispelled by several well-designed large, randomized, controlled studies that demonstrated the equivalence of laparoscopic colectomy to the conventional open
approach in terms of recurrence and survival.5-8 Undoubtedly, laparoscopic colectomy is now considered an alternative standard approach in the treatment of colon cancer.
Obviously, surgery of the rectum is more challenging for
several reasons, mostly because dissection of the mesorectal
plane is limited by the confines of the bony pelvis and the
goal of preserving the autonomic nerves. Moreover, neoadjuvant treatment sequelae, especially those secondary to
radiotherapy, may affect the surgical field. Thus, the role of
MIS for rectal cancer is still controversial and substantial
evidence is lacking. We summarize the current status of MIS
in rectal cancer, the findings of contemporary published
studies, and the appropriate application of various existing
techniques.
Laparoscopic Rectal Cancer Surgery:
Where Are We Now?
At present, the American Society of Colon and Rectal

Surgeons (ASCRS) has not endorsed laparoscopic proctectomy for rectal cancer because of concerns about achieving
adequate mesorectal excision and clear surgical margins
using this technique.9 ASCRS has encouraged well-designed
trials to examine the safety, efficacy, and benefits of MIS in
rectal cancer surgery, especially in regard to long-term
oncologic outcomes. A majority of these prospective randomized, controlled trials are ongoing, and the results have not
yet been defined. Thus, MIS for rectal cancer has not become
214
United Kingdom MRC CLASICC Trial
The United Kingdom Medical Research Council Trial of

Conventional vs. Laparoscopic-Assisted Surgery in Colorectal Cancer (United Kingdom MRC CLASICC) was a prospective randomized trial that included both patients with colon
cancer and patients with rectal cancer.7 Of the 794 enrolled
patients, 381 had rectal cancer and the conversion rate was
34%. Within the actual treatment group, 87 patients underwent open total mesorectal excision (TME) and 189 had
laparoscopic-assisted TME. The primary endpoints were
rate of positive-circumferential and longitudinally resected
margins. No significant difference in either margin of resection was identified when comparing the two procedures.
Among patients undergoing anterior resection, the rate of
positive circumferential resection (CRM) was slightly higher
in the laparoscopic than in the open group (12% vs. 6%; p
0.80). In-hospital mortality rates between laparoscopic and
open surgery were not significantly different (4% vs. 5%; p
0.57), but the mortality rate was higher in patients who
were converted to open surgery.
Regarding long-term outcomes, 5-year results of the
United Kingdom MRC CLASICC have already been published.11 There was no significant difference in 5-year overall survival (OS) between the laparoscopic and open groups
(60.3% vs. 52.9%; p 0.132). Five-year disease-free survival
did not differ either (53.2% vs. 52.1%, respectively; p
0.953). In converted patients, the overall survival rate was
significantly worse, compared to the other patients who
initially received randomized treatment (p 0.05). There
was no difference in rate of local recurrence between patients who had laparoscopic compared with open anterior
resection, although as previously noted, the rate of positive
CRM was slightly higher in the laparoscopic group (9.4% vs.
7.6%; p 0.74). No data was shown on local recurrence in
patients who had abdominoperineal resection.
From the Department of Surgery, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand; Department of Surgery, Memorial Sloan-Kettering Cancer Center,
New York, NY.
Address reprint requests to Martin R. Weiser, MD, Department of Surgery, Memorial
Sloan-Kettering Cancer Center, 1275 York Ave., Room C-1075, New York, NY, 10065; email:
weiser1@mskcc.org.
1092-9118/10/1-10
MINIMALLY INVASIVE SURGERY OF RECTAL CANCER
An important finding of this study was the high rate of

conversion from laparoscopic to open surgery intraoperatively (34%), and the poorer outcomes in patients who were
converted. The reasons for intraoperative conversion included tumor fixation, patient obesity, technical difficulty,
practitioner uncertainty regarding adequate tumor clearance, and uncertainty regarding anatomic landmarks.
COREAN Trial
The Comparison of Open vs. laparoscopic surgery for mid

and low REctal cancer After Neoadjuvant chemoradiotherapy (COREAN) trial is a multicenter randomized, controlled
trial comparing open with laparoscopic surgery in Korean
patients with mid- or low-rectal cancer (i.e., tumor within 9
cm of the anal verge) following preoperative chemoradiotherapy.12 The primary endpoint was 3-year disease-free
survival. Seven high-volume surgeons were qualified to
enroll in the study. Three hundred and forty patients with
T3/N0 2 rectal cancer were randomly assigned to laparoscopic and open groups in a ratio of 1:1. The conversion rate
was 1.2% (2 in 170 patients), secondary to difficult dissection
in the narrow pelvis and/or bleeding. The rates of coloanal
anastomosis and abdominoperineal resection were equivalent between the laparoscopic and open surgery groups
(19.4% vs. 19.4% and 11.2% vs. 14.1%, respectively; p
0.7085). Operative time was significantly longer in the
laparoscopic group (244.9 min vs. 197 min, p 0.006), but
estimated blood loss was significantly less (217.5 mL vs. 200
mL, p 0.0001). The rate of positive CRM did not differ
significantly between the two groups (2.9% laparoscopic vs.
4.1% open; p 0.770), nor did distal resection margin or
pathologic stage. The short-term benefits of laparoscopic
surgery were notable: speedier return of bowel function, less
postoperative pain, and less need for/use of analgesic drugs.
The COREAN trial is the first study to compare laparoscopic with open TME in patients who have undergone
neoadjuvant chemoradiotherapy. The promising short-term
outcomes of laparoscopic surgery for rectal cancer include a
low rate of conversion and small number of positive CRMs.
Both may be the result of neoadjuvant treatment, combined
with optimal resection technique employed by experienced
surgeons. Long-term follow-up data regarding recurrence
and survival are awaited, and will be important factors in
this large trial.
COLOR II Trial
The European Colon Cancer Laparoscopic or Open Resection (COLOR) II trial is an international randomized, multicenter study comparing the outcomes of curative
laparoscopic and conventional open surgery for rectal cancer.13 The primary endpoint is 3-year locoregional recurrence. Secondary endpoints are recurrence-free and OS at 3,
5, and 7 years; rate of port- and wound-site metastases;
distant metastases; morbidity and mortality within 8 weeks
following resection; macroscopic evaluation of the resected
specimen; quality of life; and cost. Currently, the final
results have been obtained but are not yet reported.
JCOG 0404 Trial
The Japanese Clinical Oncology Group Study (JCOG

0404) is a multicenter randomized, controlled trial comparing laparoscopic and open surgery in Japanese patients with
colon or rectal cancer.14 This study was activated in October
2004. The planned sample size is 818 cases; 409 cases per
arm with 5 years of follow-up after 3 years of accrual. The
primary endpoint is OS. Secondary endpoints are relapsefree survival, short-term clinical outcomes, adverse events,
and rates of conversion.
ACOSOG-Z6051 Trial
KEY POINTS
Minimally invasive colon surgery for cancer has become standard.

Rectal surgery is more challenging than colon surgery because of anatomic limitation of the pelvis and
need to preserve autonomic nerves while maintaining
a clear margin.
Minimally invasive rectal cancer surgery has proven
feasible.
The UK MRC CLASICC trial demonstrated equivalent long-term recurrence and survival results between laparoscopic and open rectal surgery, and the
results of several ongoing trials (the COREAN,
COLOR II, JCOG 0404, and ACOSOG-Z6051 trials)
are awaited.
Greater degree of rotation and instrument movement
along with three-dimensional optics has led to the use
of robotic technology in rectal cancer surgery. The UK
MRC ROLARR trial is currently comparing laparoscopic to robotic resection.
The optimal learning curve and method to train
surgeons in minimally invasive surgery is still in flux
and include surgeon mentoring and didactic sessions.
The American College of Surgeons Oncology Group

(ACOSOG)-Z6051 trial is a prospective, randomized trial
aimed at evaluating the noninferiority of laparoscopic
compared with open surgery for rectal cancer.15 Primary
measurements include completeness of TME and circumferential and distal resection margins. The secondary endpoints are short-term benefits; disease-free and overall
survival; pelvic recurrence rates; and quality of life, including sexual and bowel function. The results are awaited.
Meta-analysis
The data on laparoscopic resection for mid- and low-rectal

cancers are limited in predominantly retrospective series,
prospective nonrandomized, controlled trials, and small randomized, controlled studies. As predicted, the conclusion
from all meta-analyses is that patients in the laparoscopic
group have better short-term outcomes than patients in the
open surgery group. This is similar to the results of laparoscopic procedures in other diseases.16-19 In terms of recurrence and survival, one meta-analysis (including a 5-year
follow-up of the MRC CLASICC trial) evaluated 2,095 patients with rectal cancer (1,096 undergoing laparoscopic; 999
undergoing open surgery) from 12 randomized, controlled
studies.20 Less blood loss, a more rapid return to oral diet,
and briefer hospital stay were identified as short-term
benefits of laparoscopic versus open surgery. Long-term
215
TRAKARNSANGA AND WEISER
outcomes; local, wound-site, and distant recurrences; 5-year

overall and disease-free survival; and urinary and sexual
function did not differ significantly between the two groups.
Summary of Current Evidence
The United Kingdom MRC CLASICC trial finding of

increased CRM positivity within the laparoscopic anterior
resection group raised questions about the oncologic competence of this procedure. The question of adequacy is answered by the 5-year follow-up results. The COREAN trial
concludes that laparoscopic surgery for rectal cancer, performed by skillful laparoscopic surgeons in high-volume
centers, is feasible and has promising short-term results.
As we await the long-term results of the aforementioned
ongoing trials (the COREAN, COLOR II, JCOG 0404, and
ACOSOG-Z6051 trials), one topic must be considered before
applying the results to general practice: how to prepare and
qualify operators so that patients receive the maximum
benefits of MIS.
Learning Curve for Laparoscopic Surgery for Rectal
Cancer: Is There a Magic Number?
As previously noted, rectal resection is more demanding

than colectomy because of the anatomy of the pelvis, the
desirability of preserving autonomic nerve function, and the
sequelae of neoadjuvant treatment. In conventional open
surgery, the retractor is a key factor in achieving adequate
exposure, especially in male patients with a bulky mesorectal tumor and narrow pelvis. Appropriate TME should be
performed by sharp (not blunt) dissection to achieve the best
oncologic outcome. All of these considerations make laparoscopic rectal resection more challenging.
One study from Japan analyzed the learning curve for
laparoscopic low anterior resection.21 Single surgeons performed 250 operations, with patients divided into five
groups. The learning curve analysis demonstrated that
operative time stabilized after 50 cases. The conversion rate
was significantly lower after 150 cases (p 0.05), correlating with male sex and advanced T stage. Other studies have
suggested that an adequate learning curve is passed at
somewhere between 20 and 60 cases.6,7,22,23 In laparoscopic
TME, adequacy of exposure without use of the retractor is
necessary. Therefore, not only the experience of the surgeon
but the aptitude of the entire surgical team is important.
The mirror image from the camera and the alignment of the
first assistants working instruments in opposite positions
may disorient the assistant operator. One study concluded
that assisting in more than 30 to 40 cases is sufficient to
overcome mirror-image movements.24
Existing published reports show no final agreement regarding the optimal number of cases necessary to achieve
technical expertise. Extrapolated from the COST study,
ASCRS and the Society of Gastrointestinal and Endoscopic
Surgeons (SAGES) recommend that a practitioner complete
at least 20 laparoscopic resections of benign colon lesions
before being credentialed for colon cancer resection.9 Undoubtedly, completing a large number of operations correlates with better outcomes. One good example is the
correlation between high-volume surgeons (all ranked
among the top one-third in operative experience in their
respective cancer centers) and impressive short-term results
216
12
shown in the COREAN trial. This begs the question: How

can operators pass the learning curve without putting patients at risk? Simulation and animal and cadaveric training
courses are currently available worldwide; however, it is not
known if this kind of training sufficiently qualifies a surgeon
to perform resection in human patients. From our perspective, the answer is still uncertain. Intraoperative supervision by an experienced surgeon as well as appropriate case
selection should be major considerations in any training
program, even at an advanced level. This is essential to safe
clinical practice.25
Is There Any Benefit to Hand-Assisted Laparoscopy?
It has been suggested that hand-assisted laparoscopic

colectomy has potential benefits, including a shorter learning curve for the practitioner, less operative time, and lower
conversion rate compared to conventional laparoscopic surgery. This may be especially true in complex procedures
such as total proctocolectomy.26-28 Unfortunately, several
difficult issues are associated with hand-assisted laparoscopy. Placing a hand into the abdominal cavity may have
iatrogenic effects including a traumatic affect on the immune system, which may eliminate the long-term benefits of
MIS (such as fewer postoperative adhesions). Interleukin 6
(IL-6) and C-reactive protein (CRP) are sensitive markers of
the immune systems acute inflammatory response, and
exaggeration of these may actually have a deleterious effect
on wound healing and increase the risk of postoperative
infection.29,30 Several human studies have shown significant
lowering of IL-6 and CRP in laparoscopic compared with
open surgery (p 0.05, p 0.007, and p 0.001).29 In two
animal studies comparing hand-assisted laparoscopic surgery (HALS) with laparoscopic surgery, slightly higher levels of IL-6 and CRP were reported with HALS; however, the
levels of these cytokines were significantly lower than the
levels seen in open surgery (p 0.04 and p 0.05).30,31 No
comparative human study has been reported.
Regarding long-term complications, one retrospective
study compared 266 hand-assisted operations with 270
laparoscopic operations for colorectal disease. After a 27month median follow-up, the incidence of small bowel obstruction was not significantly different between the handassisted versus laparoscopic groups (4.1% vs. 7.4%; p
0.10), nor was the rate of postoperative incisional hernia
(6% vs. 4.8%; p 0.54).32 Based on these results, we may
conclude that hand-assisted surgery does not appear to be
inferior to laparoscopic surgery, and retains the potential
benefits of MIS.
HALS for rectal cancer is challenging because, as noted
above, the pelvic space is restricted by its bony anatomic
structure. On the positive side, the hand may be used as an
effective retractor to achieve adequate exposure. Unfortunately, in a deep and narrow pelvis, the hand may not be a
useful tool because of the limited anatomic space. A multicenter prospective, randomized study comparing handassisted with straight laparoscopic-assisted proctectomy for
rectal cancer is currently in the recruitment phase.33 The
planned sample size is 128 cases; estimated accrual will end
by December 2012. The primary outcome is operative time.
The secondary outcomes include adequacy of resection margins; in-hospital mortality and morbidity; and preoperative
as well as 3- and 6-month postoperative follow-ups of uri-
MINIMALLY INVASIVE SURGERY OF RECTAL CANCER
nary and sexual function. The investigators hypothesize

that HALS results in shorter operative time while retaining
the benefits associated with laparoscopic surgery.
As we await the results of this trial, several publications
have reported on the role of robotic-assisted laparoscopic
TME. A prospective randomized, controlled trial comparing
robotic-assisted and laparoscopic TME is also ongoing. Comparative analysis, based on the ultimate data comparing
hand-assisted and robotic-assisted laparoscopic surgery for
rectal cancer, will be challenging. Only one three-arm retrospective study comparing HALS, robotic-assisted, and laparoscopic TME has been reported (with 30 patients with
rectal cancer per arm).34 No significant differences in pathologic outcome or complications were identified. To ascertain
the equivalence of HALS and robotic surgery, however, a
prospective randomized, controlled trial is needed.
Robotic-Assisted Laparoscopic Surgery for Rectal
Cancer: the Ideal Solution?
Robotic-assisted laparoscopic prostatectomy has become

extremely popular. An excellent three-dimensional visual
system using EndoWrist instruments eliminates many of
the limitations of operating with straight laparoscopic instruments in a restricted anatomic area. The extraperitoneal part of the rectum is situated in the pelvic cavity.
Thus, the idea of robotic-assisted laparoscopic surgery for
rectal cancer appeals to surgeons. Loss of free movement to
several quadrants of the abdomen, as well as increases in
cost without benefits when compared to laparoscopic surgery, limits the use of robotic-assisted laparoscopic colectomy for colon cancer. A retrospective study comparing 40
robotic-assisted to 135 laparoscopic right hemicolectomies
reported no significant difference in short-term benefits,
including estimated blood loss, rate of conversion, complications, and hospital stay. Moreover, operative time and costs
were significantly higher in the robotic versus the laparoscopic group (p 0.001 vs. p 0.003, respectively).35 In
concurrence with the findings of a meta-analysis of seven
nonrandomized studies, the average operative time was 39
minutes longer and $792 more expensive than in conventional laparoscopy, with no improvement in short-term benefits.36
As noted, however, the rectum seems suited to robotic
surgery and the learning curve is less steep. Data on 50
robotic-assisted laparoscopic rectal surgery cases suggest
that the learning curve is passed after 15 to 25 operations.37
Unfortunately, the utilization of this technology has been
limited by cost. Most of the current evidence regarding the
benefits of robotic-assisted laparoscopic rectal cancer surgery consists of case series. Case-match and nonrandomized
studies conclude that robotic-assisted surgery is feasible and
comparable to laparoscopic TME.38-40 A large prospective,
randomized, controlled trial is needed, however, to assess
the equivalence of robotic surgery to conventional laparoscopic surgery.
The United Kingdom MRC ROLARR trial
The United Kingdom Medical Research Council Trial of

Robotic compared with Laparoscopic Resection for Rectal
cancer (ROLARR) trial is a multicenter prospective, randomized, controlled trial of robotic-assisted compared with
laparoscopic surgery in the curative treatment of rectal
cancer.41 Four hundred patients will be recruited and randomly assigned with a 1:1 ratio. This trial is currently in the
randomization phase, which will end by mid-2012. The
primary endpoint is conversion rates to open surgery. Secondary endpoints are intraoperative and postoperative complications; oncologic outcomes, including circumferential
margin, 3-year overall and disease-free survival; and quality
of life.
Techniques: Total Robotic or Hybrid Approach?
There is debate over port placements, docking techniques,

and techniques for take-down of the splenic flexure in
robotic-assisted laparoscopic TME. The two major techniques are 1) totally robotic surgery and 2) a hybrid approach, depending on which instrument is used in takedown. Some surgeons are practicing a hybrid approach
consisting of a robotic surgical system for vessel control and
pelvic dissection, and straight laparoscopy for splenic flexure mobilization. The idea is to decrease the prolonged
operative time caused by multiple dockings of the robot.
The technique of totally robotic surgery is challenging
because of limitations in port placement and positioning of
the robot. Most of the time, more than one docking of the
robot is needed to complete the operation. Some centers have
proposed the feasibility of a single-stage technique.42,43
However, there are currently no published studies comparing docking techniques. Surgeon preference and experience
are still the main factors in selecting an operation. From our
perspective, there is no single ideal technique that will fit all
patients. Knowledge of the benefits and limitations of each
technique is crucial, and the surgeon must be prepared to
tailor surgery to each individual patient.
Conclusion
Minimally invasive surgery for rectal cancer is challenging because of the anatomic restrictions of the bony pelvis
and the necessity of autonomic nerve preservation. Various
techniques have been proposed, including straight laparoscopic, hand-assisted, and robotic-assisted laparoscopic surgery. In laparoscopic rectal cancer surgery, the short-term
benefits are similar to those associated with other minimally
invasive techniques. Current data indicate that long-term
oncologic outcomes are similar in terms of recurrence and
survival. Several additional prospective, randomized, controlled trials are in progress. We believe that the results will
demonstrate the noninferiority of laparoscopic surgery compared to open surgery.
The data from small case series and one nonrandomized,
controlled trial indicate that the emerging techniques of
hand-assisted and robotic-assisted laparoscopy are feasible
and comparable to the results achieved with conventional
laparoscopic surgery. Prospective randomized, controlled
trials of both techniques are ongoing. At this time, it is not
possible to determine which procedure is best. Surgeon
preference and availability of instruments are crucial in
choosing the right procedure for each individual patient.
217
TRAKARNSANGA AND WEISER
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Atthaphorn Trakarnsanga*
Martin R. Weiser*
REFERENCES
1. Jacobs M, Verdeja JC, Goldstein HS. Minimally invasive colon resection
(laparoscopic colectomy). Surg Laparosc Endosc. 1991;1:144-150.
2. Paolucci V, Schaeff B, Schneider M, et al. Tumor seeding following
laparoscopy: international survey. World J Surg. 1999;23:989-997.
3. Wexner SD, Cohen SM. Port site metastases after laparoscopic colorectal
surgery for malignancy. Br J Surg. 1995;82:295-298.
4. Kuhry E, Schwenk WF, Gaupset R, et al. Long-term results of laparoscopic colorectal cancer resection. Cochrane Database of Syst Rev. 2008;16:
CD003432.
5. Lacy AM, Garca-Valdecasas JC, Delgado S, et al. Laparoscopy assisted
colectomy versus open colectomy for treatment of non-metastatic colon
cancer: a randomised trial. Lancet. 2002;359:2224-2229.
6. Clinical Outcomes of Surgical Therapy Study Group. A comparison of
laparoscopically assisted and open colectomy for colon cancer. N Engl J Med.
2004;350:2050-2059.
7. Jayne DG, Guillou PJ, Thorpe H, et al. Randomized trial of laparoscopicassisted resection of colorectal carcinoma: 3-year results of the UK MRC
CLASSIC Trial Group. J Clin Oncol. 2007;25:3061-3068.
8. Bonjer HJ, Hop WC, Nelson H, et al. Laparoscopically assisted vs. open
colectomy for colon cancer: a meta-analysis. Arch Surg. 2007;142:298-303.
9. Tjandra JJ, Kilkenny JW, Buie WD, et al. Practice parameters for the
management of rectal cancer (revised). Dis Colon Rectum. 2005;48:411-423.
10. Row D, Weiser MR. An update on laparoscopic resection for rectal
cancer. Cancer Control. 2010;17:16-24.
11. Jayne DG, Thorpe HC, Copeland J, et al. Five-year follow-up of the
Medical Research Council CLASICC trial of laparoscopically assisted versus
open surgery for colorectal cancer. Br J Surg. 2010;97:1638-1645.
12. Kang SB, Park JW, Jeong SY, et al. Open versus laparoscopic surgery
for mid or low rectal cancer after neoadjuvant chemoradiotherapy (COREAN
trial): short-term outcomes of an open-label randomised controlled trial.
Lancet Oncol. 2010;11:637-645.
13. COLOR II: Laparoscopic Versus Open Rectal Cancer Removal. http://
clinicaltrials.gov/ct2/show/NCT00297791. Accessed September 2, 2011.
14. Kitano S, Inomata M, Sato A, et al. Randomized controlled trial to
evaluate laparoscopic surgery for colorectal cancer: Japan Clinical Oncology
Group Study JCOG 0404. Jpn J Clin Oncol. 2005;35:475-477.
15. American College of Surgeons. Laparoscopic-Assisted Resection or
Open Resection in Treating Patients With Stage IIA, Stage IIIA, or Stage IIIB
Rectal Cancer. http://clinicaltrials.gov/ct2/show/NCT00726622. Accessed September 2, 2011.
16. Aziz O, Constantinides V, Tekkis PP, et al. Laparoscopic versus open
surgery for rectal cancer: a meta-analysis. Ann Surg Oncol. 2006;13:413-424.
17. Breukink S, Pierie J, Wiggers T. Laparoscopic versus open total
mesorectal excision for rectal cancer. Cochrane Database Systemat Rev.
2006;CD005200.
18. Anderson C, Uman G, Pigazzi A. Oncologic outcomes of laparoscopic
surgery for rectal cancer: a systemic review and meta-analysis of the literature. Eur J Surg Oncol. 2008;34:1135-1142.
19. Huang MJ, Liang JL, Wang LH, et al. Laparoscopic-assisted versus
open surgery for rectal cancer: a meta-analysis of randomized controlled trials
on oncologic outcomes. Int J Colorectal Dis. 2011;26:415-421.
20. Ohtani H, Tamamori Y, Azuma T, et al. A meta-analysis of the shortand long-term results of randomized controlled trials that compared
laparoscopy-assisted and conventional open surgery for rectal cancer. J
Gastrointest Surg. 2011;15:1375-1385.
21. Kayano H, Okuda J, Tanaka K, et al. Evaluation of the learning curve
in laparoscopic low anterior resection for rectal cancer. Surg Endosc. 2011;
25:2972-2979.
22. Park IJ, Choi GS, Lim KH, et al. Multidimensional analysis of the
218
learning curve for laparoscopic colorectal surgery: Lesson from 1,000 cases of
laparoscopic colorectal surgery. Surg Endosc. 2009;23:839-846.
23. Li JC, Hon SS, Ng SS, et al. The learning curve for laparoscopic
colectomy: Experience of surgical fellow in an university colorectal unit. Surg
Endosc. 2009;23:1603-1608.
24. Hwang MR, Seo GJ, Yoo SB, et al. Learning curve of assistants in
laparoscopic colorectal surgery: overcoming mirror imaging. Surg Endosc.
2010;24:2575-2580.
25. Miskovic D, Wyles SM, Ni M, et al. Systematic review on mentoring and
simulation in laparoscopic colorectal surgery. Ann Surg. 2010;252:943-951.
26. Meshikhes AW. Controversy of hand-assisted laparoscopic colorectal
surgery. World J Gastroenterol. 2010;16:5662-5668.
27. Aalbers AG, Biere SS, van Berge Henegouwen MI, et al. Hand-assisted
or laparoscopic-assisted approach in colorectal surgery: a systematic review
and meta-analysis. Surg Endosc. 2008;22:1769-1780.
28. Cima RR, Pendlimari R, Hobular SD, et al. Utility and short-term
outcomes of hand-assisted laparoscopic colorectal surgery: a single-institution
experience in 1103 patients. Dis Colon Rectum. 2011;54:1076-1081.
29. Novitsky YW, Litwin DE, Callery MP. The net immunologic advantage
of laparoscopic surgery. Surg Endosc. 2004;18:1411-1419.
30. Orenstein SB, Kaban GK, Litwin DE, et al. Evaluation of serum
cytokine release in response to hand-assisted, laparoscopic, and open surgery
in a porcine model. Am J Surg. 2011;202:97-102.
31. Matsumoto ED, Marulis V, Tunc L, et al. Cytokine response to surgical
stress: comparison of pure laparoscopic, hand-assisted laparoscopic, and open
nephrectomy. J Endourol. 2005;19:1140-1145.
32. Sonoda T, Pandey S, Trencheva K, et al. Longterm complications of
hand-assisted versus laparoscopic colectomy. J Am Coll Surg. 2009;208:62-66.
33. Hand-assisted versus Pure Laparoscopic Assisted Proctectomy for
Rectal Cancer. http://clinicaltrials.gov/ct2/show/study/NCT00651677?term
hand&show_descY. Accessed September 2, 2011.
34. Patel CB, Ragupathi M, Ramos-Valadez DI, et al. A three-arm (laparoscopic, hand-assisted and robot) matched-case analysis of intraoperative
and postoperative outcomes in minimally invasive colorectal surgery. Dis
Colon Rectum. 2011;54:144-150.
35. deSouza AL, Prasad LM, Park JJ, et al. Robotic assistance in right
hemicolectomy: is there a role? Dis Colon Rectum. 2010;53:1000-1006.
36. Maeso S, Reza M, Mayol JA, et al. Efficacy of the Da Vinci surgical
system in abdominal surgery compared with that of laparoscopy: A systematic
review and meta-analysis. Ann Surg. 2010;252:254-262.
37. Bokhari MB, Patel CB, Ramoz-Valadez DI, et al. Learning curve for
robotic-assisted laparoscopic colorectal surgery. Surg Endosc. 2011;25:855860.
38. Pigazzi A, Luca F, Patriti A, et al. Multicentric study on robot
tumor-specific mesorectal excision for the treatment of rectal cancer. Ann
Surg Oncol. 2010;17:1614-1620.
39. Baik SH, Kwon HY, Kim JS, et al. Robotic versus laparoscopic low
anterior resection of rectal cancer: Short-term outcome of a prospective
comparative study. Ann Surg Oncol. 2009;16:1480-1487.
40. Antoniou SA, Antoniou GA, Koch OO, et al. Robot-assisted laparoscopic
surgery of the colon and rectum. Surg Endosc. 2012;26:1-11.
41. Robotic versus Laparoscopic Resection for Rectal Cancer. http://ctru.
leeds.ac.uk/rolarr. Accessed September 2, 2011.
42. Choi DJ, Kim SH, Lee RJ, et al. Single-stage totally robotic dissection
for rectal cancer surgery: Technique and short-term outcome in 50 consecutive patients. Dis Colon Rectum. 2009;52:1824-1830.
43. Park YA, Kim JM, Kim SA, et al. Totally robotic surgery for rectal
cancer: From splenic flexure to pelvic floor in one setup. Surg Endosc.
2010;24:715-720.
Radiation in Rectal Cancer: What Are the

Options and If/When Can It Be Avoided?
By Karyn A. Goodman, MD
Overview: Treatment of rectal cancer has improved greatly

over recent decades. This review looks at the pivotal trials in
the development of the current standard of therapy as well
as new directions for more individualized therapy for rectal
cancer. Rates of local recurrence and overall survival (OS) for
surgery alone have improved with the use of (neo)adjuvant
5-fluorouracil (5-FU)-based chemoradiation. New surgical
techniques have improved outcomes, but preoperative radiotherapy still confers an additional benefit. Despite benefits in
the metastatic setting, the addition of oxaliplatin to 5-FU and
ECTAL CANCER treatment has made great strides

over the last 50 years from the time of surgery alone,
employing a blunt dissection technique that was associated
with local recurrence rates of 30% to 60% for locally advanced disease. Adjuvant 5-FU-based chemoradiation reduced local failure rates to 10% to 12% and improved OS
by 10% to 15%.1-3 These results led to the National Cancer
Institute consensus statement in 1990 recommending adjuvant chemoradiation for all patients with T3 to T4 or
node-positive rectal cancer.4 This was the first attempt to
risk stratify patients with rectal cancer.
Subsequently, the introduction of the total mesorectal
excision (TME) and vastly improved surgical techniques
have reduced the risk of positive margins and node disease
left behind. As a result, local recurrence rates have dropped
to 10% to 15% with surgery alone. Despite better surgical
technique, the Dutch TME study (CKVO 95 04) still showed
a benefit in local control for preoperative radiotherapy
compared with surgery alone.5 These investigators randomly selected 1,805 eligible patients to receive either
surgery alone or short-course radiation therapy (5 5 Gy)
followed by surgery and concluded that the addition of
radiation significantly decreases the rate of local recurrence
at 2 years even with high-quality surgery (p 0.001). In
addition, the MRC07 trial randomly selected 1,350 patients
with rectal cancer to receive either neoadjuvant short-course
radiation therapy (5 5 Gy) or selective postoperative
concurrent chemoradiotherapy for patients with a positive
circumferential resection margin. Preoperative radiotherapy was associated with improved 3-year local control compared with selecting patients to receive adjuvant long-course
radiation therapy.6 Thus, the use of radiotherapy even in
the setting of a high-quality TME appears to improve local
control rates.
To determine the optimal sequence for surgery and chemoradiotherapy, several randomized trials compared preoperative with postoperative 5-FU-based chemoradiotherapy.
The German Rectal Study demonstrated an improvement in
local relapse rates and reduced acute and overall toxicity
using preoperative chemoradiation compared with postoperative chemoradiation.7 This phase III randomized trial has
established the standard of care for T3/4 or node-positive
rectal cancer as neoadjuvant infusional 5-FU with conventionally fractionated radiotherapy to 5,040 cGy. With this
approach, investigators reported local recurrence rates of
6% and distant recurrence rates of 36%; therefore, bringing
radiotherapy has not shown improved outcomes, although it

increased toxicity. Preoperative therapy also allows for the
identification of predictive and prognostic markers for response to treatment, which has great potential to individualize
treatment. Currently, the search for validated biomarkers and
the refinement of risk stratification are the focus of ongoing
study. Tailored therapy may include modification of the pelvic
radiotherapy field, nonoperative therapy, or avoidance of
radiotherapy.
rates of local recurrence well below rates of distant recurrence. The subsequent publications of the EORTC 22921
and FFCD 9203 further confirmed the benefit of combining
5-FU-based chemotherapy with preoperative radiation by
improving the rates of pathologic complete response (pCR)
and local control with acceptable increase in toxicity.8,9
Intensifying Neoadjuvant Therapy: Is More Better?
The next step in the evolution of therapy for rectal cancer

has been the attempt to improve our current standard of
care with more intensive therapy. Given the multitude of
new chemotherapy combinations and molecularly targeted
systemic therapies being investigated in metastatic colorectal cancer, it was natural to consider incorporating these
regimens into the adjuvant/neoadjuvant setting.10
The endpoints for evaluating the benefit of combination
chemotherapy agents for neoadjuvant therapy have been
tailored to allow for earlier assessment of efficacy such as
pCR. For example, pCR following preoperative chemoradiation is predictive of OS and should therefore be an appropriate surrogate for evaluating the effect of neoadjuvant
chemoradiation regimens for rectal cancer.11-16 Other clinically relevant endpoints may include sphincter preservation
and toxicity.
The wide array of new cytotoxic and targeted agents has
propagated numerous small phase II studies evaluating
pCR rates. Based on promising results of these phase II
studies, there have now been four recently reported phase
III randomized trials that have added oxaliplatin to 5-FUbased preoperative chemoradiation with the intent of improving pCR and disease-free survival. Both the ACCORD
trial and the STAR trial found no difference in pCR rate with
the addition of oxaliplatin to standard 5-FU infusional
chemotherapy and 50.4 Gy.17,18 The toxicity was significantly greater (ACCORD: 25% v. 11%, p .001; STAR:
24% v. 8%, p 0.001) in the oxaliplatin arm as well. The
preliminary reports of the NSABP R-04 and the German
From the Department of Radiation Oncology, Memorial Sloan-Kettering Cancer Center,

New York, NY.
Address reprint requests to Karyn A. Goodman, MD, Department of Radiation Oncology,
Memorial Sloan-Kettering Cancer Center, 1275 York Ave., New York, NY, 10065; email:
goodmank@mskcc.org.
1092-9118/10/1-10
219
KARYN A. GOODMAN
CAO/ARO 04 studywhich added oxaliplatin to continuous

intravenous infusion 5-FU or capecitabine and 50.4 Gy of
radiotherapy demonstrated that oxaliplatin did not result
in an overwhelming improvement in pCR rates, although
it was statistically significant (13% v. 17%, p 0.033) in the
German study.19,20 The NSABP study showed significantly
more grade 3 and 4 diarrhea with oxaliplatin (15% v. 7%, p
0.0001), similar to the STAR and ACCORD trials. Interestingly, the toxicity for the oxaliplatin arm was not higher in
the German study, likely because of a lower cumulative dose
of oxaliplatin and a split course treatment with the 5-FU.
Oxaliplatin did not improve sphincter preservation one of
the goals of preoperative therapyin any of the studies.
Clearly, there is no role for oxaliplatin in combination with
neoadjuvant 5-FU-based chemoradiation. It appears to only
increase toxicity without improving pCR rates, sphincter
preservation, or local control.
Individualization of Neoadjuvant Therapy
The neoadjuvant approach is particularly appealing for

assessing the biology of the tumor and identifying active
regimens, which can then be used for the patient in the
postoperative setting. Preoperative therapy also allows us
to identify predictive and prognostic markers for treatment
response. This leads to the potential for individualizing
treatment regimens using information about the tumor and
the patient. Basic prognostic information, such as tumor
stage, histologic grade, and location, are being reinforced by
the use of biomarkers, which may help us identify patients
most likely to benefit from certain drugs. Although there are
many markers that have been evaluated, none have been
validated for clinical use. This is an area of greatly needed
research. Future studies that add new agents may have
the best chance of showing effect if we can use predictive
KEY POINTS
220
Preoperative chemoradiation improved local control

rates even after total mesorectal excisionfor patients with locally advanced rectal cancer.
Oxaliplatin combined with 5-fluorouracil based
chemoradiation does not improve pathologic response
rates, sphincter preservation, or disease-free survival
and increases the toxicity of neoadjuvant chemoradiation.
Pooled analyses of patients who received adjuvant
treatment for rectal cancer have demonstrated intermediate and high risk disease, potentially identifying subgroups of patients who may not benefit from
adjuvant radiotherapy.
Nonoperative approaches for patients with clinical
complete responses to chemoradiation may allow for
a more individualized approach for patients with
distal rectal cancers.
The PROSPECT trial is investigating the selective
use of chemoradiation in patients treated with neoadjuvant folinic acid, 5-FU, and oxaliplatin
(FOLFOX) chemotherapy for locally advanced rectal
cancers.
biomarkers as eligibility criteria to enrich the population of

patients who will potentially respond.
Until validated biomarkers are available, further refinements in risk stratification based on clinical factors should
be pursued to allow for better identification of patients
who may or may not benefit from more aggressive multimodality therapy. In an attempt to stratify patients with locally
advanced rectal cancer, investigators analyzed pooled data
from 2,551 patients enrolled in three North American rectal
trials. Patients with intermediate-risk rectal cancer defined as T1 to T2N1 or T3N0 had only approximately 6%
to 8% risk of local recurrence and may not derive any benefit
from adjuvant radiation.21 In addition, validated nomograms based on clinical and pathologic features are now
available to assess risks for local or distant recurrence and
neoadjuvant chemoradiation.22 In Northern Europe, features on rectal magnetic resonance imaging (MRI) have been
used to stratify patients into low-, intermediate-, and highrisk groups. Based on MRI findings of extramural extent
of the tumor, the relation to the mesorectal fascia, and the
presence of multiple suspicious lymph nodes, low-risk patients are treated with surgery alone, patients with
intermediate-risk rectal tumors with short-course 5 Gy
5 preoperative protocol, and patients with high-risk tumors
based on MRI findings with long-course preoperative
chemoradiation.23
Patient-specific risk factors such as age, performance
status, and other medical comorbidities must be considered
when adding chemotherapy or targeted agents to neoadjuvant therapy. The benefits of bevacizumab are likely to be
outweighed by cardiovascular or thrombotic risks in the
patient with a history of cardiac disease. For young, premenopausal women who are trying to preserve fertility, the
risks and benefits of pelvic radiation must be addressed. For
these young, female patients, the argument to avoid radiotherapy could be made.
Selective Use of Therapies for Rectal Cancer
Tailoring therapy can also be achieved by modifying the

pelvic radiotherapy field. With better imaging techniques,
it may not be necessary to treat the whole pelvis for all
patients. There may be patients whose tumors have a lower
risk of lateral pelvic sidewall involvement and, therefore,
can be spared treating internal iliac nodes. Also, after the
use of induction chemotherapyan area of growing interestit may be possible to direct radiation at the postchemotherapy volume, potentially making the treatment
fields smaller. The use of intensity-modulated radiotherapy
(IMRT) can minimize dose to radiosensitive structures in
the pelvis and potentially minimize toxicity but requires
more in-depth knowledge of pelvic anatomy and nodal drainage patterns. IMRT may also allow dose escalation to higher
risk areas and to achieve better pCR rates in patients not
receiving surgery.
The option of nonoperative therapy is also emerging as
an option for patients who have clinical complete responses
to neoadjuvant therapy. This approach was initially reported by investigators from Brazil. They reported on 361
patients with rectal cancer, 99 (27%) of whom demonstrate
a clinical complete response. These patients were observed
without radical resection. The local recurrence rate was
remarkably low (5%), and all five of these patients have been
salvaged with either surgery or additional radiation and

have no evidence as yet of subsequent recurrence.24 However, the time to recurrence in this group is quite long. Most
data from patients who receive surgery demonstrate the
majority of local recurrences within 2 to 3 years of treatment. Yet in this group, the mean interval is 52 months.
This highlights the need for continued close surveillance of
these patients, many years out from treatment. There is
interest in studying nonoperative management in Europe
and the United States, particularly in the low-lying tumors
that would require an abdominoperineal resection and permanent colostomy.
Finally, tailored therapy may come in the way of selective
therapy for patients who have a good response to induction

chemotherapy. In the near future, we will see the opening of
the PROSPECT trial: a large phase III randomized trial of
no radiotherapy for patients with good response to induction
chemotherapy. This study will accrue 1,000 patients to
determine if preoperative folinic acid, 5-FU, and oxaliplatin
(FOLFOX) is equivalent to standard preoperative chemoradiation. This study is based on excellent outcomes in a pilot
study of 30 patients who received preoperative FOLFOX and
bevacizumab and had a 27% pCR rate.25 The PROSPECT
trial is large enough to allow for a better understanding
of prognostic and predictive factors to help identify which
patients benefit from different therapeutic approaches.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Karyn A. Goodman*
REFERENCES
1. Douglass HO Jr, Moertel CG, Mayer RJ, et al. Survival after postoperative combination treatment of rectal cancer. N Engl J Med. 1986;315:12941295.
2. Krook JE, Moertel CG, Gunderson LL, et al. Effective surgical adjuvant
therapy for high-risk rectal carcinoma. N Engl J Med. 1991;324:709-715.
3. OConnell MJ, Martenson JA, Weiand HS, et al. Improving adjuvant
therapy for rectal cancer by combining protracted-infusion fluorouracil with
radiation therapy after curative surgery. N Engl J Med. 1994;331:502-507.
4. NIH consensus conference. Adjuvant therapy for patients with colon and
rectal cancer. JAMA. 1990;264:1444-1450.
5. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer.
N Engl J Med. 2001;345:638-646.
6. Sebag-Montefiore D, Stephens RJ, Steele R, et al. Preoperative radiotherapy versus selective postoperative chemoradiotherapy in patients with
rectal cancer (MRC CR07 and NCIC-CTG C016): a multicentre, randomised
trial. Lancet. 2009;373:811-820.
7. Stahl M, Stuschke M, Lehmann N, et al. Chemoradiation with and
without surgery in patients with locally advanced squamous cell carcinoma of
the esophagus. J Clin Oncol. 2005;23:2310-2317.
8. Bosset JF, Calais G, Mineur L, et al. Enhanced tumorocidal effect of
chemotherapy with preoperative radiotherapy for rectal cancer: preliminary
results-EORTC 22921. J Clin Oncol. 2005;23:5620-5627.
9. Gerard JP, Conroy T, Bonnetain F, et al. Preoperative radiotherapy with
or without concurrent fluorouracil and leucovorin in T3-4 rectal cancers:
results of FFCD 9203. J Clin Oncol. 2006;24:4620-4625.
10. Andre T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil,
and leucovorin as adjuvant treatment for colon cancer. N Engl J Med.
2004;350:2343-2351.
11. Rodel C, Martus P, Papadoupolos T, et al. Prognostic significance of
tumor regression after preoperative chemoradiotherapy for rectal cancer.
J Clin Oncol. 2005;23:8688-8696.
12. Janjan NA, Crane C, Feig BW, et al. Improved overall survival among
responders to preoperative chemoradiation for locally advanced rectal cancer.
Am J Clin Oncol. 2001;24:107-112.
13. Garcia-Aguilar J, Hernandez de Anda E, Sirivongs P, et al. A pathologic
complete response to preoperative chemoradiation is associated with lower
local recurrence and improved survival in rectal cancer patients treated by
mesorectal excision. Dis Colon Rectum. 2003;46:298-304.
14. Ruo L, Tickoo S, Klimstra DS, et al. Long-term prognostic significance
of extent of rectal cancer response to preoperative radiation and chemotherapy. Ann Surg. 2002;236:75-81.
15. Vecchio FM, Valentini V, Minsky B D, et al. The relationship of
pathologic tumor regression grade (TRG) and outcomes after preoperative
therapy in rectal cancer. Int J Radiat Oncol Biol Phys. 2005;62:752-760.
16. Guillem JG, Chessin DB, Cohen AM, et al. Long-term oncologic outcome following preoperative combined modality therapy and total mesorectal
excision of locally advanced rectal cancer. Ann Surg. 2005;241:829-836.
17. Gerard JP, Azria D, Gourgou-Bourgade S, et al. Comparison of two
neoadjuvant chemoradiotherapy regimens for locally advanced rectal cancer:
results of the phase III trial ACCORD 12/0405-Prodige 2. J Clin Oncol.
2010;28:1638-1644.
18. Aschele C, Cionini L, Lonardi S, et al. Primary tumor response to
preoperative chemoradiation with or without oxaliplatin in locally advanced
rectal cancer: pathologic results of the STAR-01 randomized phase III trial.
J Clin Oncol. 2011;29:2773-2780.
19. Roh MS, Yothers GA, OConnell MJ, et al. The impact of capecitabine
and oxaliplatin in the preoperative multimodality treatment in patients with
carcinoma of the rectum: NSABP R-04. J Clin Oncol. 2011;29 (suppl 15; abstr
3503).
20. Roedel C, Becker H, Fietkau R, et al. Preoperative chemoradiotherapy
and postoperative chemotherapy with 5-fluorouracil and oxaliplatin versus
5-fluorouracil alone in locally advanced rectal cancer: First results of the
German CAO/ARO/AIO-04 randomized phase III trial. J Clin Oncol. 2011;29
(suppl 15; abstr 3505).
21. Gunderson LL, Sargent DJ, Tepper JE, et al. Impact of T and N stage
and treatment on survival and relapse in adjuvant rectal cancer: a pooled
analysis. J Clin Oncol. 2004;22:1785-1796.
22. Valentini V, van Stiphout RG, Lammering G, et al. Nomograms for
predicting local recurrence, distant metastases, and overall survival for
patients with locally advanced rectal cancer on the basis of European
randomized clinical trials. J Clin Oncol. 2011;29:3163-3172.
23. Smith N, Brown G. Preoperative staging of rectal cancer. Acta Oncol.
2008;47:20-31.
24. Habr-Gama A, Perez RO, Proscurshim I, et al. Patterns of failure and
survival for nonoperative treatment of stage c0 distal rectal cancer following
neoadjuvant chemoradiation therapy. J Gastrointest Surg. 2006;10:13191328.
25. Schrag D, Weiser MR, Goodman KA, et al. Neoadjuvant FOLFOX-bev,
without radiation, for locally advanced rectal cancer. J Clin Oncol. 2010;28
(suppl 15; abstr 3511).
221
STAGE III COLON CANCER: WHAT WORKS, WHAT

DOESNT AND WHY, AND WHATS NEXT (eQ&A)
CHAIR
Thierry Andre, MD
Pitie-Salpetriere Hospital
Paris, France
SPEAKERS
Bert H. ONeil, MD
University of North Carolina at Chapel Hill
Chapel Hill, NC
Boston, MA
Stage III Colon Cancer: What Works, What

Doesnt and Why, and Whats Next
By Thierry Andre , MD, Bert H. ONeil, MD, and Jeffrey A. Meyerhardt, MD, MPH
Overview: Adjuvant treatment for patients with stage III

colon cancer, one of the most common malignancies, is an
important issue in oncology. The use of adjuvant chemotherapy in this setting has undoubtedly improved prognosis. This
article describes the development of adjuvant therapy and
progress in the past decade as well as failures in multiple

agents that have demonstrated efficacy in the metastatic
setting. Finally, the current clinical trials will be reviewed, as
well as complementary therapies including diet and exercise
for survivors of colorectal cancer.
respectively.4,5 These results constituted an impressive

therapeutic advance, and 5-FU/levamisole adjuvant chemotherapy became the standard of care for patients with stage
III colon cancer.
Two other studies demonstrated the benefit of adjuvant
chemotherapy with 5-FU and leucovorin (LV) compared
with no treatment for patients with colon cancer.9,10 The
International Multicenter Pooled Analysis of colorectal cancer Trials (IMPACT) showed a 22% relative risk reduction in
mortality with 5-FU/LV.9
OLORECTAL CANCER is the third most common

cancer in both men and women, and the fourth leading
cause of cancer deaths in the world.1 The estimated worldwide incidence of colorectal cancer is 1.2 million per year.1
In Western countries, the median age at diagnosis is 71, and
nearly 25% of all colon cancers are diagnosed at stage III.2
The main prognostic factor for estimating survival or relapse
after surgery for localized disease is the tumor, node, and
metastasis staging system.3 The 3-year disease-free survival
(DFS) for patients with stage III colon cancer without
any postoperative chemotherapy ranges between 44% and
52%.4,5
This review focuses on adjuvant therapy for patients with
stage III colon cancer. The goal is to illustrate how adjuvant
therapy has improved survival 3-fold in patients with stage
III colon cancer and to acknowledge those who contributed
to this achievement. We will also review efforts that have
not succeeded and discuss next steps.
Early Chemotherapy and 5-Fluorouracil/Levamisole
The first drugs evaluated in colon cancer trials were the

alkylating agent thiotepa and the antimetabolites fluorodeoxyuridine and 5-fluorouracil (5-FU), which was discovered
by Charles Heidelberger in 1957.6 The first adjuvant trials
using various routes of administration were performed in
the late 1960s and early 1970s. However, all proved unsuccessful except for trials using 5-FU. According to the criteria
of that time, positive results were reported with intraluminal or intravenous administration of 5-FU, but the small
number of undersized randomized studies failed to show any
statistical significance.7 In 1988, Buyse and colleagues carried out a meta-analysis of 17 trials involving 6,791 patients
with colorectal cancer (mean of 400 patients per trial), and
found a small benefit in overall survival (OS) with the
5-FU based regimens (hazard ratio [HR] 0.83).8 They also
concluded that much larger trials were needed.
Adjuvant Therapy for Patients with Colon Cancer: the
First Step (1990)
The first study demonstrating the value of adjuvant chemotherapy in patients with stage III colon cancer (TX, N1 or
N2, M0) was published by Moertel in 1990.4 This study
showed an increase in OS and DFS in patients receiving
5-FU/levamisole chemotherapy during 1 year compared with
patients treated with levamisole alone or not receiving
any treatment. After a mean follow-up of 6.5 years, patients treated with 5-FU/levamisole showed a 40% reduction
in their recurrence rate and an estimated 33% reduction in overall death rate.4 The 3-year DFS and 5-year
OS estimated from the survival curves were 64% and 63%,
Modulation of 5-FU with Levamisole or Leucovorin
After the publication of NSABP C-04 study results in

1999,11 which showed a small DFS advantage of 5-FU/LV
compared with FU/levamisole in patients with high-risk
stage II and III colon cancer, the addition of levamisole to
5-FU was withdrawn. The combination of 5-FU (bolus or
short infusion) and LV, administered either daily for 5 days
per month (according to the Mayo Clinic regimen) or weekly
for 6 months (according to the Roswell Park regimen),
became the new standard treatments for patients with stage
III colon cancer.11,12 In fact, the Intergroup study 0089
(INT-0089) showed that the Roswell Park regimen, the
Mayo Clinic regimen, 5-FU/LV/levamisole, and 5-FU/levamisole (control arm) treatments were equivalent. In this
four-arm study, the authors concluded that 5-FU/LV (either
the Roswell Park or the Mayo Clinic regimens) could replace
5-FU/levamisole.12 Interestingly, the biweekly LV5FU2 regimen (LV followed by both a bolus and a 22-hour infusion of
5-FU for 2 consecutive days) was also compared with
monthly 5-FU/LV (a modified Mayo Clinic regimen) in the
GERCOR C96-1 trial, which included patients with stage II
and III colon cancer.13 There was no significant improvement found in DFS, but the LV5FU2 regimen became
another accepted standard because of its improved safety
profile. Finally, the ACCENT (Adjuvant Colon Cancer
ENdpoinTs) meta-analysis, which included individual data
from more than 20,000 patients followed for more than 8
years, confirmed a 10% absolute improvement in OS for
patients with stage III colon cancer who received adjuvant
From the Service dOncologie Medicale, Hopital Saint-Antoine, Assistance Publique des
Hopitaux de Paris, Paris, France and Universite Pierre et Marie Curie (Paris 6); University
of North Carolina, Chapel Hill, NC; Dana-Farber Cancer Institute, Harvard Medical
School, Boston, MA.
Address reprint requests to Jeffrey A. Meyerhardt, MD, MPH, Dana-Farber Cancer
Institute, 44 Binney Street, Boston, MA 02215; email: Jeffrey_Meyerhardt@dfci.harvard.
edu.
1092-9118/10/1-10
223
ANDRE, ONEIL, AND MEYERHARDT

Table 1. Comparison of Fluoropyrimidines Alone (5-FUl/LV) and Fluoropyrimidines (5-FU/LV)/Oxaliplatin Combination in Patients with
Stage III Colon Cancer: Survival in 3 Phase III Randomized Controlled Trials
ENDPOINTS
Disease-Free Survival
Follow-up duration, y
DFS without vs. with oxaliplatine, %
Absolute difference in DFS, %
HR (95% CI)
p value
Overall Survival
Follow-up duration, y
OS without vs. with oxaliplatine, %
Absolute difference in OS, %
HR (95% CI)
p value
MOSAIC61 Stage III
MOSAIC18 Stage III
NSABP 0720 Stage III
NO169819 Stage III
3
65.3 vs. 72.2
6.3
0.76 (0.620.92)
.005
5
58.9 vs. 66.4
7.5
0.78 (0.650.93)
.005
5
57.8 vs. 64.4
6.6
0.78 (0.680.90)
.0007
3
66.5 vs. 70.9
4.4
0.80 (0.690.93)
.0045
3
ND
ND
6
68.7 vs. 72.9
4.2
0.80 (0.650.97)
.023
5
73.8 vs. 76.5
2.7
0.85 (0.721.00)
.052
4.75
ND
3.4
0.87 (0.721.05)
.1486
Abbreviations: 5-FU, 5-fluorouracil; CI, confidence interval; DFS, disease-free survival; HR, hazard ratio; LV, leucovorin; ND, not done; OS: overall survival.
chemotherapy with fluoropyrimidines compared with surgery alone.14

Oral Fluoropyrimidines, Capecitabine, and
Uracil/Tegafur (UFT) Are Equivalent to 5-FU/LV
Oral drugs are more convenient than intravenous chemotherapy, at least when used as single agents. The X-ACT
trial compared bolus 5-FU/LV (the Mayo Clinic regimen)
with oral capecitabine for 6 months in patients with stage III
colon cancer. DFS was at least equivalent (HR 0.87,
p 0.001 for noninferiority).15 In the NSABP C-06 study,
oral uracil and tegafur (UFT) plus LV was compared to the
Roswell Park regimen and the results showed similar DFS
and OS.16
A 3-Year DFS Is the New Endpoint in Adjuvant Trials
The traditional endpoint for clinical trials of patients

with adjuvant colon cancer treatment was 5-year OS. The
ACCENT meta-analysis of adjuvant studies demonstrated
that 3-year DFS was a predictor of 5-year OS results and
could be an appropriate primary endpoint for adjuvant
studies in colon cancer. The data led to the approval of
3-year DFS as the primary endpoint in adjuvant therapy
studies for patients with stage III colon cancer by the United
States Food and Drug Administration (FDA).17 Using 3-year
DFS as the main endpoint reduces trial duration, drug
development time, and cost, and as a result, new and better
treatments can be made available to patients more rapidly.
Oxaliplatin Plus Fluoropyrimidines Is Better than
Fluoropyrimidines Alone: The Second Step (2004)
KEY POINTS
224
Adjuvant chemotherapy for patients with stage III

colon cancer has substantially evolved in the past 2
decades, increasing disease-free and overall survival.
Until the early 1990s, research focused on modulation and duration of 5-fluorouracil and standard of
care became 5-fluorouracil and leucovorin in 1990.
The addition of oxaliplatin to 5-fluorouracil/leucovorin further improved disease-free and overall survival for patients with stage III disease; however,
additional toxicities during therapy and cumulative
neuropathy require active monitoring and adjustments in caring for patients with stage III disease.
Three drugs that have shown benefit in metastatic
colorectal cancer (irinotecan, bevacizumab, and cetuximab) failed to provide significant additional benefit in the management of stage III colon cancer, and
the reasons for the discordance between metastatic
disease and adjuvant therapy remains unclear.
Adjunctive therapies to standard treatment are actively being considered in colon cancer survivorship,
including physical activity, diet, obesity, vitamin D,
and nonsteroidal anti-inflammatory drugs, and data
are forthcoming to further understand the role these
may play in the outcomes of patients with stage III
colon cancer.
Combined fluoropyrimidines and oxaliplatin led to significant improvements in survival in three phase III
randomized controlled trials (Table 1).18-20 The MOSAIC
(Multicenter International Study of Oxaliplatin/5-FU/LV in
the Adjuvant Treatment of Colon Cancer) trial compared the
efficacy of the LV5FU2 regimen and the same regimen plus
oxaliplatin (FOLFOX4) in patients with stage II and III
colon cancer. Patients were randomly assigned to receive
12 biweekly cycles of LV5FU2 or FOLFOX4.18 For patients
with stage III disease, FOLFOX4 improved DFS by 24%
(HR 0.76, p 0.005). On the basis of these results,
FOLFOX4 was approved as adjuvant therapy after surgery
for patients with stage III colon cancer. Updated results
showed 5-year DFS rates of 58.9% and 66.4% for LV5FU2
and FOLFOX4, respectively (HR 0.78, p 0.005). A
benefit in OS for patients treated with FOLFOX4 was also
observed: the 6-year OS rates comparing LV5FU2 and
FOLFOX4 were 68.7% and 72.9%, respectively (HR 0.80,
p 0.023).18 The National Surgical Adjuvant Breast and
Bowel Project (NSABP) trial C-07 evaluated the FLOX
regimen (i.e., oxaliplatin added to a weekly bolus of 5-FU/
LV) in patients with stage II and III colon cancer.20 For both
stages, the benefit provided by oxaliplatin on 3-year DFS
was similar to that reported in the MOSAIC study (HR 0.80,
p 0.004). A longer follow-up indicated a 6.6% increase in
DFS at 5 years (57.8% vs. 64.4%, p 0.0007) for patients
with stage III disease, with a nonsignificant benefit in OS:
5-year OS rates with 5-FU/LV and FLOX were 73.8% and
76.5%, respectively (HR 0.85, p 0.052).
ADJUVANT THERAPY FOR STAGE III COLON CANCER
For both studies, the combination of 5-FU/oxaliplatin,

however, proved substantially more toxic than 5-FU alone,
resulting in a greater risk of severe cytopenia, nausea,
vomiting, diarrhea, and peripheral neuropathy.18,20
Finally, the superiority of XELOX (i.e., capecitabine/
oxaliplatin) over bolus 5-FU/LV as adjuvant treatment for
patients with stage III colon cancer was shown in the
NO16968 trial.19 Haller and colleagues reported a 3-year
DFS of 66.5% for 5-FU/LV (Mayo Clinic or Roswell Park
regimens) compared with 70.9% for XELOX (HR 0.80, p
0.0045). As anticipated with oxaliplatin, neurosensory toxicity occurred in the majority of patients in these three
studies, but the frequency of peripheral sensory neuropathy
decreased during the follow-up period.18 Preliminary
patient-reported outcomes from the NSABP C-07 trial suggest that symptoms of oxaliplatin-associated neurotoxicity
persisting for 6 years after treatment failed to reach clinical
significance.21 The total dose of oxaliplatin administered
in NSABP C-07 was approximately 30% lower than in
MOSAIC, yet provided comparable efficacy. The per-protocol
planned oxaliplatin dose was 1,020 mg/m2 (12 cycles) in
MOSAIC and 765 mg/m2 (nine cycles) in NSABP C-07,
although the median dose of oxaliplatin received per patient
was 810 mg/m2 (9.5 cycles) in MOSAIC and 667 mg/m2 (7.8
cycles) in NSABP C-07.18,20 Stopping oxaliplatin in cases of
neuropathy at grades 2 (e.g., paresthesia/dysesthesia interfering with function, but not with activities of daily living
[i.e., permanent dysesthesia]) or 3 (e.g., paresthesia/dysesthesia with pain or functional impairment) must be done in
clinical practice. If oxaliplatin has to be stopped because of
peripheral neuropathy, 5-FU/LV or capecitabine should be
administered per protocol (i.e., for a total of 6 months).
Could the mFOLFOX6 Regimen Replace FOLFOX4?
The mFOLFOX6 regimen (85 mg/m2 of intravenous oxaliplatin with 400 mg/m2 of LV over 2 hours, followed by 400
mg/m2 bolus injection of 5-FU, then 2,400 mg/m2 intravenous 5-FU for 46 hours) is more convenient for patients and
less expensive than the FOLFOX4 regimen (i.e., 1 day
compared with 2 days in the outpatient unit). In addition, no
evidence of further toxicity has resulted from the increased
dose of 5-FU. To date, no study has compared FOLFOX4 to
mFOLFOX6 in the adjuvant setting. However, in the
NSABP C-08 study,22 which evaluated the mFOLFOX6
regimen with or without bevacizumab in patients with stage
III disease, the 3-year DFS was nearly identical to that
observed in patients treated with FLOX in the oxaliplatin
arms of NSABP C-07 or those treated with FOLFOX4 in the
MOSAIC trial.18,20
Can Older Patients Benefit from Adjuvant Therapy?
The value and feasibility of 5-FU based adjuvant chemotherapy for patients older than age 70 was demonstrated in
a pooled analysis by Sargent and colleagues,23 who found no
evidence of interaction between age and the efficacy of
chemotherapy. Moreover, except for leucopenia in one study,
the incidence of toxic effects of chemotherapy was not higher
in older patients.
Adjuvant treatment with FOLFOX4 was well tolerated in
patients older than age 70, with higher rates of only neutropenia and thrombocytopenia compared with younger patients (49% vs. 43%, p 0.04%, and 5% vs. 2%, p 0.04,
respectively).24 A combined analysis of the two pivotal

adjuvant trials comparing 5-FU/oxaliplatin with 5-FU alone
failed to demonstrate a benefit in DFS and OS in older
patients, despite a positive trend for time to recurrence.25
On the other hand, the NO1698 trial,19 and the posthoc
analysis of MOSAIC data26 showed efficacy in both the
young and the older adult populations. Therefore no formal
recommendation can be offered. From a clinical point of
view, the decision to add oxaliplatin to fluoropyrimidines as
adjuvant treatment should be at least based on performance
status and comorbidities. In addition to these cancer-specific
prognostic factors, the use of a comprehensive geriatric
assessment is strongly recommended to evaluate the appropriateness of chemotherapy.
Lack of Benefit of Irinotecan in the Adjuvant Setting
As discussed, 5-FU leads to a relatively large improvement in OS for patients with stage III colon cancer. Interestingly, this benefit occurs despite a relatively low
radiographic response rate of 15% to 20%.27 With the addition of oxaliplatin and irinotecan for patients with metastatic disease, response rates more than doubled, suggesting
synergy in the cytotoxic effects on colorectal cancer cells.27
As such, it was anticipated that oxaliplatin would improve
the chance of remaining disease-free longer after surgery
and increase OS when added to 5-FU.18-20 Perhaps a presaging event to the current situation, however, was the story
of irinotecan in the adjuvant setting. When compared headto-head in the metastatic setting, irinotecan has proven
itself to be essentially equivalent to oxaliplatin (i.e., as an
addition to 5-FU).27 Despite this, four randomized trials of
irinotecan as an adjuvant addition for patients with stages
II and III colorectal cancer failed to demonstrate benefit,
demonstrating that improvements in response and OS in the
metastatic setting might not reliably translate to clear
benefit in the adjuvant setting.28-30 To this day, the failure
of irinotecan remains a puzzle, but it is also important to
understand that even a doubling of response rate with
oxaliplatin led to a fairly modest improvement in DFS
(HR 0.78 to 0.8) and OS (HR 0.8 to 0.85) in the MOSAIC
and NSABP C-07 studies, respectively.18,20 Perhaps, then,
what we can really conclude about adjuvant therapy for
patients with stage III colorectal cancer is that it is amazing
how large the benefit of 5-FU has been given the very low
response rate in stage IV disease.
Lack of Benefit of Biologics in the Adjuvant Setting:
Where Did We Go Wrong?
The first biologic agent to fail in the adjuvant setting was

bevacizumab. Mouse models of metastasis (tail-vein injection) performed in the 1990s suggested substantial inhibition of metastasis by the depletion of vascular endothelial
growth factor (VEGF) that could potentially translate to
benefit in the human adjuvant situation.31 It is welldocumented that bevacizumab led to improvements in
progression-free survival and OS in several trials of patients
with metastatic colorectal cancer, with an impressive HR for
death of 0.66 in the pivotal randomized study of bevacizumab.32 As such, hopes were high for benefit in the adjuvant setting.
The first published study of adjuvant bevacizumab was
NSABP C-08.22 This study randomly assigned 2,710 pa-
225
tients with stages II and III colorectal cancer after surgery

to a standard 6-month course of FOLFOX or FOLFOX plus
5 mg/kg of bevacizumab every other week followed by 6
months of single-agent bevacizumab. The study accrued
quickly, which reflected enthusiasm for the agent in this
setting. Disappointingly, the primary endpoint of 3-year
DFS did not improve significantly with bevacizumab (HR
0.87, p 0.08), although there was improvement in DFS at
1 year, suggesting either a biologic effect during bevacizumab treatment that did not lead to cure or that relapse
was observed later because of later surveillance scanning in
patients on the bevacizumab arm (i.e., median time to first
imaging was earlier in the control arm). Subsequent to this,
the 3,451-patient European AVANT (AVastinAdjuvaNT)
trial, which had a similar design to the NSABP C-08 trial
but included a third capecitabine/oxaliplatin/bevacizumab
arm, reported similarly negative results, including a transient benefit during the first year.33 Unlike NSABP C-08,
AVANT actually demonstrated a nonsignificant trend toward a decrement in survival for patients in both
bevacizumab-containing arms compared with FOLFOX
alone. In neither study did excess toxicity of bevacizumab
appear to be a contributing issue.
So what went wrong with bevacizumab? One question is
whether the scientific rationale for (relatively) short-term
VEGF inhibition was sound. Conceptually, we believe that
adjuvant therapy works by killing tumor cells, something
that bevacizumab is not known to do. If one were to consider
the static effects of VEGF inhibition to be valid, then ideally
bevacizumab would be continued for a long period, with the
expectation of potential relapse whenever one chose to end
therapy. We may never know if prolonged use of bevacizumab would be of benefit. In some experimental models,
bevacizumab has been shown to work in metastatic tumors
by normalizing aberrant tumor neovasculature,34 a mechanism that would not be expected to be of benefit in tumor
micrometastases. At this time, whether further study of
bevacizumab (or other VEGF pathwaytargeting strategies)
will be pursued remains unclear.
Antibodies that prevent the binding of ligands to the
epidermal growth factor receptor (EGFR) are the other class
of agent that has recently been studied in the adjuvant
setting. Cetuximab and panitumumab both have clear activity in subsets of patients with metastatic disease,35 and both
have been shown to increase response rates by roughly 10%
when added to FOLFOX or FOLFIRI (5-FU/leucovorin/
irinotecan) and when analysis is restricted to patients with
KRAS wild-type tumors. This increase in response rate,
along with single-agent responses that are similar to those
seen with 5-FU, would have suggested that the EGFR
antibodies might cause an increase in tumor-cell killing that
would make them good adjuvant therapy agents. With this
rationale in mind, study N0147 was designed by the North
Central Cancer Treatment Group (NCCTG) and conducted
by multiple U.S. and Canadian cooperative groups. N0147
went through several design iterations before settling as a
study of FOLFOX with or without cetuximab for patients
with resected stage III colorectal cancer.36,37 An interesting
subgroup of patients randomly assigned to receive FOLFIRI
with or without cetuximab was removed when it became
clear that irinotecan was an ineffective addition to 5-FU in
this setting.38 Over the course of the N0147 study, the KRAS
story emerged, and the study was amended to include only
226
patients with KRAS wild-type tumors. The final reported

sample size included 1,864 patients with KRAS wild-type
tumors. One issue that became clear was related to toxicity.
A clear signal emerged in patients older than 70 with high
rates of study discontinuation for toxicity, which resulted in
yet another amendment. After all was said and done, the
results of N0147 were again disappointing. Even patients
with KRAS wild-type tumors treated with cetuximab did not
benefit; in fact the trend was toward harm (DFS HR 1.2,
p NS).36 In addition, there was clear harm in patients
with KRAS-mutant cancers.37 In patients older than 70, the
HR was 1.79 (p 0.03), significantly favoring the FOLFOXonly arm.36
So again, what went wrong with cetuximab? In this case,
unexpected toxicities that could have shortened chemotherapy contributed to the negative outcome. As with irinotecan,
however, the ultimate cause of failure is unclear. One hint
that this study might have been negative was the relatively
small incremental improvements in outcome in most trials
of patients with metastatic cancer (a disconnect from the
relatively robust increase in radiographic response). Could it
be that micrometastases have not yet become dependent on
EGFR ligands for their growth? Could early-stage tumors
be biologically different and less likely to depend on EGFR
signaling?
Curiously, data from the N0147 study now suggest that
cetuximab might benefit patients when given with the
discontinued irinotecan backbone, although interpreting
these data has to be tempered by the small sample size.38
In a poster presented at the ASCO Gastrointestinal Cancers
Symposium 2011, N0147 investigators demonstrated a
fairly robust benefit trend in the FOLFIRI subgroup (3-year
DFS HR 0.4, p 0.05; 3-year OS HR 0.3, p 0.08).38
Given the irinotecan data, whether this information will be
followed up in larger trials remains to be seen.
Current Adjuvant Therapy Trials
To date, adjuvant therapy trials for patients with colon

cancer, as in most other cancers, develop because of promising data in metastatic disease. Whether this strategy is
optimal is debatable since certain agents might have more
efficacy in a micrometastatic setting, though this remains
the current approach. Hampering the testing of new agents
in adjuvant colon cancer, therefore, has been the recent lack
of new agents with definitive efficacy in metastatic disease.
The last drug to be approved by the FDA was panitumumab
in 2006.
Given the lack of agents to test in the adjuvant setting,
questions have arisen related to the necessary duration of
therapy. In the late 1980s and early 1990s, INT-0089 demonstrated noninferiority between 6 months of adjuvant chemotherapy with 5-FU/LV and 12 months with 5-FU/
levamisole.12 In addition, the GERCOR trial showed that
the LV5FU2 regimen administered for either 6 or 9 months
achieved similar results.13 Therefore, a 6-month duration of
treatment became the standard. Given the cumulative neuropathy from oxaliplatin that can remain with patients for
years after completing adjuvant therapy,18 the question of
needed duration of adjuvant therapy became pertinent.
Further, one small trial by Chau and colleagues demonstrated noninferiority of a protracted venous infusion regimen of 5-FU for 3 months compared to the Mayo Clinic
(5-FU/LV) regimen in adjuvant therapy for patients with

Table 2. Ongoing Adjuvant Trials for Colon Cancer Testing Duration of FOLFOX or XELOX Therapy (3 versus 6 Months of Therapy)
TOSCA
SCOT
CALGB/SWOG 80,702
GERCOR
HORG
Country
Italy
United Kingdom plus other

European sites
United States and Canada
France
Greece
Principal Investigator (s)
Alberto Sobrero
Timothy Iveson and

Jim Paul
Jeffrey Meyerhardt and

Anthony Shields
Thierry Andre and

Julien Taieb
Ioannis Souglakos
Accrual Goal
3,500
9,500
2,500
2,000
1,000
Inclusion Criteria
High risk stage II and III colon

cancer
Stage II and III colon and

rectal cancer
Stage III colon cancer
Stage III colon

cancer
High risk stage II

and stage III
colon cancer
Additional Features
Previously included bevacizumab

randomization but dropped
after other bevacizumab trials
reports
stage II and III colorectal cancer.39 In response to these

concerns, at least five randomized trials have opened and
are accruing patients to test the noninferiority of 3 months
compared with 6 months of a fluoropyrimidine/oxaliplatin
combination therapy (Table 2). Investigators have agreed to
pool data from all patients with stage III disease enrolled in
the trials to test for noninferiority through a preplanned
pooling project, the International Duration Evaluation of
Adjuvant Chemotherapy (IDEA) prospective pooled analysis. Noninferiority will be declared if the two-sided 95%
confidence interval for the HR comparing 3 to 6 months of
therapy lies entirely below 1.10, which translates to an
approximately 1.5% difference in DFS.
Adjunctive Therapy Considerations
Epidemiologic and scientific research indicates that diet

and other lifestyle factors influence the risk of developing
colorectal cancer.40 Obesity, consumption of red meat, a
Western-pattern diet, alcohol, and smoking influence ones
risk of developing colorectal cancer; physical activity, calcium, vitamin D, postmenopausal estrogen use, aspirin,
nonsteroidal anti-inflammatory drugs (NSAIDs), and possibly folate decrease ones risk. Until recently, whether these
factors influence outcomes in patients already diagnosed
with colorectal cancer was largely unknown. Data on factors
that may influence disease recurrences and mortality for
survivors of colon cancer are emerging.
Energy-Balance Factors and Colon Cancer Outcomes
Multiple prospective cohort studies have tested the influence of physical activity before and after diagnosis and
changes in physical activity before and after diagnosis.41-45
Haydon and colleagues reported that physical activity before
diagnosis was associated with a 51% improvement in DFS in
patients with stage II and III colorectal cancer.46 In contrast, physical activity before diagnosis did not influence
colorectal cancerspecific mortality in women diagnosed
with stages I to III colorectal cancer participating in the
Nurses Health Study, but activity after diagnosis did improve outcomes. Compared to women engaged physical activity less than 3 metabolic equivalent task (MET)-hours
per week (i.e., considered fairly inactive), those engaged in
at least 18 MET-hours per week (i.e., the equivalent to 1
hour of moderate walking daily at least 6 days per week)
had an adjusted HR for colorectal cancerspecific mortality
of 0.39 (95% CI, 0.18 to 0.82) and an adjusted HR for overall
2 2 randomization - duration
question and celecoxib or
placebo
mortality of 0.43 (95% CI, 0.25 to 0.74).41 Further, women

who increased their activity after diagnosis had an HR of
0.48 (95% CI, 0.24 to 0.97) for death from colorectal cancer
and an HR of 0.51 (95% CI, 0.30 to 0.85) for death from
any cause, compared to those with no change in activity.
Similar results were seen in a cohort of male health professionals participating in the Health Professionals Follow-up
Study.44 These results were further supported by data from
a prospective cohort study within an adjuvant therapy trial
sponsored by the National Cancer Institute (Cancer and
Leukemia Group B [CALGB] trial 89803).42 Among 832
patients with stage III colon cancer who survived and were
recurrence-free approximately 6 months after adjuvant
chemotherapy, physical activity after diagnosis improved
DFS by approximately 50% beyond surgery and adjuvant
chemotherapy. In addition, a recent cohort from Australia
of more than 1,800 survivors of stage I to III colorectal
cancer demonstrated that physical activity after diagnosis
improved overall mortality by approximately 25%.45 These
data led to a multinational trial in Canada and Australia,
the Colon Health and Life-Long Exercise Change
(CHALLENGE) trial, designed to determine the effects of
a 3-year structured and supervised physical activity intervention on disease outcomes in 962 high-risk survivors of
stage II and III colon cancer who completed adjuvant chemotherapy within the previous 2 to 6 months.47 The trial is
currently open to accrual.
Although the initial assumption may be that physically
active survivors of colorectal cancer have lower body mass
indexes or that exercise lowers body weight, most studies
examining obesity and colorectal cancer survival have
shown only modest associations, if any.48 Only one study has
reported on the effect of change in weight on outcomes in
survivors of colorectal cancer. Patients enrolled in CALGB
89803 reported weight during chemotherapy and approximately 6 months after completion of chemotherapy and
change in weight or body mass index was not associated
with either DFS or OS.49
Although various dietary factors have been associated
with the development of colorectal cancer, only one large,
prospective cohort study has tested whether diet is associated with outcomes in survivors of colon cancer.50 In CALGB
89803, participants completed a food frequency questionnaire during adjuvant therapy and 6 months after the
completion of adjuvant therapy. Using these data, two major
dietary patterns were identified: prudent pattern, character-
227
ized by high fruit, vegetable, poultry, and fish intakes, and

Western pattern, characterized by high meat, fat, refined
grains, and dessert intakes. All patients were assigned a
relative value along the spectrum of both dietary patterns
and the two patterns were not correlated with each other.
Compared with patients in the lowest quintile of the
Western-pattern diet, those in the highest quintile experienced an adjusted HR for DFS of 3.25 (95% CI, 2.04 to 5.19;
p for trend 0.0001). In contrast, the prudent-pattern diet
was not significantly associated with cancer recurrence or
mortality. Studies on the components of the Westernpattern diet that influenced these findings are underway.
NSAIDs, Vitamin D, and Colon Cancer Outcomes
Outside of energy-balance factors, research is emerging on

the effects of aspirin, NSAIDs, and vitamin D. Prospective
cohort studies have demonstrated that aspirin and NSAIDs
reduce the risk of developing colorectal cancer, and several
randomized trials have shown that these agents reduce the
number of subsequent polyps in patients with a history of
adenomatous polyps.51-54 Initial studies have also shown a
beneficial association of these agents for survivors of colorectal cancer.55 In the CALGB 89803 cohort of patients with
stage III colon cancer, regular use of aspirin (defined as
consistent use both while receiving adjuvant therapy and for
6 months after the completion of adjuvant therapy) had a
significant 54% improvement in DFS compared with patients that did not regularly use aspirin (p 0.001).55
Similarly, for patients with stages I to III colorectal cancer
from both the Nurses Health Study and Health Professional
Follow-up Study, regular aspirin users had a more modest
but still significant improvement in colorectal cancer mortality compared to nonusers (HR 0.72; 95% CI, 0.54 to
0.97, p 0.03).56
The Vioxx in Colorectal Cancer Therapy: Definition of
Optimal Regime (VICTOR) trial was a randomized controlled trial in the United Kingdom evaluating the role of
rofecoxib, a cyclooxygenase 2 inhibitor, for patients with
stage II and III colon cancer who completed adjuvant therapy.57 The trial activated in April 2002 but was terminated
in September 2004 after the withdrawal of rofecoxib from
the market. Before study closure, 2,434 patients were randomly assigned to placebo, 2 years of rofecoxib, or 5 years of
rofecoxib, with a median duration of time on trial of 7.4
months. The HR for DFS with a median follow-up of 3 years
was 0.90 (95% CI, 0.77 to 1.06), favoring rofecoxib among
patients with stage II and III disease. However the early
closure of the trial and the limited time patients received the
study medication in relationship to the intent of the trial
limit conclusions from this trial. Given the observational
data and the potential benefit seen in VICTOR, the CALGB

and Southwestern Oncology Group (SWOG) have developed
a trial (CALGB/SWOG 80702, Table 2) that tests the duration of FOLFOX therapy, with a second randomization of
celecoxib compared with placebo initiated concurrent with
the start of FOLFOX and continuing for 3 years. The trial
will test for superiority in DFS of celecoxib compared with
placebo.
Increasing evidence suggests an association between vitamin D and development of colorectal cancer. A metaanalysis of five epidemiologic studies found a 51% decrease
in the risk of colorectal cancer associated with plasma
25(OH)D levels in the highest quintile compared to those in
the lowest quintile (p 0.0001).58 Although studies on
vitamin D in patients with colon cancer are limited, two
recent studies by Ng and colleagues using the Nurses
Health Study and the Health Professional Follow-up Study
cohorts suggest that either higher levels of serum vitamin D
before diagnosis or greater predictive vitamin D serum
levels are associated with improved colorectal cancerspecific mortality, overall mortality, or both.59,60
Conclusion
Adjuvant chemotherapy has taken an important place in

the treatment of patients with stage III colon cancer. The
5-year survival of patients with stage III colon cancer has
increased from less than 50% with surgery alone (before
1990) to more than 70% with surgery followed by adjuvant
fluoropyrimidines combined with oxaliplatin (since 2003).
However, despite this progress, multiple disappointments
have emerged including the realization that three of the four
agent classes approved for metastatic colorectal cancer since
1995 have failed to improve the chance of cure for patients
with stages II and III colorectal cancer. We must learn from
the experiences of these trials to avoid future expensive
failures, and must also begin to consider that not all colon
cancers are the same. The future, as is the case for most
areas of cancer, must involve rational trial designs based
on well-characterized and clinically validated biology. The
study of agents with marginal survival benefit in unselected
populations is discouraged based on the experiences outlined above. Although most of the current trials are not
adding new agents to FOLFOX (with the exception of the
CALGB/SWOG 80702 trial), the focus on the potential for
less therapy and presumed less long-term neuropathy is
important. Finally, beyond standard chemotherapies, continued research on other adjunctive therapies is critical to
further understand the role of energy balance and other host
factors in colon cancer survival.
Author
Thierry Andre
Bert H. ONeil
Jeffrey A. Meyerhardt
228
Employment or
Leadership
Positions
Consultant or
Advisory Role
Roche; Sanofi
Amgen;
Bayer/Onyx
Stock
Ownership
Honoraria
Baxter
International;
Roche; Yacult
Amgen
Bayer
Research
Funding
AstraZeneca;
Bristol-Myers
Squibb
Expert
Testimony
Other
Remuneration
REFERENCES
1. Ferlay J, Shin HR, Bray F, et al. Estimates of worldwide burden of
cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893-2917.
2. Chou JF, Row D, Gonen M, et al. Clinical and pathologic factors that
predict lymph node yield from surgical specimens in colorectal cancer: a
population-based study. Cancer. 2010;116:2560-2570.
3. Edge SB, Compton CC. The American Joint Committee on Cancer: the
7th edition of the AJCC Cancer Staging Manual and the future of TNM. Ann
Surg Oncol. 2010;17:1471-1474.
4. Moertel CG, Fleming TR, Macdonald JS, et al. Levamisole and fluorouracil for adjuvant therapy of resected colon carcinoma. N Engl J Med.
1990;322:352-358.
5. Moertel CG, Fleming TR, Macdonald JS, et al. Fluorouracil plus levamisole as effective adjuvant therapy after resection of stage III colon carcinoma:
a final report. Ann Intern Med. 1995;122:321-326.
6. Heidelberger C, Chaudhuri NK, Danneberg P, et al. Fluorinated pyrimidines, a new class of tumour-inhibitory compounds. Nature. 1957;179:663666.
7. Moertel CG. Chemotherapy of gastrointestinal cancer. N Engl J Med.
1978;299:1049-1052.
8. Buyse M, Zeleniuch-Jacquotte A, Chalmers TC. Adjuvant therapy of
colorectal cancer. Why we still dont know. JAMA. 1988;259:3571-3578.
9. Efficacy of adjuvant fluorouracil and folinic acid in colon cancer. International Multicentre Pooled Analysis of Colon Cancer Trials (IMPACT)
investigators. Lancet. 1995;345:939-944.
10. OConnell MJ, Mailliard JA, Kahn MJ, et al. Controlled trial of
fluorouracil and low-dose leucovorin given for 6 months as postoperative
adjuvant therapy for colon cancer. J Clin Oncol. 1997;15:246-250.
11. Wolmark N, Rockette H, Mamounas E, et al. Clinical trial to assess the
relative efficacy of fluorouracil and leucovorin, fluorouracil and levamisole,
and fluorouracil, leucovorin, and levamisole in patients with Dukes B and C
carcinoma of the colon: results from National Surgical Adjuvant Breast and
Bowel Project C-04. J Clin Oncol. 1999;17:3553-3559.
12. Haller DG, Catalano PJ, Macdonald JS, et al. Phase III study of
fluorouracil, leucovorin, and levamisole in high-risk stage II and III colon
cancer: final report of Intergroup 0089. J Clin Oncol. 2005;23:8671-8678.
13. Andre T, Colin P, Louvet C, et al. Semimonthly versus monthly
regimen of fluorouracil and leucovorin administered for 24 or 36 weeks as
adjuvant therapy in stage II and III colon cancer: results of a randomized
14. Sargent D, Sobrero A, Grothey A, et al. Evidence for cure by adjuvant
therapy in colon cancer: observations based on individual patient data from
20,898 patients on 18 randomized trials. J Clin Oncol. 2009;27:872-877.
15. Twelves C, Wong A, Nowacki MP, et al. Capecitabine as adjuvant
treatment for stage III colon cancer. N Engl J Med. 2005;352:2696-2704.
16. Lembersky BC, Wieand HS, Petrelli NJ, et al. Oral uracil and tegafur
plus leucovorin compared with intravenous fluorouracil and leucovorin in
stage II and III carcinoma of the colon: results from National Surgical
Adjuvant Breast and Bowel Project Protocol C-06. J Clin Oncol. 2006;24:20592064.
17. Sargent DJ, Wieand HS, Haller DG, et al. Disease-free survival versus
overall survival as a primary end point for adjuvant colon cancer studies:
individual patient data from 20,898 patients on 18 randomized trials. J Clin
Oncol. 2005;23:8664-8670.
18. Andre T, Boni C, Navarro M, et al. Improved overall survival with
oxaliplatin, fluorouracil, and leucovorin as adjuvant treatment in stage II or
III colon cancer in the MOSAIC trial. J Clin Oncol. 2009;27:3109-3116.
19. Haller DG, Tabernero J, Maroun J, et al. Capecitabine plus oxaliplatin
compared with fluorouracil and folinic acid as adjuvant therapy for stage III
colon cancer. J Clin Oncol. 2011;29:1465-1471.
20. Kuebler JP, Wieand HS, OConnell MJ, et al. Oxaliplatin combined
with weekly bolus fluorouracil and leucovorin as surgical adjuvant chemotherapy for stage II and III colon cancer: results from NSABP C-07. J Clin
Oncol. 2007;25:2198-2204.
21. Land SR, Kopec JA, Cecchini RS, et al. Neurotoxicity from oxaliplatin
combined with weekly bolus fluorouracil and leucovorin as surgical adjuvant
chemotherapy for stage II and III colon cancer: NSABP C-07. J Clin Oncol.
2007;25:2205-2211.
22. Allegra CJ, Yothers G, OConnell MJ, et al. Phase III trial assessing
bevacizumab in stages II and III carcinoma of the colon: results of NSABP
protocol C-08. J Clin Oncol. 2011;29:11-16.
23. Sargent DJ, Goldberg RM, Jacobson SD, et al. A pooled analysis of
adjuvant chemotherapy for resected colon cancer in elderly patients. N Engl
J Med. 2001;345:1091-1097.
24. Goldberg RM, Tabah-Fisch I, Bleiberg H, et al. Pooled analysis of safety
and efficacy of oxaliplatin plus fluorouracil/leucovorin administered bi-
monthly in elderly patients with colorectal cancer. J Clin Oncol. 2006;24:

4085-4091.
25. Jackson McCleary NA, Meyerhardt J, Green E, et al. Impact of older
age on the efficacy of newer adjuvant therapies in 12,500 patients (pts) with
stage II/III colon cancer: findings from the ACCENT database. J Clin Oncol.
2009;27:15s (suppl; abstr 4010).
26. Tournigand C, Andre T, Bachet J, et al. FOLFOX4 as adjuvant therapy
in elderly patients (pts) with colon cancer (CC): subgroup analysis of the
MOSAIC trial. J Clin Oncol. 2010;28:15s (suppl; abstr 3522).
27. Lucas AS, ONeil BH, Goldberg RM. A decade of advances in cytotoxic
chemotherapy for metastatic colorectal cancer. Clin Colorectal Cancer. 2011;
10:238-244.
28. Papadimitriou CA, Papakostas P, Karina M, et al. A randomized phase
III trial of adjuvant chemotherapy with irinotecan, leucovorin and fluorouracil versus leucovorin and fluorouracil for stage II and III colon cancer: a
Hellenic Cooperative Oncology Group study. BMC Med. 2011;9:10.
treatment for stage III colon cancer: results of CALGB 89803. J Clin Oncol.
2007;25:3456-3461.
30. Van Cutsem E, Labianca R, Bodoky G, et al. Randomized phase III trial
comparing biweekly infusional fluorouracil/leucovorin alone or with irinotecan in the adjuvant treatment of stage III colon cancer: PETACC-3. J Clin
Oncol. 2009;27:3117-3125.
31. Warren RS, Yuan H, Matli MR, et al. Regulation by vascular endothelial growth factor of human colon cancer tumorigenesis in a mouse model of
experimental liver metastasis. J Clin Invest. 1995;95:1789-1797.
32. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus
irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer.
N Engl J Med. 2004;350:2335-2342.
33. De Gramont A, Van Cutsem E, Tabernero J, et al. AVANT: Results
from a randomized, three-arm multinational phase III study to investigate
bevacizumab with either XELOX or FOLFOX4 versus FOLFOX4 alone as
adjuvant treatment for colon cancer. J Clin Oncol. 2011; (suppl; abstr 362)
34. Carmeliet P, Jain RK. Principles and mechanisms of vessel normalization for cancer and other angiogenic diseases. Nat Rev Drug Discov. 2011;10:
417-427.
35. Stein S, Cohen DJ, Hochster HS. Treatment paradigms with epidermal
growth factor receptor-targeted therapies in colorectal cancer. Clin Colorectal
Cancer. 2010; 9 Suppl 1:S44-S50.
36. Alberts SR, Sargent DJ, Smyrk TC, et al. Adjuvant mFOLFOX6 with or
without cetuxiumab (Cmab) in KRAS wild-type (WT) patients (pts) with
resected stage III colon cancer (CC): results from NCCTG Intergroup Phase
III Trial N0147. J Clin Oncol. 2010;28:18s (suppl; abstr CRA3507).
37. Goldberg RM, Sargent DJ, Thibodeau SN, et al. Adjuvant mFOLFOX6
plus or minus cetuximab (Cmab) in patients (pts) with KRAS mutant (m)
resected stage III colon cancer (CC): NCCTG Intergroup Phase III Trial
N0147. J Clin Oncol. 2010;28:15s (suppl; abstr 3508).
38. Huang J, Sargent DJ, Mahoney MR, et al. Adjuvant FOLFIRI with or
without cetuximab in patients with resected stage III colon cancer: NCCTG
Intergroup phase III trial N0147. J Clin Oncol. 2011; (suppl; abstr 363).
39. Chau I, Norman AR, Cunningham D, et al. A randomised comparison
between 6 months of bolus fluorouracil/leucovorin and 12 weeks of protracted
venous infusion fluorouracil as adjuvant treatment in colorectal cancer. Ann
Oncol. 2005;16:549-557.
40. Wei EK, Giovannucci E, Wu K, et al. Comparison of risk factors for
colon and rectal cancer. Int J Cancer. 2004;108:433-442.
41. Meyerhardt JA, Giovannucci EL, Holmes MD, et al. Physical activity
and survival after colorectal cancer diagnosis. J Clin Oncol. 2006;24:35273534.
42. Meyerhardt JA, Heseltine D, Niedzwiecki D, et al. Impact of physical
activity on cancer recurrence and survival in patients with stage III colon
cancer: findings from CALGB 89803. J Clin Oncol. 2006;24:3535-3541.
43. Haydon AM, Macinnis R, English D, et al. Effect of physical activity
and body size on survival after diagnosis with colorectal cancer. Gut. 2006;
55:62-67.
44. Meyerhardt JA, Giovannucci EL, Ogino S, et al. Physical activity and
male colorectal cancer survival. Arch Intern Med. 2009;169:2102-2108.
45. Baade PD, Meng X, Youl PH, et al. The impact of body mass index and
physical activity on mortality among patients with colorectal cancer in
Queensland, Australia. Cancer Epidemiol Biomarkers Prev. 2011;20:1410-1420.
46. Haydon AM, Macinnis RJ, English DR, et al. Physical activity, insulinlike growth factor 1, insulin-like growth factor binding protein 3, and survival
from colorectal cancer. Gut. 2006;55:689-694.
47. Courneya KS, Booth CM, Gill S, et al. The Colon Health and Life-Long
229

Exercise Change trial: a randomized trial of the National Cancer Institute of
Canada Clinical Trials Group. Curr Oncol. 2008;15:271-278.
48. Meyerhardt JA, Ma J, Courneya KS. Energetics in colorectal and
prostate cancer. J Clin Oncol. 2010;28:4066-4073.
49. Meyerhardt JA, Niedzwiecki D, Hollis D, et al. Impact of body mass
index and weight change after treatment on cancer recurrence and survival in
patients with stage III colon cancer: findings from Cancer and Leukemia
Group B 89803. J Clin Oncol. 2008;26:4109-4115.
50. Meyerhardt JA, Niedzwiecki D, Hollis D, et al. Association of dietary
patterns with cancer recurrence and survival in patients with stage III colon
cancer. JAMA. 2007;298:754-764.
51. Baron JA, Cole BF, Sandler RS, et al. A randomized trial of aspirin to
prevent colorectal adenomas. N Engl J Med. 2003;348:891-899.
52. Baron JA, Sandler RS, Bresalier RS, et al. A randomized trial of
rofecoxib for the chemoprevention of colorectal adenomas. Gastroenterology.
2006;131:1674-1682.
53. Bertagnolli MM, Eagle CJ, Zauber AG, et al. Celecoxib for the prevention of sporadic colorectal adenomas. N Engl J Med. 2006;355:873-884.
54. Chan AT, Giovannucci EL, Meyerhardt JA, et al. Long-term use of
aspirin and nonsteroidal anti-inflammatory drugs and risk of colorectal
cancer. JAMA. 2005;294:914-923.
230
55. Fuchs C, Meyerhardt JA, Heseltine DL, et al. Influence of regular

aspirin use on survival for patients with stage III colon cancer: findings from
Intergroup trial CALGB 89803. J Clin Oncol. 2005;23: (suppl abstr 3530).
56. Chan AT, Ogino S, Fuchs CS. Aspirin use and survival after diagnosis
of colorectal cancer. JAMA. 2009;302:649-659.
57. Kerr DJ, Dunn JA, Langman MJ, et al. VICTOR: a phase III placebocontrolled trial of rofecoxib in colorectal cancer patients following surgical
resection. J Clin Oncol. 2008;26: (suppl abstr 4038).
58. Gorham ED, Garland CF, Garland FC, et al. Optimal vitamin D status
for colorectal cancer prevention: a quantitative meta analysis. Am J Prev
Med. 2007;32:210-216.
59. Ng K, Meyerhardt JA, Wu K, et al. Circulating 25-hydroxyvitamin d
levels and survival in patients with colorectal cancer. J Clin Oncol. 2008;26:
2984-2991.
60. Ng K, Wolpin BM, Meyerhardt JA, et al. Prospective study of predictors
of vitamin D status and survival in patients with colorectal cancer. Br J
Cancer. 2009;101:916-923.
61. Andre T, Boni C, Mounedji-Boudiaf L, et al. Oxaliplatin, fluorouracil,
and leucovorin as adjuvant treatment for colon cancer. N Engl J Med.
2004;350:2343-2351.
EVOLVING PARADIGMS IN THE MANAGEMENT OF

UNRESECTABLE PANCREATIC CANCER
CHAIR
The University of Chicago Medical Center
Chicago, IL
SPEAKERS
Theodore S. Hong, MD
Boston, MA
Lauren A. Wiebe, MD
Milwaukee, WI
A New Direction for Pancreatic Cancer

Treatment: FOLFIRINOX in Context
By Hedy Lee Kindler, MD
Overview: Since 1996, the cornerstone of chemotherapy for

advanced pancreatic cancer has been gemcitabine, which has
a genuine, but modest effect on survival and quality of life. It
has been remarkably difficult to improve on these outcomes.
Many phase III studies of gemcitabine doublets have been
uniformly negative, with the exception of a trial of gemctabine
plus erlotinib, which provided only marginal benefit. In 2010,
the FOLFIRINOX regimen (bolus and infusional 5-fluorouracil,
irinotecan, and oxaliplatin) emerged as a major treatment
advance for patients with metastatic pancreatic cancer. In a
trial with 342 patients, FOLFIRINOX yielded a longer median
overall survival (11.1 vs. 6.8 months, hazard ratio [HR] 0.57,
p < 0.001), a superior progression-free survival (6.4 vs. 3.3
months, HR 0.47, p < 0.001), a higher objective response rate
(31.6% vs. 9.4%, p < 0.001), and a significant increase in time
until definitive deterioration in quality of life, compared with

gemcitabine. FOLFIRINOX is also more cost-effective than
gemcitabine. Because of higher rates of grade 3 to 4 neutropenia (46% vs. 21%), febrile neutropenia (5% vs. 1%), and
diarrhea (13% vs. 2%) with FOLFIRINOX, vigilant patient
selection, education, and monitoring are essential. Retrospective single-institution series confirm the substantial activity of
FOLFIRINOX in metastatic, locally advanced, and previouslytreated patients; demonstrate its safety in individuals with
biliary stents; and elucidate how physicians routinely modify
drug doses without clear evidence or guidelines. Ongoing and
planned studies will prospectively evaluate FOLFIRINOX in
the adjuvant, locally advanced, and borderline resectable
settings, will add targeted agents to FOLFIRINOX, and will
evaluate how to adjust doses to ameliorate toxicity.
Background
4/ACCORD 11 trial, presented first at ASCO in 201017 and

published in the New England Journal of Medicine in
2011,18 represent a substantial treatment advance for patients with pancreatic cancer. The multidrug combination FOLFIRINOX significantly improved median and
progression-free survival, objective response rates, and quality of life, albeit with greater toxicity. Given all of the
negative trials that preceded it, these data alone are a major
achievement. What is truly unprecedented is the magnitude
of the benefit achieved with this regimen.
In this article, we will review the data behind this pivotal
study, examine how oncologists are currently using this
regimen, and assess the ways that FOLFIRINOX may be
incorporated into pancreatic cancer treatment in the future.
HERE HAS been a long-standing, well-deserved therapeutic nihilism regarding chemotherapy for advanced
pancreatic cancer. In countless trials over the past few
decades, many drugs and drug combinations have demonstrated minimal to no activity against this devastating
disease.
Since 1996, the cornerstone of therapy has been gemcitabine, an agent with a genuine, but modest impact. In the
pivotal trial that compared gemcitabine to weekly bolus
5-fluorouracil (5-FU), gemcitabine treatment yielded a response rate of 5% and a median overall survival of 5.7
months.1 Although these results do represent a real advance
over 5-FU, gemcitabine is principally given because it provides a clinical benefit, a combination of an improvement of
pain and performance status, and a stabilization of weight.
Despite these very modest outcomes, it has been difficult
for any agent to displace gemcitabine for advanced pancreatic cancer. Most drugs simply do not work in this disease.
Although innumerable phase II trials have reported promising activity for various gemcitabine-based cytotoxic and
targeted doublets, phase III trials of these combinations
have been uniformly disappointing, generally yielding
greater toxicity for the multidrug regimen with no improvement in overall survival.2-7
This dismal outlook changed slightly in 2005, when the
National Cancer Institute of Canada PA.3 trial demonstrated a small improvement in survival for patients treated
with gemcitabine plus erlotinib, an epidermal growth factor
receptor tyrosine kinase inhibitor.8 Although the results
were statistically significant, with a hazard ratio of 0.82, the
absolute improvement in median overall survival, 5.91
months with gemcitabine compared with 6.24 months for
gemcitabine/erlotinib, was minimal. This combination also
came with a substantial economic cost9 and did not improve
quality of life. One could easily question whether such an
incremental improvement in overall survival was worth the
expense and toxicity.10 Over the next 5 years, several more
negative phase III trials were reported,11-16 and it certainly
appeared that any major improvements were a long way off.
It was in this context that the results of the PRODIGE
232
Development of the FOLFIRINOX Regimen: Phase I

and II Studies
It would seem logical to combine 5-FU, irinotecan, and

oxaliplatin: these 3 drugs have activity in several gastrointestinal malignancies, including pancreatic cancer, without
many overlapping toxicities. The doses for the FOLFIRINOX regimen were established in a phase I trial of 34
evaluable patients with advanced solid tumors enrolled
between 1998 and 2000.19 Although the investigators may
have been primarily interested in developing this combination for metastatic colon cancer, they also observed two
responses (one complete and one partial) among the six
patients with pancreatic cancer enrolled on the study.
These encouraging data prompted the same investigators
to evaluate this regimen in a single-arm phase II trial.20
Forty-six chemotherapy-nave patients with pancreatic cancer with a World Health Organization (WHO) performance
status of 0 or 1 enrolled at nine French centers from 2000 to
2002. Most subjects (76%) had metastatic disease, a perfor-
From the University of Chicago, Chicago, IL.

Address reprint requests to Hedy Lee Kindler, MD, Section of Hematology/Oncology,
University of Chicago, 5841 S. Maryland Ave., MC 2115, Chicago, IL 60637; email:
hkindler@medicine.bsd.uchicago.edu.
1092-9118/10/1-10
FOLFIRINOX FOR PANCREATIC CANCER
mance status (PS) of 1 (74%), and liver metastases (61%).

Responses, assessed using WHO criteria21 by an external
review committee, were observed in 26% of the patients
(95% CI, 13% to 39%) and included 4% complete responses.
Thirty-five percent of the subjects had stable disease, yielding a disease control rate of 61%. The median progressionfree survival was 8.2 months (95% CI, 5.3 to 11.6 months).
The median overall survival was 10.2 months (95% CI, 8.1 to
14.4 months); in patients with metastatic disease, the median overall survival was 9.9 months. Grades 3 4 neutropenia and diarrhea developed in 52% and 17% of patients,
respectively. Quality of life, measured by the EORTC QLQC30, was improved in all functional scales except for cognitive functioning.
The PRODIGE 4/ACCORD 11 Study
The impressive activity observed with FOLFIRINOX in

the phase I and II trials19,20 prompted the development of
a randomized phase II/III study that compared this new
regimen with the benchmark therapy, gemcitabine.18 Key
eligibility criteria included no prior chemotherapy, an
ECOG Performance Status of 0 or 1, age less than 76 years,
measurable metastatic disease, and a total bilirubin less
than 1.5 times the upper limit of normal.
Patients, stratified by center, performance status (0 vs. 1),
and location of the primary tumor (head vs. body/tail)
were randomized 1:1 to either FOLFIRINOX (oxaliplatin
KEY POINTS
In a pivotal trial in 342 patients with metastatic

pancreatic cancer and a good performance status,
FOLFIRINOX (bolus and infusional 5-fluorouracil,
irinotecan, and oxaliplatin) produced a longer median
overall survival (11.1 vs. 6.8 months), a superior
progression-free survival (6.4 vs. 3.3 months), a
higher objective response rate (31.6% vs. 9.4%), and a
significant increase in the time until definitive deterioration in quality of life compared with gemcitabine.
FOLFIRINOX is more cost-effective than gemcitabine.
Vigilant patient selection, education, and monitoring
are essential with FOLFIRINOX treatment, because
of higher rates of grade 3 4 neutropenia, febrile
neutropenia, and diarrhea compared with gemcitabine.
Retrospective single-institution series confirm the
substantial activity of FOLFIRINOX in metastatic,
locally advanced, and previously-treated patients;
demonstrate its safety in patients with biliary stents;
and elucidate how physicians routinely modify drug
doses without clear evidence or guidelines.
Ongoing and planned studies will prospectively evaluate FOLFIRINOX in the adjuvant, locally advanced,
and borderline resectable settings, will add targeted
agents to FOLFIRINOX, and will evaluate how to
adjust doses to ameliorate toxicity.
85 mg/m2, leucovorin 400 mg/m2, irinotecan 180 mg/m2,

bolus 5-fluorouracil 400 mg/m2 followed by infusional
5-fluorouracil 2,400 mg/m2 given over 46 hours, every 14
days) or gemcitabine (1,000 mg/m2 over 30 minutes weekly
for 7 of 8 weeks, then weekly for 3 of 4 weeks). Each cycle
was defined as a 2-week interval for both regimens. Six
months of treatment were recommended for responding
patients. Filgrastim was not routinely administered for
primary prophylaxis, though it was permitted for high-risk
patients. Computed tomography (CT) scans were obtained
every 2 months. RECIST criteria were employed for response assessment.22 Quality of life was measured by
EORTC QLQ-C30 questionnaires completed every 2 weeks.
Response was the primary efficacy endpoint of the phase
II portion of the study, which was planned to proceed to
phase III if at least 12 objective responses occurred in the
first 40 FOLFIRINOX-treated patients. Overall survival
was the primary phase III endpoint. Phase II patients were
included in the phase III analysis.
The study was designed to have an 80% power to detect an
increase in median overall survival from 7 to 10 months (HR
0.70, 0.05). For the final analysis, 360 patients would be
required to reach 250 events; an interim analysis was
planned after 167 events occurred.23 In September 2009, the
Independent Data Management Committee recommended
that accrual be stopped early, because a planned interim
analysis determined that the primary endpoint was
achieved with a p value of less than 0.001.18,23
Between 2005 and 2009, 342 patients enrolled at 48
French centers. Patient characteristics were balanced between the arms for age, sex, performance status, tumor
location, biliary stenting, metastatic sites, and baseline Ca
19 9 level, except that a greater percentage of patients on
the gemcitabine arm had measurable pulmonary metastases
(29% vs. 19%). The median age was 61. Approximately 60%
of the subjects in both arms had a PS of 1, and approximately 87% had liver metastases. Only 38% of patients had
tumors of the pancreatic head, and only 14% had a biliary
stent.
The median number of 2-week treatment cycles was 10 in
the FOLFIRINOX arm and 6 in the gemcitabine arm (p
0.001). The median relative dose intensity was approximately 80% for each of the component drugs in the FOLFIRINOX regimen and 100% for gemcitabine.
Independent radiologic review confirmed that 15 of the
first 44 FOLFIRINOX-treated patients (34%) in the phase II
portion of the trial had an objective response, meeting the
criteria for the study to proceed to phase III. Patients
treated with FOLFIRINOX achieved a much higher objective response rate (31.6%) than those who received singleagent gemcitabine (9.4%).
Median overall survival was significantly longer in the
patients treated with FOLFIRINOX (11.1 months vs. 6.8
months, HR 0.57, 95% CI, 0.45 to 0.73; p 0.001). Overall
survival rates at 6, 12, and 18 months were also superior for
FOLFIRINOX-treated patients (76%, 48%, and 19% respectively), compared with 58%, 21%, and 6%, respectively for
those who received gemcitabine. Patients on the multidrug
regimen also achieved a superior progression-free survival
(6.4 months vs. 3.3 months, HR 0.47; 95% CI, 0.37 to
0.59; p 0.001). The beneficial effect of FOLFIRINOX was
similar in all patient subgroups. These data are summarized
in Table 1.
233
HEDY LEE KINDLER

Table 1. FOLFIRINOX versus Gemcitabine: Efficacy
FOLFIRINOX
Gemcitabine
171
0.6%
31%
39%
71%
11.1
48%
19%
6.4
171
0%
9%
42%
51%
6.8
21%
6%
3.3
Patients
Complete response
Partial response (PR)
Stable disease (SD)
Disease control (PR SD)
Median overall survival (months)
1-yr survival
18-mo survival
Progression-free survival (months)
Hazard
Ratio
18
p value
0.0001
0.57
0.0003
0.0001
0.47
0.0001
Patients who received FOLFIRINOX experienced significantly higher rates of grade 3 and 4 neutropenia (46% vs.
21%), febrile neutropenia (5% vs. 1%), thrombocytopenia
(9% vs. 4%), diarrhea (13% vs. 2%), and sensory neuropathy
(9% vs. 0%) than those who received gemcitabine. The
presence of a biliary stent did not increase the risk of
infection in either arm, and no cholangitis was reported.
Filgrastim was given to 43% of FOLFIRINOX-treated patients. Toxicity data are summarized in Table 2.
Significantly more patients on the gemcitabine arm had a
definitive decrease in their scores on the Global Health
Status and Quality of Life scale compared with those on the
FOLFIRINOX arm (66% vs. 31%, HR 0.47, p 0.001). A
significant increase in the time until definitive deterioration
in quality of life was observed in the FOLFIRINOX-treated
subjects for all functional and symptom scales.
FOLFIRINOX Usage in Clinical Practice
Almost immediately after Conroy presented the FOLFIRINOX data at ASCO in 2010, Xcenda, LLC analyzed the
prescribing plans of American oncologists using the NMCR
Challenging Cases live research vehicle.24 From July 31 to
August 28, 2010, they assessed the prescribing plans of more
than 370 U.S. oncologists for first-line therapy of patients
with metastatic pancreatic cancer and PS 1 or 2. They
observed that the FOLFIRINOX data produced an immediate change in the distribution of planned first-line prescribing. For the PS 1 scenario, FOLFIRINOX had an 18% share;
in the previous year, those patients would mostly have
received gemcitabine/erlotinib. As expected, plans for FOLFIRINOX were negligible (3%) for PS 2 patients.
In the phase I, II, and III trials described above, FOLFIRITable 2. FOLFIRINOX versus Gemcitabine: Selected Grade 3
and 4 Toxicities 18
Toxicity
Hematologic
Neutropenia
Febrile neutropenia
Thrombocytopenia
Anemia
Non-hematologic
Fatigue
Vomiting
Diarrhea
Sensory neuropathy
Increased alanine aminotransferase
Thrombosis
Abbreviation: NS, not significant.
234
FOLFIRINOX
(171 patients)
Gemcitabine
(171 patients)
p value
46%
5%
9%
8%
21%
1%
4%
6%
0.001
0.03
0.04
NS
24%
15%
13%
9%
7%
7%
18%
8%
2%
0%
21%
4%
NS
NS
0.001
0.001
0.001
NS
NOX was tested in European patients who had predominantly body/tail tumors and a good performance status,
which may not be fully representative of other populations of
patients with pancreatic cancer. In the absence of data from
additional prospective clinical trials, retrospective singleinstitution series from highly selected academic centers
provide some insight into the current state of FOLFIRINOX
usage in the United States and elsewhere.25-30 These data
confirm the considerable activity of this regimen in the
metastatic, locally advanced, and previously-treated settings, demonstrate the safety of this combination in patients
with indwelling biliary stents, and elucidate how some
physicians are routinely modifying drug doses without clear
evidence or guidelines.
Between July 2010 and April 2011, investigators at Massachusetts General Hospital treated 29 patients with pancreatic cancer with FOLFIRINOX; 59% had metastatic
disease.25 The median age was 60. Most (62%) were
chemotherapy-nave, 17% and 21%, respectively, had received one and two prior regimens. Almost half (48%) had a
PS of 1; 55% had head or uncinate tumors, and 28% had
biliary stents. A median of eight cycles were delivered. The
objective response rate on formal radiologic review was 38%.
In chemotherapy-nave patients, the response rate was 56%;
stable disease was reported in 39%, for a disease control rate
of 94%. In previously-treated patients, the response rate was
9%; the disease control rate was 73%. Response rates were
similar in patients with locally advanced and metastatic
disease (42% and 35%, respectively). Emergency room visits
or hospitalizations were required in 41% of patients, most
commonly for neutropenic fever or dehydration (14% each).
Seven of the 10 patients who developed grades 3 or 4
neutropenia had not received prophylactic growth factors
from the start of FOLFIRINOX treatment.
Washington University physicians used a registry to document the tolerance and efficacy of FOLFIRINOX.26
Twenty-nine patients, with a median age of 57, received at
least one cycle of FOLFIRINOX. Most (93%) had an ECOG
PS of 0 or 1, 38% had pancreatic head tumors, and 24% had
biliary stents. The majority (52%) had metastatic disease;
the rest were locally advanced. The regimen was empirically
modified in the majority of patients because of concern for
potential toxicities. The 5-FU bolus was deleted in 48%.
Irinotecan was dose-reduced in four patients who had the
UGT1A1*28/*28 genotype, and in nine patients who did not.
Prophylactic growth factor support was administered to 48%
of patients beginning with the first cycle; 10% initiated
granulocyte-colony stimulating factor (G-CSF) in subsequent cycles. Grades 3 and 4 neutropenia developed in 14%,
but only one patient experienced a neutropenic fever. Fourteen percent of the patients were hospitalized. Of the patients who began treatment on the full-dose regimen, only
25% required any dose adjustment in future cycles, suggesting to these investigators that prophylactic dose adjustments may not have been necessary. On independent
review, 26% achieved partial responses and 63% had stable
disease. Thus, despite frequent dose reductions, the majority
of patients achieved disease control with FOLFIRINOX.
Investigators at Yale University observed that oncologists
in community and academic practices were reluctant to use
full-dose FOLFIRINOX because of its toxicity profile.27 To
ascertain the potential effect of dose attenuation on toxicity
and efficacy, they performed a retrospective review of pa-
tients with pancreatic cancer who were treated with FOLFIRINOX at their institution between June 2010 and June
2011, and compared patient characteristics, toxicities, and
response rates with those reported in the pivotal phase III
trial.18 Thirty-five patients, with a median age of 61, were
treated with FOLFIRINOX; 68% had an ECOG PS of 0, 57%
had pancreatic head tumors, 46% had locally advanced
disease, and only 14% had received prior chemotherapy.
Only 17% of patients received full-dose FOLFIRINOX
with the first cycle. Irinotecan was reduced in 93% of
patients and omitted in 3%; oxaliplatin was reduced in 34%;
bolus 5-FU was reduced in 31% and omitted in 24%; leucovorin was decreased in 37%; and the 5-FU continuous
infusion was decreased in 10%. A median of 10 cycles was
delivered. The median relative doses of oxaliplatin, irinotecan, bolus 5-FU, and infusional 5-FU were 90%, 68%, 68%,
and 100%, respectively (in the phase III trial the median
relative doses of oxaliplatin, irinotecan, and 5-FU [bolus and
infusion] were 78%, 81%, and 82%, respectively). Yale patients experienced significantly less grade 3/4 fatigue (p
0.0089) and neutropenia (p 0.0001) compared with patients in the phase III trial. Despite routine dose modifications, the response rate, progression-free survival, and
overall survival were not significantly different from historic
controls. The authors concluded that modest dose attenuations of FOLFIRINOX reduce toxicity but do not appear to
compromise its efficacy.
The activity of FOLFIRINOX in previously treated patients has been described in two retrospective series from
France.28,29 In a retrospective review of 27 patients who
received second-line FOLFIRINOX from 2003 to 2009, a
median of six cycles were delivered. Grades 3 4 neutropenia
developed in 56%, and one patient experienced grade 5
febrile neutropenia; 44% received growth factors as secondary prophylaxis. Partial responses were achieved in 19%;
44% had stable disease. The median time to progression was
5.4 months, and median overall survival was 8.5 months.28
In a second French series, 13 previously-treated patients,
only 69% of whom had a PS of 0 or 1, were treated with
FOLFIRINOX, mostly (85%) in the second-line setting.
There were no objective responses, but 46% had stable
disease. No grade 4 toxicities were reported.29
FOLFIRINOX Is Cost-effective
Using a Markov model, Attard and colleagues assessed

the cost-effectiveness of FOLFIRINOX compared with gemcitabine in Canadian patients undergoing first-line treatment for metastatic pancreatic cancer.31 Since the initial
treatment choice affects subsequent therapy, second-line
treatment was also included in their analysis.
Their first model, based on the ACCORD 11 trial data,
assumed that in one arm patients received first-line FOLFIRINOX and second-line gemcitabine, and in the other
arm, they received first-line gemcitabine and second-line
platinum-based chemotherapy; in both groups G-CSF usage
was allowed.
The second analysis, which mirrored current Ontario
treatment patterns, assumed that first-line FOLFIRINOX
was followed by second-line gemcitabine in one arm, and
first-line gemcitabine was followed by best supportive care
in the other arm; no G-CSF was permitted.
In both scenarios, first-line FOLFIRINOX produced more

life years and quality-adjusted life years (QALY) than firstline gemcitabine. The costs per QALYs for FOLFIRINOX
were about $45,000 to $55,000. Thus, even though the
component drugs of the FOLFIRINOX regimen cost more
than single-agent gemcitabine, FOLFIRINOX is more costeffective than gemcitabine. FOLFIRINOX has therefore
received a favorable funding recommendation in most Canadian provinces.
Future Directions with FOLFIRINOX
Given the impressive activity of FOLFIRINOX in the

metastatic setting, plans are underway to prospectively
evaluate this regimen in the adjuvant, locally advanced, and
borderline resectable settings, studies are ongoing to add
targeted agents to FOLFIRINOX, and trials are in development to determine how to adjust doses and ameliorate
toxicity.
PRODIGE 24-ACCORD 24/0610 will be the first trial to
evaluate FOLFIRINOX in the adjuvant setting. This study
will use a modified regimen, called mFOLFIRINOX, in
which the bolus 5-FU has been omitted and all other drug
doses remain unchanged. Eligible patients with resected
head, body, or tail lesions, PS 0 1, age less than 80 years,
total bilirubin less than1.5 X ULN, and Ca 19 9 less than
180, will be stratified by center, node status, postoperative
Ca 19 9 level (90 vs. 91180), and surgical margin (R0 vs.
R1), and randomized to 24 weeks of gemcitabine or mFOLFIRINOX. The primary endpoint will be 3-year disease-free
survival. A 30-patient lead-in safety analysis will soon be
initiated to ascertain that the rate of grade 3 4 diarrhea is
less than 5%. The study requires 490 patients to demonstrate a 10% increase in disease-free survival at 3 years,
from 17% to 27%. It will be conducted in France and Canada.
The safety of FOLFIRINOX in borderline resectable patients has been described in a retrospective series.30
A021101 is a 50-patient, single-arm, neoadjuvant phase II
U.S. Intergroup study of mFOLFIRINOX followed by chemoradiation then surgery and postoperative gemcitabine for
patients with borderline resectable pancreatic cancer. This
benchmark trial will assess the feasibility of a multiinstitutional effort in this patient subgroup and provide a
foundation for future trials. The primary endpoint is 1-year
overall survival.
FOLFIRINOX may also serve as a platform for the addition of targeted agents, though caution must be used because of the potential for overlapping toxicities. The Cancer
and Leukemia Group B (CALGB) will be leading a phase
IB/randomized phase II trial of mFOLFIRINOX plus placebo
or ganitumumab (a monoclonal antibody to the insulin
growth factor 1 receptor), in patients with previously untreated metastatic pancreatic cancer. This will be the first
U.S. cooperative group trial to confirm the European experience with FOLFIRINOX.
A phase IB dose-finding study of FOLFIRINOX plus
IPI-926, a hedgehog pathway inhibitor, is currently ongoing.
Once a phase II dose is determined, this combination will be
incorporated in a randomized phase II study in development
in CALGB and ECOG, in which patients with locally advanced diseaase are randomized to FOLFIRINOX plus IPI926 or placebo, followed by chemoradiation. A phase IB
study of FOLFIRINOX plus the hedgehog inhibitor LDE225
is also accruing patients.
235
HEDY LEE KINDLER
The Southwest Oncology Group (SWOG) is in the process

of designing a randomized trial that compares FOLFIRINOX with FOLFOX, to determine whether irinotecan is an
essential component of the regimen. Another approach is to
ascertain which patients would be most likely to develop
toxicity from irinotecan and adjust doses accordingly.
UGT1A1 is the enzyme responsible for clearing SN-38, the
active metabolite of irinotecan; germline polymorphisms in
the UGT1A1 gene are known to reduce enzymatic activity.32
Thus, patients treated with FOLFIRINOX who have different UGT1A1 genotypes may tolerate different doses of
irinotecan. Using genotype-guided dosing, investigators at
the University of Chicago will soon open a phase I study to
establish the optimal safe doses of irinotecan in the mFOLFIRINOX regimen for each of three UGT1A1 genotype
groups (*1*1, *1*28, and *28*28).
Conclusion
After so many negative randomized trials of gemcitabine

doublets, the unprecedented outcomes achieved with the
FOLFIRINOX regimen clearly represent a major treatment
advance for those patients with pancreatic cancer who have
a good performance status. No other randomized study has
ever achieved a median survival of nearly a year. In no other
trial have we even whispered about an 18-month survival in

a proportion of patients with metastatic disease. No other
phase III study has achieved such a high objective response
rate. And despite substantial, though manageable toxicities,
FOLFIRINOX also helps patients feel better for longer than
if they received gemcitabine, a drug used principally for its
effect on symptoms. Remarkably, this new drug combination
is even cost-effective.
The investigators in this study are to be commended not
only for the decade they spent in developing the highly
active FOLFIRINOX regimen. They are also to be applauded
for a very well-designed pivotal study, which tested this
therapy in a uniform population of patients (all metastatic),
who, with their good performance status, were most likely to
tolerate the rigors of the multidrug combination and were
thus most able to benefit from it.
Unanswered questions remain about the optimal way to
dose the component drugs to minimize toxicity while preserving activity. Upcoming studies will address the potential
role of this regimen in other disease settings and will use
FOLFIRINOX as a platform on which to add new agents.
It has been a long journey, but with the advent of FOLFIRINOX, we are finally beginning to make progress against
this dismal disease.
Author
Hedy Lee Kindler
Employment or
Leadership
Positions
Consultant or
Advisory Role
AstraZeneca;
Bristol-Myers
Squibb; Clovis
Oncology;
Genentech;
GlaxoSmithKline;
Merck
Stock
Ownership
Honoraria
Research
Funding
CanBas;
Genentech;
Infinity; Lilly;
Merck;
Morphotek
Expert
Testimony
Other
Remuneration
REFERENCES
1. Burris HA 3rd, Moore MJ, Andersen J, et al. Improvements in survival
and clinical benefit with gemcitabine as first-line therapy for patients with
advanced pancreas cancer: A randomized trial. J Clin Oncol. 1997;15:24032413.
2. Berlin JD, Catalano P, Thomas JP, et al. Phase III study of gemcitabine
in combination with fluorouracil versus gemcitabine alone in patients with
advanced pancreatic carcinoma: Eastern Cooperative Oncology Group Trial
E2297. J Clin Oncol. 2002;20:3270-3275.
3. Rocha Lima CM, Green MR, Rotche R, et al. Irinotecan plus gemcitabine
results in no survival advantage compared with gemcitabine monotherapy in
patients with locally advanced or metastatic pancreatic cancer despite increased tumor response rate. J Clin Oncol 2004;22:3776-3783.
4. Louvet C, Labianca R, Hammel P, et al. Gemcitabine in combination
with oxaliplatin compared with gemcitabine alone in locally advanced or
metastatic pancreatic cancer: Results of a GERCOR and GISCAD phase III
5. Oettle H, Richards D, Ramanathan RK, et al. A phase III trial of
pemetrexed plus gemcitabine versus gemcitabine in patients with unresectable or metastatic pancreatic cancer. Ann Oncol. 2005;16:1639-1645.
6. Heinemann V, Quietzsch D, Gieseler F, et al. Randomized phase III trial
of gemcitabine plus cisplatin compared with gemcitabine alone in advanced
pancreatic cancer. J Clin Oncol. 2006;24:3946-3952.
7. Herrmann R, Bodoky G, Ruhstaller T, et al. Gemcitabine plus capecitabine compared with gemcitabine alone in advanced pancreatic cancer: A
randomized, multicenter, phase III trial of the Swiss Group for Clinical
Cancer Res and the Central European Cooperative Oncology Group. J Clin
Oncol. 2007;25:2212-2217.
8. Moore MJ, Goldstein D, Hamm J, et al. Erlotinib plus gemcitabine
compared with gemcitabine alone in patients with advanced pancreatic
236
cancer: a phase III trial of the National Cancer Institute of Canada Clinical
Trials Group. J Clin Oncol. 2007;25:1960-1966.
9. Grubbs SS, Grusenmeyer PA, Petrelli NJ, Gralla RJ. Is it cost-effective
to add erlotinib to gemcitabine in advanced pancreatic cancer? J Clin Oncol.
2006; 24:18s(suppl; abstr 6048).
10. Miksad RA, Schnipper L, Goldstein M. Does a statistically significant
survival benefit of erlotinib plus gemcitabine for advanced pancreatic cancer
translate into clinical significance and value? J Clin Oncol. 2007;25:45064507.
11. Kindler HL, Niedzwiecki D, Hollis D, et al. Gemcitabine plus bevacizumab compared with gemcitabine plus placebo in patients with advanced
pancreatic cancer: Phase III trial of the Cancer and Leukemia Group B
(CALGB 80303). J Clin Oncol. 2010;28:3617-3622.
12. Kindler HL, Ioka T, Richel DJ, et al. Axitinib plus gemcitabine versus
placebo plus gemcitabine in patients with advanced pancreatic adenocarcinoma: A double-blind randomized phase 3 study. Lancet Oncology. 2011;12:
256-262.
13. Poplin E, Feng Y, Berlin J, et al. Phase III, randomized study of
gemcitabine and oxaliplatin versus gemcitabine (fixed-dose rate infusion)
compared with gemcitabine (30-minute infusion) in patients with pancreatic
carcinoma E6201: a trial of the Eastern Cooperative Oncology Group. J Clin
Oncol. 2009;27:3778-3785.
14. Philip PA, Benedetti J, Corless CL, et al. Phase III study comparing
gemcitabine plus cetuximab versus gemcitabine in patients with advanced
pancreatic adenocarcinoma: Southwest Oncology Group-directed intergroup
trial S0205. J Clin Oncol. 2010;28:3605-3610.
15. Cunningham D, Chau I, Stocken DD, et al. Phase III randomized
comparison of gemcitabine versus gemcitabine plus capecitabine in patients
with advanced pancreatic cancer. J Clin Oncol. 2009;27:5513-5518.
16. Van Cutsem E, Vervenne WL, Bennouna J, et al. Phase III trial of

bevacizumab in combination with gemcitabine and erlotinib in patients with
metastatic pancreatic cancer. J Clin Oncol. 2009;27:2231-2237.
17. Conroy T, Desseigne F, Ychou M, et al. Randomised phase III trial
comparing FOLFIRINOX (F : 5FU/leucovorin [LV], irinotecan [I], oxaliplatin
[O]) versus gemcitabine (G) as first-line treatment for metastatic pancreatic
adenocarcinoma : Preplanned interim analysis results of the PRODIGE
4/ACCORD 11 trial. J Clin Oncol. 2010;28: 15S (suppl; abstr 4010).
18. Conroy T, Desseigne F, Ychou M, et al. FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Med. 2011;364:1817-1825.
19. Ychou M, Conroy T, Seitz JF, et al. An open phase I study assessing
the feasibility of the triple combination: oxaliplatin plus irinotecan plus
leucovorin/5-fluorouracil every 2 weeks in patients with advanced cancers.
Ann Oncol. 2003;14:481-489.
20. Conroy T, Paillot B, Francois E, et al. Irinotecan plus oxaliplatin and
leucovorin-modulated fluorouracil in advanced pancreatic cacnerA Groupe
Tumeurs Digestives of the Federation Nationale des Centres de Lutte Contre
le Cancer Study. J Clin Oncol. 2005;23:1228-1236.
21. Miller AB, Hoogstraten B, Staquet M, et al. Reporting results of cancer
treatment. Cancer. 1981;47:201-214.
22. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to
evaluate the response to treatment in solid tumors. European Organization
for Research and Treatment of Cancer, National Cancer Institute of the
United States, National Cancer Institute of Canada. J Natl Cancer Inst.
2000;92:205-216.
23. Conroy T, Gavoille C, Adenis A. Metastatic pancreatic cancer: Old
drugs, new paradigms. Curr Opin Oncol. 2011;23:390-395.
24. Bendell JC, Britton S, Green MR, et al. Immediate impact of the
FOLFIRINOX phase III data reported at the 2010 ASCO Annual Meeting on
prescribing plans of American oncology physicians for patients with metastatic pancreas cancer. J Clin Oncol. 2011;29 (suppl, abstract 286).
25. Faris JE, Hong TS, McDermott S, et al. FOLFIRINOX in locally
advanced or metastatic pancreatic cancer. J Clin Oncol. 2012;30 (suppl; abstr
313).
26. Peddi PF, Tan BR, Picus J, et al. Washington University experience
with FOLFIRINOX in pancreatic cancer. J Clin Oncol. 2012;30 (suppl; abstr
354).
27. Gunturu KS, Thumar JR, Hochster HS, et al. Single-institution experience with FOLFIRINOX in advanced pancreatic cancer. J Clin Oncol.
2012;30 (suppl; abstr 330).
28. Assaf E, Verlindhe-Carvalho M, Delbaldo C, et al. 5-Fluorouracil/
leucovorin combined with irinotecan and oxaliplatin (FOLFIRINOX) as
second-line chemotherapy in patients with metastatic pancreatic adenocarcinoma. Oncology. 2011;80:301-306.
29. Breysacher G, Kaatz O, Lemarignier C, et al. Safety and clinical
effectiveness of FOLFIRINOX in metastatic pancreas cancer after first-line
therapy. J Clin Oncol. 2010;28 (suppl; abstr 269).
30. Hosein PJ, Kawamura C, Macintyre J, et al. Pilot study of neoadjuvant
FOLFIRINOX in unresectable locally advanced pancreatic carcinoma. J Clin
Oncol. 2011;29 (suppl; abstr 324).
31. Attard CL, Brown S, Alloul K, et al. Cost-effectiveness of FOLFIRINOX
for first-line treatment of metastatic pancreatic cancer. J Clin Oncol. 2012;30
(suppl; abstr 199).
32. Fujita K, Sparreboom. Pharmacogenetics of irinotecan disposition and
toxicity: A review. Curr Clin Pharmacol. 2010;5:209-219.
237
A Matter of Timing: Is There a Role for

Radiation in Locally Advanced Pancreatic
Cancer, and If So, When?
By Theodore S. Hong, MD, Jennifer Y. Wo, MD, and Eunice L. Kwak, MD, PhD
Overview: The role of radiation therapy in the management of

locally advanced pancreatic cancer is controversial. Despite
its localized presentation, locally advanced pancreatic cancer
is characterized by high rates of metastases. Historic data
have been mixed, and newer studies have called into question
the use of radiation therapy. However, it appears that patients
more likely to benefit from chemoradiation can be identified
ANCREATIC CANCER is the fourth leading cause of

cancer deaths in the United States, with 5-year survival
rates of less than 5%.1 Although surgical resection offers
the best chance at long-term survival, only 10% to 20% of
patients have resectable disease. Nearly half of all patients
with pancreatic cancer have clinically evident metastatic
disease, and the remaining 40% of patients present with
localized but unresectable disease because of vascular involvement, which is considered locally advanced pancreatic
cancer (LAPC). However, even patients with apparently
localized disease have high rates of occult metastases, with
staging laparoscopy studies demonstrating that at least
30% of patients harbor metastatic peritoneal disease that is
undetectable by imaging.2
With the greater appreciation of the systemic nature of
localized pancreatic cancer, there has been a renewed interest in evaluating the optimal timing of radiation therapy,
including the early use of systemic chemotherapy both to
address metastatic disease and to select patients most likely
to benefit from local therapy.
Historic Perspective
Although the trend toward upfront or neoadjuvant chemotherapy has gained traction in recent years, the benefit of
chemoradiation therapy has always been controversial.
Early randomized trials suggested mixed benefits from
chemoradiation. Chemoradiation first came to the forefront
of therapy for LAPC based on a randomized trial conducted
by the Gastrointestinal Tumor Study Group (GITSG) in
1974.3 In this study, 228 patients with LAPC were randomly
assigned to receive 60 Gy of radiation therapy alone compared with 40 Gy of radiation with concurrent 5-fluorouracil
(5-FU) and 60 Gy of radiation with 5-FU. Patients assigned
to the two chemoradiation arms also received maintenance
5-FU. Both chemoradiation arms were associated with significantly improved survival compared with the radiation
alone arm (p 0.05 for 40 Gy arm, p 0.001 for 60 Gy arm).
No significant difference was reported between the 60 Gy
and 40 Gy chemoradiation arms.
In 1977, the Eastern Cooperative Oncology Group (ECOG)
initiated a randomized trial comparing 5-FU alone to chemoradiation therapy. Ninety-one patients with LAPC were
randomly assigned to groups receiving either weekly 5-FU
alone or 40 Gy of radiation with bolus 5-FU followed by
weekly 5-FU.4 This study showed no difference in survival
with or without radiation therapy (median survival [MS]
8.3 months vs. 8.2 months, not significant [NS]). However,
this study was criticized because of the poor survival in both
238
with an induction phase of chemotherapy. Data evaluating this

approach suggest that approximately 30% of patients will
develop metastatic disease within the first 3 to 4 months of
chemotherapy. Patients without progression who receive
chemoradiation therapy may experience improved survival.
Future directions include the validation of this strategy and
the integration of biologic agents.
groups. Because of these results, GITSG initiated another

randomized trial comparing a three-drug regimen of streptozocin, mitomycin C, and 5-FU (SMF) or radiation therapy
to 54 Gy with concurrent 5-FU followed by SMF chemotherapy.5 In this small 43-patient study, the 1-year survival was
improved with the inclusion of chemoradiation compared to
chemotherapy alone (41% vs. 19%, p 0.02). Based on this
study, chemoradiation therapy was felt to be superior to
chemotherapy alone.
Role of Radiation Questioned: the 2000 01
FFCD/SFRO Trial
Because of the limitations of the data defining the role of

radiation therapy and the lack of formal evaluation of
radiation therapy in the modern treatment era, the Federation Francophone de Cancerologie Digestive (FFCD) and
the Societe Francophone de Radiotherapie Oncologique
(SFRO) evaluated the role of chemoradiation therapy in
patients with LAPC.6 In this trial, patients were randomly
assigned to receive either chemoradiation followed by maintenance gemcitabine or gemcitabine alone until progression.
Radiation was delivered to 60 Gy in 2 Gy/fraction with an
infusion of 5-FU (300 mg/m2/day) on days 1 to 5 for 6 weeks
and cisplatin (20 mg/m2/day) on days 1 to 5 during weeks 1
and 5. This phase III study was designed to demonstrate
an improvement from 6 months with gemcitabine alone to
12 months with chemoradiation. However, the study was
stopped early because of slow accrual, and a review by an
independent data monitoring committee followed. At that
time, it was noted that the chemoradiation arm had a
shorter overall survival (OS). Survival analysis showed
worse median survival with chemoradiation (8.6 months)
compared with gemcitabine alone (13.0 months) by intentto-treat analysis (p 0.03). Because of concerns of a possibly
confounding toxicity, a per-protocol analysis was performed
restricting analysis to those who received at least 75% of
protocol treatment. By this analysis, patients receiving
chemoradiation therapy still experienced a shorter median
survival (9.5 months vs. 15.0 months, p 0.006).
From the Departments of Radiation Oncology and Medicine, Massachusetts General

Hospital and Harvard Medical School, Boston, MA.
Address reprint requests to Theodore S. Hong, MD, Department of Radiation Oncology,
Massachusetts General Hospital, 100 Blossom Street, Cox 3, Boston, MA 02114; email:
tshong1@partners.org.
1092-9118/10/1-10
RADIATION AND PANCREATIC CANCER
Although these results suggest that chemoradiation

should not be used, the study design limits the ability to
interpret the outcome. The choices of chemotherapy agents
and radiation dose were nonstandard and may have compromised tolerability in the chemoradiation arm. For instance,
although highly emetogenic, cisplatin has not demonstrated
significant activity in pancreatic cancer and the dose of
radiation used was higher than tested in most clinical trials.
Furthermore, this nonstandard regimen was not tested in
phase II trials before its use in this randomized trial. The
toxicity of this regimen led to fewer than half of the patients
receiving 75% of the chemoradiation dose and contributed to
the poor tolerance of subsequent maintenance chemotherapy. The median total dose of gemcitabine received by
patients in the chemoradiation arm was well below half the
total dose received by patients in the chemotherapy-alone
group.
Despite the flaws, this trial underscores a key lesson: even
when localized in presentation, pancreatic cancer is a
largely systemic disease, and it is important for patients to
receive chemotherapy early in the treatment course.
Does Anyone Benefit from Radiation Therapy?
The GERCOR Study
In the late 1990s through the early 2000s, the Groupe

Cooperateur Multidisciplinaire en Oncologie (GERCOR) ran
a number of phase II and III trials evaluating different
chemotherapy regimens.7 Following the trend of including
patients with locally advanced disease in chemotherapy
trials but acknowledging the controversial role of radiation
therapy, investigators recommended patients for these trials
who had no progression after 3 months of receiving chemotherapy and who had ECOG performance status of 2 or
better. However, investigators could also continue chemotherapy at their discretion. Radiation therapy was delivered
to 45 Gy with a boost to a total dose of 55 Gy with continuous
infusion 5-FU. Computed tomography (CT) planning was
mandated.
In a retrospective analysis of prospective trials, 167 of the
497 patients enrolled on these trials had locally advanced
disease. After 3 months of chemotherapy, 71% of patients
with LAPC did not have progressive disease and had performance status scores eligible for chemoradiation. Of the
remaining 29% of patients ineligible for chemoradiation, 45
of the 53 patients were ineligible because they had progres-
KEY POINTS
Most patients with locally advanced pancreatic cancer will succumb to metastatic disease.
A subset of patients with more locally aggressive
disease may be identified by the use of upfront
chemotherapy.
The use of radiation therapy may be beneficial in
patients with nonprogressive disease.
DPC4 may be a biomarker that can predict biologic
behavior.
Future directions include improving the evaluation of
new radiation techniques and integrating biologic
therapies.
sive disease. Of the 128 patients eligible for chemoradiation,

72 (56%) patients received chemoradiation and 56 (44%)
patients continued receiving chemotherapy alone. Patients
in both groups were well matched by chemotherapy regimen, age, performance status, weight loss, and response to
chemotherapy. The median progression-free survival (PFS)
for the chemoradiation and chemotherapy groups were 10.8
months and 7.4 months, respectively (p 0.005). The
median OS was 15 months for the chemoradiation group and
11.7 months for the chemotherapy group (p 0.0009).
Because of its retrospective nature, this study was limited
by potential patient selection bias; however, the two treatment groups were well balanced for performance status, age,
and chemotherapy response. This study suggests that some
patients might benefit from chemoradiation therapy. Although previous studies have recognized that some patients
have rapidly metastatic disease, this study suggests that an
effective strategy to evaluate which patients have aggressive
systemic biology is to start with systemic chemotherapy and
to move only those patients whose disease does not progress
after several months of chemotherapy to receive chemoradiation. This study caused a substantial paradigm shift in
the way radiation therapy was timed. Increasingly, institutions would initiate radiation therapy with chemotherapy,
primarily to identify patients with rapid, early metastatic
progression who would not benefit from aggressive local
therapy.
Further Evidence of the Benefit of Radiation Therapy:
ECOG 4201
During this period, investigators in the United States

were also investigating the role of radiation therapy. Using
a gemcitabine platform, ECOG 4201 directly compared
gemcitabine alone with gemcitabine-based chemoradiation
therapy followed by gemcitabine.8 The chemotherapy-alone
group received one 7-week induction cycle, followed by five
additional 4-week cycles (3 weeks on, 1 week off). The
chemoradiation arm received 50.4 Gy of radiation as 1.8
Gy/fraction with concurrent gemcitabine (600 mg/m2)
weekly during radiation treatment followed by five additional cycles of gemcitabine. The intended study design
included 316 patients but was closed after enrolling only 74
patients because of slow accrual.
As expected, the radiation arm had greater toxicity, with
41% of patients experiencing grade 4 toxicity compared to
9% in the chemotherapy-alone arm. Despite this, both arms
received a median of three cycles of chemotherapy. Median
survival was improved in the radiation arm (11.1 months vs.
9.2 months, p 0.017). There was no difference in PFS.
This trial was more straightforward than the FFCD trial
because it directly compared radiation therapy with gemcitabine to gemcitabine alone. The chemoradiation regimen
demonstrated safety in phase I testing, which in contrast to
the FFCD trial, is reflected in the fact that both arms
received a similar amount of chemotherapy. However, the
small number of patients enrolled in this study limited the
ability to understand the role of radiation therapy. Additionally, because radiation therapy was used at the beginning
of treatment, issues of timing were not addressed. Finally,
this small study elicited an OS benefit, despite a lack of
difference in PFS. It is unclear why this would be the case,
especially in light of the increased toxicity of the radiation
arm. However, this study was the first study to demonstrate
239
HONG, WO, AND KWAK
an OS benefit with radiation therapy against a backbone of

gemcitabine.
Further Evidence of Delaying Radiation Therapy
Investigators at the University of California, San Francisco (UCSF), seeking to optimize systemic control, performed a phase II study of fixed-dose rate gemcitabine in
combination with cisplatin for six cycles, followed by chemoradiation therapy.9 The gemcitabine was given at 1,000
mg/m2 infused at 10 mg/m2/minute followed by cisplatin (20
mg/m2) administered on days 1 and 15 of a 28-day cycle.
Patients initiated chemoradiation therapy 2 to 6 weeks after
completing six cycles of chemotherapy. Radiation was delivered to a standard 50.4 Gy with continuous infusion 5-FU.
Of the 25 patients enrolled on the study, eight (28%)
developed disease progression while receiving chemotherapy
between two and 4.5 treatment cycles (median three cycles),
consistent with the rate of metastatic progression seen in
the GERCOR study. Thirteen patients (56%) proceeded to
the chemoradiation phase. The median OS for the entire
cohort was 13.5 months, with a median time to progression
(TTP) of 10.5 months.
The patterns of treatment failure were informative for the
selection that occurs with upfront chemotherapy. Seven of
the eight patients who had progressive disease during chemotherapy had metastatic disease. In contrast, six of the 12
patients who received all six cycles of chemotherapy and
chemoradiation developed local progression, rather than
metastatic progression. This study suggests that delayed
radiation therapy enriches the group of patients receiving
radiation for those with more localized biology.
Investigators at Massachusetts General Hospital (MGH)
also compared upfront chemoradiation therapy with delayed
chemoradiation therapy. At MGH, patients with LAPC were
historically treated with early chemoradiation therapy.
With the publication of the GERCOR study in 2007, a
gradual shift to delayed chemoradiation therapy occurred.
In a retrospective analysis, investigators compared patients
who received chemotherapy before chemoradiation therapy
with those who started with chemoradiation therapy to
determine the relative outcomes associated with patient
selection based on the timing of chemoradiation treatment.10
In this study, 70 consecutive patients with unresectable
(46 patients) or borderline resectable (24 patients) LAPC
were treated with chemoradiation from 2005 to 2009. Patients typically received 50.4 Gy of radiation in 28 fractions
(91%) with concurrent 5-FU (84%) or capecitabine (14%).
Forty patients received chemoradiation alone, and 30 patients received a median of 4 months of chemotherapy before
chemoradiation, typically gemcitabine (93%). All patients
without progression after chemotherapy were offered
chemoradiation.
Fifty-three percent of the patients in the early chemoradiation group compared with 83% in the delayed chemoradiation group had categorically unresectable tumors at
diagnosis. Median OS for the early and delayed chemoradiation groups was 12.4 months and 18.7 months, respectively
(p 0.02). Median PFS for early compared with delayed
chemoradiation was 6.7 months and 11.4 months, respectively (p 0.02). On multivariate analysis, administration
of chemotherapy before chemoradiation (adjusted hazard
240
ratio [AHR] 0.49; 95% CI, 0.28 to 0.87; p 0.02) and

surgical resection (AHR 0.38; 95% CI, 0.17 to 0.85; p
0.02) were associated with increased OS.
These studies are representative of the paradigm shift
that has occurred at many institutions over the past 5 years.
The use of upfront chemotherapy represents a way to
identify patients more likely to benefit from chemoradiation
treatment.
A Marker for Different Biologies: DPC4 Status
Although clinical evidence has suggested that upfront

chemotherapy could help identify patients more likely to
benefit from chemoradiation, a biomarker predictive of clinical course had yet to be identified. To this end, investigators
at Johns Hopkins University evaluated 76 patients who had
consented to participate in their rapid autopsy program.11
Patients were grouped according to death with widely metastatic disease or death with locally progressive disease.
Grouping was correlated with KRAS and TP53 mutations
and DPC4 (which stands for Deleted in Pancreatic Cancer,
locus 4) status.
Of the 76 patients, 9 had no metastases at the time of
death, 13 had 10 metastases or fewer, 26 had 11 to 99
metastases, and 26 patients had more than 100 metastases.
Investigators found a striking correlation between DPC4
status and patterns of treatment failure. Generally, patients
with DPC4 loss had widespread disease, whereas patients
with intact DPC4 were more likely to have locally destructive disease (p 0.007).
This study highlights two major points. First, almost 30%
of patients died with either no metastases or fewer than 10
metastases, highlighting that a substantial number of patients succumb to locally destructive disease rather than
metastatic disease. Second, intact DPC4 presence as assayed by immunohistochemistry may identify a subgroup of
tumors that are destined to be locally destructive as opposed
to widely metastatic. Because this finding may have implications for the relative roles of chemotherapy and radiation
therapy, the Radiation Therapy Oncology Group (RTOG)
has proposed a study in which systemic chemotherapy with
a combination of 5-FU, oxaliplatin, irinotecan, and leucovorin (FOLFIRINOX) will be studied in patients with loss of
DPC4 and radiation dose escalation will be studied in
patients with intact DPC4.
Current Recommendations and Future Directions
The evidence that LAPC represents a heterogeneous

group has been mounting. Although the majority of patients
will succumb to metastatic disease, there appears to be a
group of patients that has more locally destructive disease.
Currently, the use of chemotherapy before radiation therapy
has the advantages of early treatment of micrometastatic
disease and better selection of patients who may benefit
from chemoradiation therapy. Efforts are currently ongoing
to further clarify the timing of radiation for LAPC. At this
writing, the LAP 07 study is evaluating the role of radiation
therapy in this very manner (Fig. 1). Led by GERCOR, LAP
07 is evaluating chemoradiation therapy after four cycles of
chemotherapy compared with two more cycles of chemotherapy. This trial involving 902 patients will hopefully provide
a definitive answer to the role of delayed chemoradiation.
In the future, the use of biomarkers, such as DPC4, may
RADIATION AND PANCREATIC CANCER
Fig. 1.
The schema of the ongoing LAP 07 randomized trial for locally advanced pancreatic cancer.
provide this biologic information at the time of diagnosis.

However, the DPC4 assay has not yet been standardized for
routine clinical use. The RTOG is currently in the process of
validating the feasibility of this assay on fine needle aspiration samples for its proposed trial of patients with LAPC.
Questions remain regarding the optimal duration of therapy. No standard recommendation exists about how much
chemotherapy should be given before chemoradiation therapy. As discussed above, the GERCOR study used 3 months.
The RTOG plans to use 2 months for the induction phase of
chemotherapy. At our institution, we have chosen to use 4
months, in part because adjuvant trials like RTOG 9704
have administered 4 months of chemotherapy when used
with 5.5 weeks of chemoradiation therapy. Additionally,
because the median TTP on gemcitabine is between 3 and 4
months, 4 months allows adequate time for metastatic
biology to declare itself, as demonstrated in the UCSF trial.
The use of 2 to 4 months of chemotherapy seems reasonable.
In addition, no consensus has been set on which chemotherapy should be used with standard fractionation radiation therapy. Classically, most studies have used 5-FU,
shifting from bolus 5-FU to continuous infusion 5-FU, which
follows the trend in rectal cancer. Gemcitabine-based
chemoradiation therapy has been studied extensively, including in the randomized ECOG study. For example, RTOG
and Alliance cooperative groups are using capecitabine as
the standard chemoradiation backbone. Whether a particular platform has substantial advantages over another remains unclear.
There has been increasing interest in the use of stereotactic body radiation therapy (SBRT), where advanced technology delivers high doses of radiation alone to the tumor. In
2004, an initial phase I study out of Stanford University
demonstrated the feasibility and safety of SBRT for LAPC.
In this study, 15 patients with LAPC and an ECOG performance status of less than 2 were treated with 15 Gy, 20 Gy,
or 25 Gy of radiation in a single fraction via CyberKnife.
One of the three patients treated at 15 Gy, two of the four
patients treated at 20 Gy, and zero of the six patients
treated with 25 Gy had local progression.12 This phase I
study was followed by a phase II study combining conven-
tionally fractionated chemoradiation with a stereotactic radiosurgery (SRS) boost.13 Sixteen patients were treated with
45 Gy of radiation in 1.8 Gy/fraction to the tumor and
regional lymphatics with concurrent 5-FU or capecitabine.
Within 1 month of chemoradiation treatment, patients were
given an SRS boost of 25 Gy of radiation alone to the tumor
using CyberKnife. Sixteen patients were treated. Fifteen of
the16 patients were free from local progression until death.
However, TTP (17.5 weeks) and median survival (33 weeks)
were modest. In a retrospective study from the Beth Israel
Deaconness Medical Center, 47 patients with LAPC were
given two cycles of gemcitabine followed by restaging.14
Patients without metastatic disease were given a third cycle
of gemcitabine while undergoing planning. Patients were
then treated with 24 Gy to 36 Gy of radiation in three
fractions followed by maintenance gemcitabine. Eight patients (17%) developed metastatic disease before undergoing
SBRT. Patients undergoing SBRT had a median OS of 20
months.
Efforts are also ongoing to reduce the toxicity of chemoradiation treatment by using advanced radiation technology
such as intensity modulated radiation therapy (IMRT).
IMRT uses multiple beam angles and a computational
process called inverse planning to create irregular dose
distributions that can conform to irregular shapes, thereby
affording the potential to decrease the toxicity associated
with chemoradiation. In a retrospective review from the
University of Maryland, 46 patients with pancreatic or
ampullary cancers treated with IMRT were evaluated.15
Investigators noted a grade 3 or 4 nausea and vomiting rate
of 0%, compared to the 11% seen with standard CT-based
radiation therapy used in RTOG 9704. This study suggests
that further improving radiation delivery can positively
influence the toxicity profile that was reported in the FFCD
and ECOG studies.
Based on the current research, for the treatment of patients with LAPC we recommend 2 to 4 months of chemotherapy followed chemoradiation therapy. Patients typically
receive gemcitabine, although select patients are now also
receiving FOLFIRINOX. Chemoradiation therapy is delivered with either continuous infusion 5-FU or capecitabine to
241
HONG, WO, AND KWAK
a dose of 50.4 Gy to 59.4 Gy of radiation in 1.8 Gy/fraction.

Future trials are now focused on the prospective evaluation
of biomarkers such as DPC4 and the integration of targeted

therapies into this platform.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Theodore S. Hong*
Jennifer Y. Wo*
Eunice L. Kwak*
REFERENCES
1. Jemal A, Siegel R, Xu J, et al. Cancer statistics, 2010. CA Cancer J Clin.
2010;60:277-300.
2. Jimenez RE, Warshaw AL, Fernandez-del Castillo C. Laparoscopy and
peritoneal cytology in the staging of pancreatic cancer. J Hepatobiliar
Pancreat Surg. 2000;7:15-20.
3. Gastrointestinal Tumor Study Group. A multi-institutional comparative
trial of radiation therapy alone and in combination with 5-fluorouracil for
locally unresectable pancreatic carcinoma. Ann Surg. 1979;189:205-208.
4. Klaassen DJ, MacIntyre JM, Catton GE, et al. Treatment of locally
unresectable cancer of the stomach and pancreas: a randomized comparison
of 5-fluorouracil alone with radiation plus concurrent and maintenance
5-fluorouracil-an Eastern Cooperative Oncology Group study. J Clin Oncol.
1985;3:373-378.
5. Gastrointestinal Tumor Study Group. Treatment of locally unresectable
carcinoma of the pancreas: comparison of combined-modality therapy (chemotherapy plus radiotherapy) to chemotherapy alone. J Natl Cancer Inst.
1988;80:751-755.
6. Chauffert B, Mornex F, Bonnetain F, et al. Phase III trial comparing
intensive induction chemoradiotherapy (60 Gy, infusional 5-FU and intermittent cisplatin) followed by maintenance gemcitabine with gemcitabine alone
for locally advanced unresectable pancreatic cancer. Definitive results of the
2000-01 FFCD/SFRO study. Ann Oncol. 2008;19:1592-1599.
7. Huguet F, Andre T, Hammel P, et al. Impact of chemoradiotherapy after
disease control with chemotherapy in locally advanced pancreatic adenocarcinoma in GERCOR phase II and III studies. J Clin Oncol. 2007;25:326-331.
8. Loehrer PJ Sr, Feng Y, Cardenes H, et al. Gemcitabine alone versus
242
gemcitabine plus radiotherapy in patients with locally advanced pancreatic

cancer: an Eastern Cooperative Group Oncology Trial. J Clin Oncol. 2011;29:
4105-4112.
9. Ko AH, Quivey JM, Venook AP, et al. A phase II study of fixed-dose rate
gemcitabine plus low-dose cisplatin followed by consolidative chemoradiation
for locally advanced pancreatic cancer. Int J Radiat Oncol Biol Phys. 2007;
68:809-816.
10. Arvold ND, Ryan DP, Niemierko A, et al. Long-term outcomes of
neoadjuvant chemotherapy before chemoradiation for locally advanced pancreatic cancer. Cancer. 2011 Oct 21.
11. Iacobuzio-Donahue CA, Fu B, Yachida S, et al. DPC4 gene status of the
primary carcinoma correlates with patterns of failure in patients with
pancreatic cancer. J Clin Oncol. 2009;27:1806-1813.
12. Koong AC, Le QT, Ho A, et al. Phase I study of stereotactic radiosurgery
in patients with locally advanced pancreatic cancer. Int J Radiat Oncol Biol
Phys. 2004;58:1017-21.
13. Koong AC, Christofferson E, Le QT, et al. Phase II study to assess the
efficacy of conventionally fractionated radiotherapy followed by a stereotactic
radiosurgery boost in patients with locally advanced pancreatic cancer. Int J
Radiat Oncol Biol Phys. 2005;63:320-3.
14. Mahadevan A, Miksad R, Goldstein M, et al. Induction gemcitabine and
stereotactic body radiotherapy for locally advanced nonmetastatic pancreas
cancer. Int J Radiat Oncol Biol Phys. 2011;81:e615-22.
15. Yovino S, Poppe M, Jabbour S, et al. Intensity-modulated radiation
therapy significantly improves acute gastrointestinal toxicity in pancreatic
and ampullary cancers. Int J Radiat Oncol Biol Phys. 2011;79:158-62.
A Myriad of Symptoms: New Approaches to

Optimizing Palliative Care of Patients with
Advanced Pancreatic Cancer
By Lauren A. Wiebe, MD
Overview: Patients with advanced pancreatic cancer (APC)

require early and frequent palliative interventions to achieve
optimal quality of life for the duration of illness. Evidencebased supportive treatments exist to maximize quality of life
for any patient, whether receiving chemotherapy or not. This
article provides a comprehensive review of symptoms with
current treatment recommendations and directions for future
development. Celiac plexus neurolysis improves pain in the
majority of patients with APC and should be moved earlier in
the analgesic paradigm. Malignant bowel obstruction can be
palliated quickly with optimal management via gastric decompression, octreotide, parenteral opioids, and standing antiemetics. Recommendations are provided for best treatment of
malignant gastroparesis, gastric outlet obstruction, and
chemotherapy-induced nausea and vomiting in this popula-
ATIENTS WITH advanced pancreatic cancer (APC)

require early and frequent palliative interventions to
achieve optimal quality of life for the duration of illness.
Despite recent notable advances in multidisciplinary antineoplastic therapy, the majority of patients with APC ultimately die after facing numerous physical and emotional
hurdles.1,2 As a physician caring for these patients, there
are opportunities to relieve suffering from predictable complications of pancreatic cancer. Evidence-based supportive
treatments exist to maximize quality of life for any patient,
whether receiving chemotherapy or not. This article provides a comprehensive review of symptoms facing patients
wtih APC with current treatment recommendations and
directions for future development.
Pain
Pain from cancer in the pancreas is often constant and

severe, felt predominantly in the midback and epigastrium.
Noxious sensory input from inflamed pancreatic tissue and
direct neural invasion is transmitted via the celiac plexus as
pain.3 Opioids and adjuvant medications remain standard of
care for analgesia; however, the titration of opioids can be
limited by systemic toxicities and may not adequately address the pain. Successful locoregional intervention minimizes systemic opioid requirements early in the disease.3
In the majority of patients, celiac plexus neurolysis (CPN)
has been shown to provide effective pain relief simultaneous
with reduction in systemic opioids.3,4 An injection of either
ethanol or phenol destroys afferent nerve fibers3,4 and disrupts pain signals for an average of 3 months, though
sometimes permanently. CPN can be performed with equivalent efficacy surgically, percutaneously under radiologic
guidance, or endoscopically via ultrasound.3 The most common risks include transient hypotension, constipation, or
diarrhea; no serious adverse events were noted in a metaanalysis.3 More than 80% of patients note significantly
improved analgesia after CPN in blinded or sham studies.3,4
To evaluate both effect and timing of CPN in pancreatic
cancer, Wyse and colleagues performed a double-blind, randomized controlled trial of patients found to have inoperable
tion. Malignant ascites can be treated initially with diuretics

and sodium-restriction in patients with an exudative process;
however, an indwelling catheter is recommended for patients
with recurrent ascites, particularly because of carcinomatosis
or a refractory process. With exocrine insufficiency contributing to weight loss, pancreatic enzyme replacement is essential
to improve nourishment in the majority of patients. Presently,
megestrol acetate is the only U.S. Food and Drug Administration (FDA)-approved therapy for the anorexia-cachexia syndrome, although future developments are promising. Finally,
patients with advanced pancreatic cancer should be screened
and treated early for depression as a common comorbid
diagnosis. Early palliative care consultation also helps address the existential and psychosocial concerns of patients
facing death from pancreatic cancer in a holistic manner.
pancreatic cancer at time of diagnostic endoscopic ultrasound (EUS). The 96 participants were randomized to either
CPN or usual medical management at time of EUS diagnosis. Persistently increasing pain scores were noted in the
control group during the study; however, patients who
underwent neurolysis reported improvements in analgesia
both 1 and 3 months later with a statistically significant
decrease of 49% in mean pain score.4
The study from Wyse and colleagues adds to the body of
evidence supporting early CPN for patients with pancreatic
cancer.3 Optimal timing would be at moment of diagnosis if
a patient reports abdominal pain attributable to inoperable
pancreatic cancer.
Because these investigators took a detailed pain history
before diagnostic EUS/endoscopic retrograde cholangiopancreatography (ERCP), patients were able to benefit from
CPN with early, lasting pain relief.4 For patients undergoing surgical exploration, CPN should be considered intraoperatively for early, seamless analgesia once a diagnosis
is secured. With recurrent pain, repeat CPN is indicated
and effective, particularly if a patient had benefit from prior
neurolysis.3
Nausea and Vomiting
Nausea or vomiting in a patient with APC can arise from

multiple etiologies. With new onset, a patient should be
evaluated for potentially reversible causes, some of which
are discussed later in this article. While receiving antineoplastic therapy, adequate support with antiemetics should
be provided per the guidelines of the American Society of
Clinical Oncology (ASCO) or the National Cancer Care
Network (NCCN), available online. As FOLFIRINOX is
From the Department of Hematology and Oncology, The Medical College of Wisconsin,
Froedtert Memorial Hospital.
Address reprint requests to Lauren Wiebe, MD, Department of Hematology and Oncology,
The Medical College of Wisconsin, 9200 W. Wisconsin Ave., Milwaukee, WI 53226; email:
laurenwiebe@gmail.com.
1092-9118/10/1-10
243
LAUREN A. WIEBE
more commonly given for APC, special attention should be

paid to this aspect of patient care.
ASCO provides evidence-based guidelines for medications
based on likelihood of emesis by chemotherapy regimen.5
Published in 2011, the recommended support for highlyemetogenic chemotherapy consists of a three-drug regimen:
dexamethasone, a 5HT3 inhibitor, and a neurokinin-1
(NK-1) inhibitor. For moderately-emetogenic chemotherapy,
data supports a single dose of palonestron followed by 3
days of dexamethasone. The guidelines suggest dexamethasone alone as premedication for patients starting a lowemetogenic regimen such as gemcitabine, though the steroid
dose may be decreased in APC patients with glucose intolerance or new diabetes.5
In the most recent edition, both the NCCN and ASCO
guidelines acknowledge the antiemetic effect of olanzapine.5,6 The authors of ASCO guidelines suggest olanzapine
for patients with refractory chemotherapy-induced nausea
and vomiting (CINV) that occurs despite appropriate prophylaxis. The NCCN guidelines include olanzapine 2.55 mg
twice daily for patients with breakthrough CINV.6 However,
subsequent to publication of these guidelines, additional
data has emerged showing superior control of delayed nausea with an olanzapine-based regimen over the standard
recommendations for patients undergoing highly emetogenic chemotherapy, which included an NK-1 inhibitor.7
An atypical antipsychotic, olanzapine blocks several neurotransmitters including serotonin, dopamine, acetylcholine, and histamine. The proposed antiemetic mechanism is
thought to be D2 and 5HT3 inhibition, used in an off-label
indication for nausea.7,8 Two recent large randomized studies demonstrate efficacy for patients receiving highly emetogenic chemotherapy. The largest effect of olanzapine
appears to be on delayed nausea with complete control as
late as day 5 or 7.7,8 Additional improvements in global
functioning, insomnia, appetite, and fatigue have been
found to be significant over placebo (p 0.01).7,8 Weight
gain and hearty appetite are described side effects of olanzapine.

Olanzapine appears to be well-tolerated in patients undergoing chemotherapy.7,8 In 123 patients, no grade 3 or 4
toxicity was found by Navari and colleagues, though patients noted fatigue, drowsiness, and dry mouth most often.7
Prescribers should familiarize themselves with potential
toxicities. Prolonged administration, for months to years,
increases the risk of diabetes and pancreatitis in the psychiatric population, though this is less concerning with the
prognosis of APC. Finally, olanzapine carries a black box
warning for increased mortality in elderly patients with
dementia-related psychosis.
Malignant Gastroparesis
Gastroparesis associated with pancreatic cancer is a welldescribed phenomenon. Up to one-half of patients with APC
experience slowed gastric emptying with no anatomic obstruction.9 Cancer in the pancreas alters normal gut motility
by direct infiltration of autonomic nerve fibers and by
altering neurohormonal pathways within the bowel.9 Prior
abdominal surgery, radiation, or comorbid diabetes also
contribute to gastroparesis.9
A careful medication review and esophagogastroduodenoscopy may rule out reversible causes of gastroparesis. The
gold standard for diagnosis is functional imaging, but APC
patients with refractory nausea, vomiting, bloating, and
early satiety should be started on prokinetics empirically
either metoclopramide 510 mg four times daily or erythromycin.9 Small high-calorie, low-fiber meals with ample
liquids are best. Patients feel better sitting upright postprandially, followed by ambulation. Palliation of severe
cases can be achieved with decompressive gastrostomy or
Roux-en-Y, although overall prognosis and the presence of
ascites should be considered carefully before interventions.9
Gastric Outlet Obstruction
KEY POINTS
244
Celiac plexus neurolysis alleviates pain in the majority of patients with pancreatic cancer and should be
performed early even at time of cancer diagnosis.
With highly emetogenic chemotherapy indicated for
advanced pancreatic cancer, nausea and vomiting
should be managed rigorously per evidence in the
available literature.
Malignant bowel obstruction can be palliated quickly
with optimal medical management: gastric decompression, octreotide, parenteral opioids, and standing
antiemetics.
In patients with pancreatic cancer, exocrine insufficiency contributes to weight loss and failure to thrive,
but is reversible with pancreatic enzyme replacement
therapy.
Depression occurs in the majority of patients with
advanced pancreatic cancer and should be treated
aggressively and holistically for improved quality of
life.
Approximately 15% of patients with APC develop gastric

outlet obstruction (GOO) during the disease trajectory and
present with refractory vomiting.10 Clinical suspicion can be
confirmed with a plain radiograph alone, although patients
often undergo ultrasound, CT scan, or endoscopy to confirm
the diagnosis. Historically, an open, palliative gastrojejunostomy (GJ) was the definitive treatment for malignant GOO,
but surgery can result in substantial morbidity and even
mortality.10,11 More recently, self-expanding metallic stents
(SEMS) have become a safe and easier alternative with
minimal morbidity and mortality.
In a systematic review of the literature, endoscopic stent
placement was feasible in 96% of patients with malignancyrelated GOO, only limited by inability to transverse the
obstruction or stent deployment failure. After 1 week, 72% of
patients who underwent SEMS were eating soft or regular
foods compared with 14% of patients post GJ. Mean hospital
stay is shorter with stent placement as compared with
surgery (7 vs. 13 days). Repeat obstructive symptoms occurred in 18% of patients with a stent compared with 1% of
surgical cases, but overall, few complications are seen in
patients with SEMS.10 GJ, particularly laparoscopic, remains a viable palliative option, but only for a select group of
healthy patients with a longer prognosis.10,11
OPTIMIZING PALLIATIVE CARE FOR APC
Fig. 1. Algorithm for assessing and managing a patient with malignant bowel obstruction (MBO). Reprinted from Ripamonti CI, Easson AM,
Gerdes H. Management of malignant bowel obstruction. European Journal of Cancer. 2008;44(8):1105-1115, with permission from Elsevier.
Malignant Bowel Obstruction
Many patients with APC develop peritoneal implants and

carcinomatosis. Unfortunately, this can result in malignant
bowel obstruction (MBO) with abdominal distension, profound cramping, and refractory vomiting causing the inability to eat and often death. Palliative surgical or endoscopic
interventions can be considered but may be limited by
location of disease, multifocal obstruction, presence of ascites, and nutritional status of the patient.11 Detailed in
Figure 1, complex decision making in an urgent setting is
difficult, and MBO can quickly change the trajectory of a
patients life.
A combined approach of gastric decompression with optimal pharmacologic therapy provides relief of suffering as
well as reversal of MBO in the majority of cases. Adequate
pain control with strong opioids is critical by either intravenous or subcutaneous route; transdermal medications
should be reserved for later when a patient has stable opioid
requirements. Antisecretory agents are critical to reduce
splanchnic blood flow, bowel secretions, fluid loss, and
cramping pain. Considered the standard of care in MBO,
octreotide has been shown to be effective for palliation in
prospective trials. An acceptable total dose of octreotide
ranges from 300 1,200 mcg per day given via intravenous or
245
LAUREN A. WIEBE
Fig. 2. Common symptoms and complications of advanced pancreatic cancer with

corresponding recommendations.
Abbreviations: WHO, World Health Organization; SEMS, self-expanding metal
stent; NG, nasogastric; SAAG, serumascites albumin gradient; IU, international
units; SSRI, selective serotonin reuptake
inhibitor.
subcutaneous bolus or infusion. Haloperidol is the gold

standard to relieve nausea and vomiting associated with
MBO, although a combination of antiemetics may be required. Corticosteroids may help with nausea and bowel
edema but are less well studied in this setting.11
The goal for managing MBO is complete relief of vomiting,
nausea, and pain. Gastric decompression via nasogastric
tube provides immediate symptom relief, but should be
temporarily used, on the order of days. With aggressive
around-the-clock pharmacologic therapy detailed above, the
goal is to eliminate symptoms within 24 hours. Without
clear signs of returning bowel function in 2 or 3 days, a
venting gastrostomy should be placed early for optimal
palliation and return of patient mobility.11 Hospice and
homecare agencies are familiar with home drainage protocols, which the patient and/or family can learn.
Malignant Ascites
Malignant ascites causes discomfort, early satiety, bowel

stasis, orthopnea, and diminished mobility. In APC, peritoneal fluid accumulation occurs by several mechanisms, including portal hypertension from intrahepatic metastases,
portal vein compression and/or thrombus, peritoneal tumor
implants with increased capillary permeability, or direct
disruption of lymphatic drainage from the peritoneum.12
Treatment goals are to palliate symptoms and preserve
normal function as long as possible.
For exudative ascites, dietary sodium restriction and
246
diuretics may provide relief. These patients have a serumascites albumin gradient (SAAG) greater than 1.1 and
elevated portal pressures from venous occlusion or diffuse
hepatic metastases. However, patients with carcinomatosis
resulting in transudative ascites and a low SAAG ( 1.1) are
often primarily refractory to diuretic combinations and
should undergo early paracentesis for palliation.12 For all
patients with malignant ascites, therapeutic paracentesis
relieves symptoms 90% of the time with minimal risk of
adverse events.12
The majority of patients (88% to 100%) with anticipated
survival greater than 2 to 3 months will benefit from
placement of an indwelling pigtail or tunneled catheter to
control ascites with easy drainage at home or clinic. In a
meta-analysis, tunneled catheters have a low risk of infectious complications (2.5%), but no randomized study has
been performed. Nontunneled devices have a reported complication rate as high as 30% in some series.12 Although an
off-label use in the United States, PleurX catheters have
been studied in this setting and are commonly used with
success for palliation of refractory malignant ascites.12
Pancreatic Exocrine Insufficiency
Patients with pancreatic cancer are at high risk for

pancreatic exocrine insufficiency (PEI). When rigorously
assessed 1 year after partial pancreatectomy, 55% of patients have PEI by diagnostic criteria.13 Radiation and
surgery may cause ductal fibrosis resulting in poor delivery
OPTIMIZING PALLIATIVE CARE FOR APC
of lipase and trypsin into the gut lumen. In situ pancreatic

tumors block secretion of enzymes, leading to failure absorbing fat-soluble vitamins such as A, D, E, and K. Although
patients may not manifest clinical steatorrhea until the
lipase concentration falls below 10% of normal, symptoms of
maldigestion can be debilitating with abdominal cramping,
bloating, and urgent stools.14 PEI also contributes greatly
to the weight loss and malnutrition seen in patients with
APC.14
Diagnostic testing can be pursued but is expensive, difficult to perform, and usually unnecessary in this population.
Patients with pancreatic cancer and weight loss should be
given empiric supplementation with pancreatic enzymes.
When taken with each meal or shortly afterward, 40,000
50,000 IU of lipase results in substantial weight gain over
placebo in several studies.14 If not clinically effective after
several weeks, the dose should be increased as tolerated.
Because a low gastric pH irreversibly inhibits lipase, enzymes should never be taken on an empty stomach. H2blockade and proton-pump inhibition both augment the
effect of enzyme supplementation; either acid-suppressant
can be added for augmented absorption of fat. Low-fat
diets have not been shown to improve symptoms, but
moderate-fat diets are acceptable paired with enzyme supplements.14 Medium-chain fatty acids are absorbed without
lipase, so referral to a specialized dietician is helpful for
guidance.
Anorexia-Cachexia Syndrome
One of the most striking symptoms observed in patients

with pancreatic cancer is the anorexia-cachexia syndrome
(ACS). With circulating proinflammatory cytokines, patients
demonstrate negative protein-energy balance along with
minimal desire to eat, resulting in profound muscle wasting
and functional decline. The hallmark of cachexia is that the
syndrome is incompletely reversed by appetite stimulants
and nutritional intervention.15 ACS can be seen in earlystage pancreatic cancer, but occurs in more than one-half of
patients with advanced disease leading to poor quality of life
and survival.15
The ability to stabilize the weight loss can improve quality
of life, but interventions remain limited with no FDAapproved therapy for ACS.15 Currently, options include
targeted dietary counseling and appetite stimulants, typically with oral progestin.16 Berenstein and Ortiz performed
a rigorous review of the literature evaluating megestrol
acetate compared with placebo in patients with advanced
cancer. In total, megestrol was found to have a relative risk
of 3.03 (95% CI 1.83 to 5.01) for improved appetite compared
with placebo. Most trials utilized a dose between 400 and
800 mg daily. Both physicians and patients should be aware
of the prothrombotic tendency of progestins.17 No other
orexigenics have a documented benefit in ACS, though
individual patient response is variable to drugs such as
dronabinol.16
Recent research on ACS is promising and may be clinically
translatable in the near future. Recent phase II interventions targeted components of the inflammatory pathway
such as IL-6, COX-2, TNF-alpha, NF-kappaB, and growth
hormone.15 A phase II dose-escalation study of thalidomide
in 35 patients with cancer-related ACS resulted in significantly improved appetite in 64% of patients after 2 weeks
(p 0.001).18 Finally, dietary omega-3 fatty acids suppress
inflammatory and angiogenic pathways in pancreatic cancer

models, suggesting a potential future role for patients with
APC and ACS.19
Depression
Patients with pancreatic cancer have the highest incidence of depression seen in any cancer population.20 Studies
report rates of depression in APC ranging from 33% to 76%
with resultant poor quality of life and difficulty achieving
pain control.21 Elevated levels of circulating cytokines such
as IL-6 and TNF-alpha are thought to alter neurohormonal
pathways in the brain causing depressive symptoms even
before the diagnosis of cancer is suspected.21 In particular,
men with pancreatic cancer have a death rate from suicide
11 times that of the general population, reflecting the need
for an improved holistic approach to care of these patients.2,20,22
In patients with advanced cancer, it may be difficult to tell
clinical depression from normal sadness experienced with a
daunting prognosis. Common somatic symptoms of APC
such as fatigue, anorexia, and weight loss can overlap with
the signs and symptoms of depression.21 Simply asking a
patient whether he or she has felt depressed most of the
time is a validated tool with good sensitivity and specificity
for depression, even in patients with terminal illness.2
Timely treatment of depression is critical in APC, as patients with depression are more likely to endorse a desire for
hastened death.2
Antidepressant medicationsmainly selective serotonin
reuptake inhibitors (SSRIs) have been shown to be effective in patients with advanced cancer. Selection of SSRI
should be dictated by the side effect profile.21 In APC
patients suffering marked anorexia, mirtazapine may be an
advantageous choice, although not studied for this indication.21 Anxiolytics are helpful in patients with underlying
anxiety disorders. In clinical practice, lorazepam is used
frequently without clear evidence to guide ongoing administration. Patients requiring more than twice daily dosing of
a short-acting benzodiazepine could be switched to clonazepam with a longer half-life to avoid rebound anxiety,
usually starting at 0.5 mg orally twice daily.
In addition to pharmacologic interventions, patients
with APC benefit greatly from supportive counseling to
strengthen innate coping strategies and help with anticipatory grief as an integral part of their cancer care.2,21 Additional support will be provided by early referral to a
palliative care provider. Depression and anxiety often occur
because of distress from unaddressed fears of death or the
symptoms that may arise in the process of dying.2 Early and
frequent discussion of symptom concerns and quality of life
preserves hope for patients with APC. Active interventions
such as Dignity Therapy have shown promise in relief of
suffering in patients with limited prognosis.2 ASCO recently
released a statement of provisional clinical opinion that
palliative care leads to better patient and caregiver outcomes.23 In the end, patients want to be individuals acknowledged with compassion and healing at this time in
their lives.
Conclusion
Patients with advanced pancreatic cancer suffer numerous symptoms throughout the illness. It is critical that these
247
LAUREN A. WIEBE
patients are cared for completely with aggressive palliation

of symptoms to maximize their remaining time. Celiac
plexus neurolysis improves pain in the majority of APC
patients and should be moved earlier in the analgesic
paradigm. Antiemetics continue to evolve for maximal prevention of CINV, including olanzapine in recent studies.
Nutritional status remains important to patients and families; weight loss related to undiagnosed pancreas exocrine
insufficiency improves with aggressive enzyme supplementation. Bowel dysfunction from gastroparesis or malignant
obstruction is treatable with palliative interventions. Finally, the existential and psychosocial concerns of patients
facing death from pancreatic cancer should be addressed in
a holistic manner. Early integration of palliative care will
help patients achieve the best quality of life in the face of
this unfortunate diagnosis.
Author
Lauren A. Wiebe*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Siegel R, Naishadham D, Jemal A. Cancer Statistics, 2012. CA Cancer
J Clin. 2012;62:10-29.
2. Chochinov HM. Dying, dignity and new horizons in palliative end-of-life
care. CA Cancer J Clin. 2006;56:84-103.
3. Yan BM, Myers RP. Neurolytic celiac plexus block for pain control in
unresectable pancreatic cancer. Am J Gastroenterol. 2007;102:430-438.
4. Wyse JM, Carone M, Paquin SC, et al. Randomized, double-blind,
controlled trial of early endoscopic ultrasound-guided celiac plexus neurolysis
to prevent pain progression in patients with newly diagnosed, painful,
inoperable pancreatic cancer. J Clin Oncol. 29:3541-3546.
5. Basch E, Prestrud AA, Hesketh PJ, et al. Antiemetics: American Society
of Clinical Oncology Practice Guideline Update. J Clin Oncol. 2011;29:41894198.
in Oncology. http://www.nccn.org/professionals/physician_gls/f_guidelines.
asp#supportive. Accessed February 29, 2012.
7. Navari RM, Gray SE, Kerr AC. Olanzapine versus aprepitant for the
prevention of chemotherapy-induced nausea and vomiting: A randomized
phase III trial. J Support Oncol. 2011;9:188-195.
8. Tan L, Liu J, Liu X, et al. Clinical research of olanzapine for prevention
of chemotherapy-induced nausea and vomiting. J Exp Clin Cancer Res.
2009;29:131.
9. Donthireddy KR, Ailawadhi S, Nasser E, et al. Malignant gastroparesis:
Pathogenesis and management of an underrecognized disorder. J Support
Oncol. 2007;5:355-363.
10. Juernink SM, van Eijck CHJ, Steyerberg EW, et al. Stent versus
gastrojejunostomy for the palliation of gastric outlet obstruction: A systematic
review. BMC Gastroenterol. 2007;7:18.
11. Ripamonti CT, Easson AM, Gerdes H. Management of malignant bowel
obstruction. Eur J Cancer. 2008;44:1105-1115.
12. Chung M and Kozuch P. Treatment of malignant ascites. Curr Treat
Options Oncol. 2008;9:215-233.
248
13. Halloran CM, Cox TF, Chauhan S, et al. Partial pancreatic resection for
pancreatic malignancy is associated with sustained pancreatic exocrine
failure and reduced quality of life: A prospective study. Pancreatology.
2011;11:535-545.
14. Dominguez-Munoz JE. Pancreatic exocrine insufficiency: Diagnosis
and treatment. J Gastroenterol Hepatol. 2011;26 Suppl. 2:12-16.
15. Dodson S, Baracos VE, Jatoi A, et al. Muscle wasting in cancer
cachexia: Clinical implications, diagnosis, and emerging treatment strategies.
Annu Rev Med. 2011;62:265-279.
16. Dy SM, Lorenz KA, Naeim A, et al. Evidence-based recommendations
for cancer fatigue, anorexia, depression and dyspnea. J Clin Oncol. 2008;26:
3886-3895.
17. Berenstein G, Ortiz Z. Megestrol acetate for treatment of anorexiacachexia syndrome. Cochrane Database of Systematic Reviews 2005, Issue 2.
Art. No.: CD004310.
18. Davis M, Lasheen W, Walsh D, et al. A phase II dose titration study of
thalidomide for cancer-associated anorexia. J Pain Symptom Manage. 2012;
43:78-86.
19. Boutros C, Somasundar P, Razzak A, et al. Omega-3 fatty acids:
Investigations from cytokine regulation to pancreatic cancer gene suppression. Arch Surg. 2010;145:515-520.
20. Clark KL, Loscalzo M, Trask PC, et al. Pyschological distress in
patients with pancreatic canceran understudied group. Pychooncology.
2010;19:1313-1320.
21. Mayr M, Schmid RM. Pancreatic cancer and depression: Myth and
truth. BMC Cancer. 2010;10:569.
22. Turaga KT, Malafa MP, Jacobsen PB, et al. Suicide in patients with
pancreatic cancer. Cancer. 2011;117:642-647.
23. Smith TJ, Temin S, Alesi ER, et al. American Society of Clinical
Oncology Provisional Clinical Opinion: The Integration of Palliative Care into
Standard Oncology Care. http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2011.
38.5161. Accessed February 29, 2012.
GLOBAL PERSPECTIVE OF LOCALLY ADVANCED

GASTRIC CANCER: DIFFERENT TREATMENT
PARADIGMS AND THEIR RATIONALE
CHAIR
Manish A. Shah, MD
New York, NY
SPEAKERS
Sam S. Yoon, MD
Boston, MA
Yoon-Koo Kang, MD, PhD
Asan Medical Center
Seoul, South Korea
Varying Lymphadenectomies for Gastric

Adenocarcinoma in the East Compared
with the West: Effect on Outcomes
By Benjamin Schmidt, MD, and Sam S. Yoon, MD
Overview: There are notable differences in surgical approaches to gastric adenocarcinoma throughout the world,
particularly in terms of the extent of lymphadenectomy (LAD).
In high-incidence countries such as Japan and South Korea,
more extensive (e.g., D2) lymphadenectomies are standard,
and these surgeries are generally done by experienced surgeons with low morbidity and mortality. In countries such as
the United States, where the incidence of gastric adenocarcinoma is 10-fold lower, the majority of patients are treated at
nonreferral centers with less extensive (e.g., D1 or D0) lymphadenectomy. There is little disagreement among gastric cancer (GC) experts that the minimum lymphadenectomy that
should be performed for gastric adenocarcinoma should be at
least a D1 lymphadenectomy, and many of these experts
recommend a D2 lymphadenectomy. More extensive lymphadenectomies provide better staging of patient disease and
T IS estimated that there are more than one million cases

of GC worldwide per year, making it the fourth most
common cancer.1 Nearly three-quarters of cases occur in
developing countries, and nearly half of cases occur in
eastern Asia (mainly in China). GC is the second leading
worldwide cause of cancer death for both men and women,
with a total of more than 700,000 deaths each year. The
incidence of gastric adenocarcinoma varies tremendously
throughout the world and country by country, with the
highest incidence occurring in South Korea at 66.5 to 72.5
per 100,000 males and 19.5 to 30.4 per 100,000 females.2
The incidence of GC in the United States is only one-tenth
that of South Korea. The estimated number of new GC cases
in the United States in 2012 was 21,320, and the estimated
number of deaths was 10,540.3
In addition to the global differences in GC epidemiology,
there are also appreciable differences in the surgical treatment of GC, particularly in the extent of LAD. This article
will examine the effect of varying LADs in Eastern and
Western countries on patient outcomes. Institutional studies from two countries, Japan and South Korea, will be used
to represent two high-incidence Eastern countries, and institutional and national database studies from the United
States will be used to represent low-volume Western countries.
Definitions
Before discussion of differences in LAD for gastric adenocarcinoma, one should define the terms to be used. The node
stations surrounding the stomach were precisely defined by
the Japanese Gastric Cancer Association (JGCA), formerly
known as the Japanese Research Society for Gastric Cancer, in 19734 (Fig. 1 and Table 1). In its most recent GC
treatment guidelines, the JGCA again changed the definitions for D levels of LAD such that they are now defined
according to the type of gastrectomy performed rather than
the location of the tumor (Table 2).5 To broadly summarize,
a D1 LAD removes the first tier of perigastric nodes and the
left gastric artery nodes whereas a D2 LAD removes the
250
likely reduce locoregional recurrence rates. Two large, prospective randomized trials performed in the United Kingdom
and the Netherlands in the 1990s failed to demonstrate a
survival benefit of D2 over D1 lymphadenectomy, but these
trials have been criticized for inadequate surgical training and
high surgical morbidity and mortality rates (10% to 13%) in the
D2 group. More recent studies have demonstrated that Western surgeons can be trained to perform D2 lymphadenectomies on Western patients with low morbidity and mortality.
The 15-year follow-up of the Netherlands trial now demonstrates an improved disease-specific survival and locoregional recurrence in the D2 group. Retrospective analyses and
one prospective, randomized trial suggest that there may be a
survival benefit to more extensive lymphadenectomies when
performed safely, but this assertion requires further validation.
second tier of nodes that generally fall along primary and

secondary branches of the celiac axis (i.e., splenic artery,
common hepatic artery, proper hepatic artery). The JGCA
guidelines recommend a D2 LAD for all gastric carcinomas
beyond a clinical T1 tumor (e.g., tumor invades lamina
propria, muscularis mucosa, or submucosa).
Differences in Surgical Volume and Extent of LAD
Japan and South Korea have two of the highest incidences

of gastric adenocarcinoma in the world, but despite the high
incidence of gastric adenocarcinoma in these countries,
patients are often referred to tertiary centers for treatment.
Two-thirds of all GC surgeries in South Korea are performed
at 16 high-volume institutions, which perform at least 200
GC surgeries per year. Thus GC surgeons at high-volume
institutions in South Korea gain tremendous experience in
the surgical management of GC. As noted earlier, the
minimum LAD performed by Japanese and Korean surgeons
for gastric adenocarcinoma (except for T1 tumors) is a D2
LAD.5 Despite performing extensive LADs, the morbidity
and mortality rates are quite low. For example, Seoul
National University Hospital (SNUH), which performs almost 1,000 GC operations per year, recently reported a
morbidity rate of 18% and a mortality rate of 0.5%.6
In contrast, the majority of GC surgeries in the United
States are performed at nonreferral centers. A high volume institution in the United States has been defined in
some studies as centers with as low as more than 15 to 20
surgeries per year.7,8 Birkmeyer and colleagues reviewed a
database of Medicare patients and found that hospitals with
more than 20 gastrectomies per year had one-third less risk
From the Department of Surgery, Massachusetts General Hospital, Boston, MA; Harvard
Medical School, Boston, MA.
Address reprint requests to Sam S. Yoon, MD, Division of Surgical Oncology, Department
of Surgery, Massachusetts General Hospital, Yawkey 7B, 55 Fruit St., Boston, MA 02114;
email: syoon@partners.org.
1092-9118/10/1-10
GASTRIC CANCER LYMPHADENECTOMY AND OUTCOMES
of perioperative death (odds ratio 0.55 0.74), yet more than

80% of patients were operated on at centers that performed
20 or fewer gastrectomies per year.7 As most U.S. surgeons
see only a few GC patients a year, they likely err on the side
of more limited LADs in order to avoid excess morbidity and
mortality. In the Intergroup 0116 trial published in 2001,
patients were randomly selected after GC surgery to undergo no further therapy or chemoradiation, and more than
50% of operations were less aggressive than a D1 LAD (aka
D0 LAD).9 Despite the performance of less extensive LADs
in the United States, surgical morbidity and mortality rates
for gastric adenocarcinoma are generally much higher in the
United States than in South Korea and Japan. An analysis
of the Nationwide Inpatient Sample from 1998 to 2003 of
more than 50,000 patents with GC found the overall mortality rate following gastric surgery was 6%.10 Singleinstitution series have reported morbidity rates following
gastrectomy of up to 40%.11
Differences in Survival
Gastric adenocarcinoma frequently metastasizes to regional nodes. For T1 lesions invading the submucosa, node
involvement is found in approximately 20% of patients.12
For T2 lesions (invading muscularis propria), the node
metastasis rate increases to more than 50%. There is some
evidence that at least some patients with node metastases
beyond the immediate perigastric (D1) nodes and into D2
nodes can be cured with surgical resection alone.13 A sizable
minority of GC patients with positive D2 nodes survive for
more than 5 years following D2 lymphadenectomy at the
Japanese National Cancer Center in Tokyo.
Numerous studies have demonstrated decreased overall
KEY POINTS
More extensive D2 lymphadenectomies are standard

in high-incidence Eastern countries such as Japan
and South Korea, leading to better staging of disease
and likely lower rates of locoregional recurrence.
In the United States (a low-incidence Western country), the vast majority of gastric resections are performed at low-volume (less than 20 cases per year)
centers with generally less extensive lymphadenectomies and higher morbidity and mortality.
Stage for stage, overall survival is worse in the
United States than in Japan and South Korea, but
much of this difference could be explained by stage
migration and clinicopathologic differences between
gastric cancers in Eastern versus Western countries.
The Dutch and U.K. D1 versus D2 lymphadenectomy
randomized trials were flawed, and further prospective
randomized trials of lymphadenectomies performed by
well-trained Western surgeons on Western patients are
needed to determine if there is an overall survival
benefit to more extensive lymphadenectomies.
Strategies to improve the surgical outcomes of patients with gastric cancer in low-incidence Western
countries include referral to tertiary centers and
improved training of surgeons.
Fig 1. Location of node stations according to the Japanese Gastric

Cancer Association.38 (A) Perigastric nodes stations 1 to 7. (B) Second
tier node stations 8 to 12 and 14.
survival (OS) after potentially curative gastrectomy for

gastric adenocarcinoma in the West compared with the East.
Table 3 demonstrates 5-year OS results stage-for-stage from
four large databases based on the sixth American Joint
Committee on Cancer (AJCC) staging system. D2 LADs are
generally performed at the National Cancer Center (NCC) in
Tokyo, Japan, and at SNUH is South Korea. The median
number of examined nodes at both these institutions is
greater than 30, and there is a remarkable similarity in the
5-year survival figures from these two institutions. In the
U.S. Surveillance Epidemiology and End Results (SEER)
database, most patients had either a D0 or D1 LAD, and the
median number of examined nodes is 10 to 11.14 For stages
I to III, the 5-year survival rate is 14% to 30% lower for
SEER database patients. At Memorial Sloan-Kettering Cancer Center, where approximately 80% of patients receive a
D2 LAD,15 stage-for-stage OS is intermediate between
SEER database patients and NCC/SNUH patients.16 Some
of these differences in 5-year survival can clearly be attributed to stage migration.17 As more nodes are harvested,
more malignant nodes can be found, leading to a higher
staging of patients.
There are also clearly some differences in the clinical and
pathologic presentation and adjuvant treatment of GC in the
West compared with the East that make comparison of
outcomes difficult. In terms of patient demographics, West-
251
SCHMIDT AND YOON

Table 1. Regional Lymph Nodes of the Stomach
Number
1
2
3
a
b
4
sa
sb
d
5
6
7
8
a
p
9
10
11
p
d
12
a
b
p
14
v
a
Description
Right paracardial
Left paracardial
Lesser curvature
Along branches of left gastric artery
Along second branch and distal part of right gastric artery
Greater curvature
Along short gastric vessels
Along left gastroepiploic vessels
Along second branch and distal part of right gastroepiploic artery
Suprapyloric along first branch and proximal part of right gastric artery
Infrapyloric along first branch and proximal part of right gastroepiploic
artery
Left gastric artery
Common hepatic artery
Anterosuperior group
Posterior group
Celiac artery
Splenic hilum
Along splenic artery
Along proximal splenic artery
Along distal splenic artery
Hepatoduodenal ligament
Along proper hepatic artery
Along bile duct
Along portal vein
Along superior mesenteric vessels
Along superior mesenteric vein
Along superior mesenteric artery
Adapted from the Japanese Gastric Cancer Associations 2011 classifications. 38
ern patients compared to Eastern patients are generally: (1)

older, (2) have a higher body mass index, (3) have a lower
incidence of H pylori infection, (4) have more proximal
tumors, (5) present with later stage disease, and (6) receive
different adjuvant therapies. Many of the factors more
common in Western patients are negative prognostic factors
for gastric adenocarcinoma. Verdecchia and colleagues compared GC incidence and survival in four regions (Campina,
Brazil; Iowa, United States; Varese province, Italy; and
Osaka Japan).18 U.S. patients were older, had more proximal cancers, more commonly presented with metastatic
disease, and had worse survival rates than Japanese patients.
One of the best attempts to determine if Eastern patients
generally have a better OS than Western patients following
more extensive LADs was performed by Strong and colleagues. They analyzed 711 U.S. patients treated at Memorial Sloan-Kettering Cancer Center and 1,646 Korean
patients treated at Seoul St. Marys Hospital.19 In this
study, the median age of U.S. patients was 10 years older
than that of Korean patients (69 vs. 59 years old). Thirtynine percent of U.S. patients had upper third or gastroesophageal junction tumors compared to only 9.4% of
Korean patients, and 59% of U.S. patients had intestinal
type tumors compared to 49% of Korean patients. D2 LAD
was performed in 84% and 89% of U.S. and Korean patients, respectively, but there were more nodes examined
in Korean patients than U.S. patients (97% of Korean
patients with 15 nodes examined compared to 78% of U.S.
patients). The T stage, N stage, and overall stage of U.S.
patients were significantly more advanced than those of
252
Table 2. Extent of Lymphadenectomy a

Extent of Gastrectomy
D1 Dissection
Total gastrectomy
17
Distal/subtotal
gastrectomy
Proximal
gastrectomy
1, 3, 4sb, 4d, 5,
6, 7
1,2,3a, 4sa, 4sb,
7
D1 Dissection
D2 Dissection
D1 8a, 9p,
11p
D1 8a, 9
D1 8a, 9p, 11p,

11d, 12a
D1 8a, 9, 11p,
12a
N/A
D1 8a, 9,
11p
Adapted from the Japanese Gastric Cancer Associations 2010 guidelines. 5
Korean patients (p 0.0001, p 0.008, p 0.0001,

respectively). Survival was worse, stage-for-stage, in U.S.
patients compared to Korean patients for AJCC stages I to
III. Interestingly, the survival of U.S. patients with middle
or upper tumors was worse than that of Korean patients, but
U.S. and Korean patients had similar OS for distal tumors.
After adjusting for clinically important prognostic factors,
Korean patients still had a 30% better disease-specific survival rate than U.S. patients. There are some potential
confounding variables in this study, including the fact that
316 U.S. patients were excluded because they received
neoadjuvant treatment. However, this study does suggest
that after controlling for prognostic factors and following
relatively uniform D2 LAD, Eastern patients still have a
better survival rate compared with Western patients.
Potential Benefits of More Extensive LADs
LAD for cancer can serve three primary purposes: staging

of disease, prevention of locoregional recurrence, and improvement in OS. There is little doubt that more extensive
LADs for gastric adenocarcinoma can lead to better staging
of disease. The 2010 (seventh edition) AJCC Cancer Staging
Manual for gastric adenocarcinoma recommends that at
least 16 nodes be examined for correct assessment of the N
category.20 Despite this, our analysis of the SEER database
found that only one-third of 18,043 resected GC patients had
16 or more nodes examined.14 It is difficult to be confident
that a GC is truly node-negative when fewer than 10 nodes
are examined,21,22 and N1 tumors can be upstaged to N2 or
even N3 tumors as more nodes are harvested.22,23 Furthermore, it is impossible to be categorized as N3b if less than 16
nodes are harvested. Thus many patients are understaged
following surgical resection of their GCs because of inadequate node sampling.
In the Unites States, the pathologist is usually the one
who finds and examines the dissected nodes. Thus a coordiTable 3. Five-Year Overall Survival Rates of Patients with GC
by AJCC 6th Edition Stage
Lymphadenectomy
Median nodes examined
IA
IB
II
IIIA
IIIB
IV
SEER,14
n 32,531
MSKCC,16
n 1,038
NCC,
Japan,a
n 6,730
SNUH,
South Korea,39
n 12,026
D0, D1
1011
78%
58%
34%
20%
8%
7%
D2 D1
22
92%
85%
49%
32%
11%
11%
D2
30
91.5%
84.6%
69.3%
50.4%
30.6%
5.4%
D2
30
94.4%
84.2%
68.5%
47.3%
31.6%
17.2%
Abbreviations: AJCC, American Joint Committee on Cancer; MSKCC, Memorial

Sloan Kettering Cancer Center; NCC, National Cancer Center; SEER, Surveillance
Epidemiology and End Results; SNUH, Seoul National University Hospital.
a
From a written communication with H.K. Yang, SNUH (May, 2011).
nated effort is required between the surgeon and pathologist

if more extensive LADs are to result in improved staging of
patients. In Japan and South Korea, following the en bloc
dissection of the stomach and nodes, surgeons generally
dissect out the individual nodal stations from the surgical
specimen, allowing the pathologist to examine and report
the number of positive and negative nodes for each nodal
station. Thus more extensive LAD must be combined with
more rigorous pathologic analysis to optimize the node
staging of GC.
There is some evidence that more extensive LADs result
in lower rates of locoregional recurrence. Locoregional recurrence after potentially curative surgery for gastric adenocarcinoma can be quite high. In a 1982 series from the
University of Minnesota, 107 patients with gastric adenocarcinoma underwent second-look laparotomy, and 80% had
recurrence.24 Of these recurrences, 88% were locoregional,
54% were peritoneal, and 29% were distant. More recently in
the U.S. Intergroup 0116 trial, 177 of 275 patients (64%) in
the surgery-only group developed recurrent disease.9 In
terms of the site of first relapse, 29% had local recurrence,
72% had regional recurrence, and only 18% had distant
recurrence. Rates of locoregional recurrence are generally
lower in reports from both Western and Eastern institutions
that perform more extensive LADs. In a series of 367
patients with recurrent gastric adenocarcinoma from Memorial Sloan-Kettering Cancer Center over 15 years, 81% of
patients had a D2 or greater LAD, and the median number
of nodes removed was 22.15 Of patients in whom disease
recurred, locoregional recurrence was the initial and only
site of recurrence in 26% of patients and was a component of
initial recurrence in 54% of patients. Yoo and colleagues
examined 508 patients in whom recurrent disease developed
after curative gastrectomy at Yonsei University in South
Korea. Nineteen percent of patients had locoregional recurrence only as the first site of recurrence, and 32.5% of
patients had locoregional recurrence combined with peritoneal or distant recurrence as the initial site of recurrent
disease. In the Japanese prospective randomized trial of
adjuvant S-1 chemotherapy, 188 (35.5%) of 530 patients
treated with surgery suffered a recurrence.25 The site of first
recurrence in these 188 patients was local in 7.9% and in
nodes in 24.5%.
The effect of more extensive LADs on OS for GC is still
controversial. Two large, prospective randomized trials in
Western countries have failed to identify a survival advantage for D2 over D1 LAD.26,27 However, these two trials had
fairly high morbidity (43% to 46%) and mortality rates (10%
to 13%) for D2 LAD. In these trials, the distal pancreas and
spleen were often resected during dissection of station 10
and 11 nodes, which likely increased morbidity. Of note in
the Dutch trial, if patients with in-hospital mortality are
excluded, patients with N2 disease had a survival advantage
when treated with a D2 LAD.28 Several studies have now
demonstrated that D2 LADs can be performed without the
need for distal pancreatectomy29 or splenectomy.30,31 Furthermore, a recent randomized trial in Taiwan demonstrated an OS advantage of more extensive LAD over D1
LAD, with the overall 5-year survival rate being 59.5%
compared with 53.6%, respectively (p 0.041).32 However,
the applicability of this trial to Western patients has been
called into question.33 The long-term follow-up of the Dutch
trial was recently reported.34 After a median follow-up of

15 years, D2 LAD was associated with lower locoregional
recurrence and gastric cancerrelated death rates (37% vs.
48%) than D1 LAD. Degiuli and colleagues in Italy have
demonstrated that Western surgeons, following extensive
training, can perform D2 LADs on Western patients with
low morbidity and almost no mortality,35,36 and survival
results from a prospective randomized trial of D1 compared
with D2 LAD from this group are pending.37
Can Western Surgeons Perform More Extensive
LADs Safely?
Italian GC Study Group approached the issue of Western

surgeons performing more extensive LADs in Western patients in a series of two prospective clinical trials.35,36
Following extensive training of 16 surgeons in D2 LAD, a
phase II trial of D2 LAD was instituted in which all
surgeries were performed by the two attending surgeons. Of
the 191 patients enrolled in the study, 106 patients (55%)
were ultimately found to be ineligible, usually as a result of
more extensive disease. The mean number of nodes removed
was 39 (range: 22 to 93). Overall postoperative morbidity
and mortality were impressively low at 20.9% and 3.1%,
respectively. Subsequent to this study, the surgeons from
the five highest-volume centers performed a randomized
trial of D1 compared with D2 LAD.37 Of 267 randomly
selected patients, total morbidity and mortality was 12.0%
and 3.0%, respectively, in the D1 group and 17.9% and 2.2%,
respectively, in the D2 group. Survival results are pending.
The experience of the Italian GC Study Group clearly
demonstrates that following a period of fairly rigorous training, Western surgeons can perform D2 LADs on Western
patients with morbidity and mortality results similar to that
of high-volume centers in Korea and Japan.
Several tertiary referral centers in Western countries
routinely perform D2 LADs for GC,15 but as noted earlier,
LADs for GC in Western countries are limited and often do
not even reach the D1 LAD threshold. There are several
reasons why more extensive LADs are not more commonly
performed. First and foremost is the lack of a proven benefit
in OS of D2 over D1 LAD based on the Dutch and U.K. trials.
Unfortunately, many Western surgeons have interpreted
the results of these trials to mean any LAD does not improve
OS. Certainly some patients with node-positive disease are
cured by surgical resection alone, and these patients would
undoubtedly not have been cured if diseased nodes were left
undissected without additional therapy. Another important
obstacle to the performance of more extensive LADs is the
relative paucity of gastric adenocarcinomas seen at any
given institution. In order for more extensive LADs to
benefit GC patients, they must be performed without excess
morbidity and mortality, and this can only be achieved with
adequate surgical training and adequate case volume. Contributing to the lack of high-volume centers for GC surgery
is a potential reluctance of general surgeons to refer GC
patients to tertiary referral centers given that gastric surgery has been historically the realm of the general surgeon.38 Finally, there are geographical and language
barriers between different countries that make dissemination of information and techniques on the surgical treatment
of GC difficult.
253
SCHMIDT AND YOON

14
Conclusion
There are clear differences in the extent of LAD performed

between Eastern and Western countries. D2 LAD is the
standard LAD performed in Japan and South Korea for all
resectable tumors except for T1 tumors. Significantly less
extensive LADs are generally performed in the United
States. How does this difference in extent of LAD affect GC
patient outcomes? There is no doubt that less extensive
LADs result in understaging of patients.17 Many U.S. medical oncologists and radiation oncologists already factor this
understaging into their medical decision making. For any
patient in the SEER database who is stage IB or greater, the
5-year OS rate is 58% or less and adjuvant therapy is likely
warranted. However, a patient treated at MSKCC who is
stage IB has an 85% 5-year OS and may not warrant
adjuvant therapy. Our analysis of the SEER database found
wide variations in survival of each AJCC stage based on
subgroup analysis, with more than half of patients being
misclassified. Thus there are inherent problems with inaccurately staged patients.
Less extensive LADs also likely result in increased locoregional recurrence, making the decisions between adjuvant
chemotherapy versus chemoradiation more difficult. GC
patients are often limited in the extent of neoadjvuant and
adjuvant treatment that can be delivered, and one must
often choose between multiagent chemotherapy regimens
versus 5-fluorouracil or capecitabine-based chemoradiation.
In terms of OS, the effects of more extensive LAD are
difficult to discern. The Dutch and U.K. trials of D1 compared with D2 LAD demonstrated that when D2 LAD is
performed with excess morbidity and mortality, there is no
survival benefit compared to D1 LAD. If D2 LAD is performed with low morbidity and mortality, there also may be
a benefit in OS, at least in Chinese patients,32 but this
potential benefit needs to be demonstrated by future prospective, randomized trials of Western patients.
Author
Benjamin Schmidt*
Sam S. Yoon*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
J Clin. 2011;61:69-90.
2. Lee J, Demissie K, Lu SE, et al. Cancer incidence among KoreanAmerican immigrants in the United States and native Koreans in South
Korea. Cancer Control. 2007;14:78-85.
J Clin. 2012;62:10-29.
4. Japanese Research Society for Gastric Cancer. The general rules for the
gastric cancer study in surgery. Jpn J Surg. 1973;3:61.
5. Japanese Gastric Cancer Association. Japanese gastric cancer treatment
guidelines 2010 (ver. 3). Gastric Cancer. 2011;14:113-123.
6. Park DJ, Lee HJ, Kim HH, et al. Predictors of operative morbidity and
mortality in gastric cancer surgery. Br J Surg. 2005;92:1099-1102.
7. Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume and
surgical mortality in the United States. N Engl J Med. 2002;346:1128-1137.
8. Smith DL, Elting LS, Learn PA, et al. Factors influencing the volumeoutcome relationship in gastrectomies: a population-based study. Ann Surg
Oncol. 2007;14:1846-1852.
9. Macdonald JS, Smalley SR, Benedetti J, et al. Chemoradiotherapy after
surgery compared with surgery alone for adenocarcinoma of the stomach or
gastroesophageal junction. N Engl J Med. 2001;345:725-730.
10. Smith JK, McPhee JT, Hill JS, et al. National outcomes after gastric
resection for neoplasm. Arch Surg. 2007;142:387-393.
11. Marcus SG, Cohen D, Lin K, et al. Complications of gastrectomy
following CPT-11-based neoadjuvant chemotherapy for gastric cancer. J
Gastrointest Surg. 2003;7:1015-1022.
12. Gotoda T, Yanagisawa A, Sasako M, et al. Incidence of lymph node
metastasis from early gastric cancer: estimation with a large number of cases
at two large centers. Gastric Cancer. 2000;3:219-225.
13. Yoon SS, Yang HK. Lymphadenectomy for gastric adenocarcinoma:
should west meet east? Oncologist. 2009;14:871-882.
14. Wang J, Dang P, Raut CP, et al. Comparison of a lymph node
ratio-based staging system with the 7th AJCC System for Gastric Cancer:
analysis of 18,043 patients from the SEER database. Ann Surg. In press.
15. DAngelica M, Gonen M, Brennan MF, et al. Patterns of initial recurrence in completely resected gastric adenocarcinoma. Ann Surg. 2004;240:
808-816.
16. Karpeh MS, Leon L, Klimstra D, et al. Lymph node staging in gastric
cancer: is location more important than number? An analysis of 1,038
patients. Ann Surg. 2000;232:362-371.
17. Bunt AM, Hermans J, Smit VT, et al. Surgical/pathologic-stage migra-
254
tion confounds comparisons of gastric cancer survival rates between Japan

and Western countries. J Clin Oncol. 1995;13:19-25.
18. Verdecchia A, Mariotto A, Gatta G, et al. Comparison of stomach cancer
incidence and survival in four continents. Eur J Cancer. 2003;39:1603-1609.
19. Strong VE, Song KY, Park CH, et al. Comparison of gastric cancer
survival following R0 resection in the United States and Korea using an
internationally validated nomogram. Ann Surg. 2010;251:640-646.
20. American Joint Committee on Cancer. AJCC Cancer Staging Manual.
7th edition. New York: Springer; 2010.
21. Bouvier AM, Haas O, Piard F, et al. How many nodes must be examined
to accurately stage gastric carcinomas? Results from a population based
study. Cancer. 2002;94:2862-2866.
22. Smith DD, Schwarz RR, Schwarz RE. Impact of total lymph node count
on staging and survival after gastrectomy for gastric cancer: data from a large
US-population database. J Clin Oncol. 2005;23:7114-7124.
23. Estes NC, Macdonald JS, Touijer K, et al. Inadequate documentation
and resection for gastric cancer in the United States: a preliminary report. Am
Surg. 1998;64:680-685.
24. Gunderson LL, Sosin H. Adenocarcinoma of the stomach: areas of
failure in a re-operation series (second or symptomatic look) clinicopathologic
correlation and implications for adjuvant therapy. Int J Radiat Oncol Biol
Phys. 1982;8:1-11.
25. Sakuramoto S, Sasako M, Yamaguchi T, et al. Adjuvant chemotherapy
for gastric cancer with S-1, an oral fluoropyrimidine. N Engl J Med. 2007;
357:1810-1820.
26. Cuschieri A, Fayers P, Fielding J, et al. Postoperative morbidity and
mortality after D1 and D2 resections for gastric cancer: preliminary results of
the MRC randomised controlled surgical trial. The Surgical Cooperative
Group. Lancet. 1996;347:995-999.
27. Bonenkamp JJ, Songun I, Hermans J, et al. Randomised comparison of
morbidity after D1 and D2 dissection for gastric cancer in 996 Dutch patients.
Lancet. 1995;345:745-748, 1995.
28. Jansen EP, Boot H, Verheij M, et al. Optimal locoregional treatment in
gastric cancer. J Clin Oncol. 2005;23:4509-4517.
29. Maruyama K, Sasako M, Kinoshita T, et al. Pancreas-preserving total
gastrectomy for proximal gastric cancer. World J Surg. 1995;19:532-536.
30. Uyama I, Ogiwara H, Takahara T, et al. Spleen- and pancreaspreserving total gastrectomy with superextended lymphadenectomy including dissection of the para-aortic lymph nodes for gastric cancer. J Surg Oncol.
1996;63:268-270.

31. Biffi R, Chiappa A, Luca F, et al. Extended lymph node dissection
without routine spleno-pancreatectomy for treatment of gastric cancer: low
morbidity and mortality rates in a single center series of 250 patients. J Surg
Oncol. 2006;93:394-400.
32. Wu CW, Hsiung CA, Lo SS, et al. Nodal dissection for patients with
gastric cancer: a randomised controlled trial. Lancet Oncol. 2006;7:309-315.
33. Roggin KK, Posner MC. D3 or not D3. That is not the question. Lancet
Oncol. 2006;7:279-280.
34. Songun I, Putter H, Kranenbarg EM, et al. Surgical treatment of
gastric cancer: 15-year follow-up results of the randomised nationwide Dutch
D1D2 trial. Lancet Oncol. 2010;11:439-449.
35. Degiuli M, Sasako M, Ponti A, et al. Morbidity and mortality after D2
gastrectomy for gastric cancer: results of the Italian Gastric Cancer Study Group
prospective multicenter surgical study. J Clin Oncol. 1998;16:1490-1493.
36. Degiuli M, Sasako M, Calgaro M, et al. Morbidity and mortality after

D1 and D2 gastrectomy for cancer: interim analysis of the Italian Gastric
Cancer Study Group (IGCSG) randomised surgical trial. Eur J Surg Oncol.
2004;30:303-308.
37. Degiuli M, Sasako M, Ponti A. Morbidity and mortality in the Italian
Gastric Cancer Study Group randomized clinical trial of D1 versus D2
resection for gastric cancer. Br J Surg. 2010;97:643-649.
38. Karpeh MS Jr. Should gastric cancer surgery be performed in community hospitals? Semin Oncol. 2005;32:S94-S96.
39. Japanese Gastric Cancer Association. Japanese classification of gastric
carcinoma: 3rd English edition. Gastric Cancer. 2011;14:101-112.
40. Ahn HS, Lee HJ, Yoo MW, et al. Changes in clinicopathological features
and survival after gastrectomy for gastric cancer over a 20-year period. Br J
Surg. 2011;98:255-260.
255
W i l l D i s e a s e H e t e r o g e n e i t y H e l p D e fi n e
Treatment Paradigms for Gastroesophageal
Adenocarcinoma? A Global Perspective
By Manish A. Shah, MD
Overview: Cancers of the upper gastrointestinal (GI) tract

form a heterogeneous group of diseases for which treatment
paradigms for localized disease continue to emerge. Recently,
several phase III studies in esophagus and gastric cancer that
have attempted to define new standards of care have been
reported. However, controversy still persists and treatment
algorithms often depend on individual preference, patient
IFFERENT TREATMENT paradigms characterize upper GI malignancies across the globe. Among the most
varied is the approach to patients with locally advanced, but
resectable, esophagogastric carcinoma. National Comprehensive Cancer Network (NCCN) guidelines for esophageal
carcinoma in patients with localized disease who are medically fit for resection allow for several standard treatment
options, including preoperative chemoradiotherapy (CRT),
definitive CRT, preoperative chemotherapy (for adenocarcinoma), and resection followed by postoperative CRT.1 For
localized, resectable gastric cancer, acceptable treatment
options are nearly equally varied, with the notable exception
of definitive CRT.2 Several important phase III studies were
recently reported at the last two annual ASCO meetings. We
will attempt to place these studies into context of the current
accepted treatment paradigms on the basis of current best
evidence.
Results of Recent Phase III Studies: Gastric Cancer
The management of localized gastric cancer has become

increasingly complicated with emerging data of the survival
advantage of systemic chemotherapy alone, as well as the
potential benefit of CRTin specific settings. Recently, investigators presented the results of the CLASSIC study, a
randomized phase III evaluation of post-operative chemotherapy with capecitabine and oxaliplatin compared with
observation in resected gastric cancer.3 In this study, 1,035
patients with stage II to III gastric cancer who had a D2 (i.e.,
extended lymphadenectomy) surgical dissection were randomly assigned to receive either observation alone (n 515)
or eight cycles of chemotherapy (n 520) consisting of
oxaliplatin 130 mg/m2 plus capecitabine 1,000 mg/m2 twice
daily for 14 days repeated every 3 weeks. The investigators
met their prespecified primary end point of improving 3-year
disease-free survival (DFS), demonstrating 74% 3-year DFS
with adjuvant chemotherapy versus 60% with observation
alone (hazard ratio [HR] 0.56; Table 1). These data are
consistent with the previous adjuvant S1 study reported by
Sakuramoto and colleagues4 and, bolstered by the pooled
individual patient-data metaanalysis,5 further support the
use of postoperative chemotherapy alone. The applicability
of the recent large phase III studies from Korea and Japan
to Western patients remains a question, especially with
markedly different epidemiology of increased proximal and
GEJ tumors in the West compared with Asia. Even when
controlling for the extent of surgical resection, apparent
differences in outcomes persist.6 Furthermore, a recent
256
referral patterns, and treatment biases. In the current era of

improving quality control and standardization of care, such
variations in practice present a substantial challenge for both
patients and physicians. In this article, I will highlight differences in disease biology for upper GI diseases, and in particular, gastric cancer.
well-performed, although underpowered, study examined

postoperative chemotherapy with cisplatin, epirubicin, fluorouracil (FU), and leucovorin (LV; PELF) in a randomized
study and did not show a survival advantage in a European
patient population of resected advanced gastric cancer.7 In
this study, 258 patients were randomly assigned to receive
four cycles of PELF chemotherapy or observation alone, and
with a median follow-up of more than 72 months, no differences in patient outcomes were observed. Specifically, at the
end of the study period, 47% of the patients who received
chemotherapy were still alive compared with 45.3% of the
surgery-alone arm.7 Thus, despite the recent compelling
phase III data, treatment paradigms in the West for locally
advanced gastric/gastroesophageal junction (GEJ) adenocarcinoma currently involve either preoperative or perioperative chemotherapy,8,9 or postoperative CRT.10,11 The role of
postoperative chemotherapy in the West remains controversial.
Regarding CRT after curative-intent resection in gastric
cancer, the addition of epirubicin and cisplatin importantly
did not demonstrate an improvement over standard adjuvant FU/radiotherapy (RT) in the CALGB national phase III
study.11 This was a 546-patient study in which patients
were randomly assigned to receive standard postoperative
CRT (n 280) consisting of bolus FU/LV (cycle 1, 3 and 4),
along with infusional FU (200 mg/m2/day as an intravenous
continuous infusion 5 weeks, cycle 2) or experimental
postoperative CRT (n 266) consisting of epirubicin, cisplatin, and FU (ECF) chemotherapy (cycle 1, 3 and 4), and the
same CRT (infusional FU RT), to demonstrate an improvement in overall survival of 30% with the addition of
ECF. There was no difference in overall survival (HR
1.03), suggesting that epirubicin and cisplatin do not add
significantly to FU adjuvant therapy. Investigators also
noted remarkably little improvement in RT planning over
the last two large United States based phase III studies
(15% major deviations in the RT plan).11 In another study
(the ARTIST Trial), investigators from Korea investigated
From the Weill Cornell Medical College; New York-Presbyterian Hospital, New York, NY.
Address reprint requests to: Manish A. Shah, MD, Weill Cornell Medical College/New
York-Presbyterian Hospital, 1305 York Avenue, 12th Floor, New York, NY 10065; email:
mas9313@med.cornell.edu.
1092-9118/10/1-10
TREATMENT PARADIGMS FOR GASTROESOPHAGEAL ADENOCARCINOMA

Table 1. Recent Phase III Studies in Patients with Localized Esophagogastric Carcinoma
Study
Esophagus cancer
Van Der Gaast
et al13
Year
Disease Type
Subgroup
No. of
Patients
Comparison
CRT (41.4 Gy)

surgery surgery
2010 Esophageal
adenocarcinoma (74%)
and SCC (23%), stage
II/III
Survival
Hazard
Ratio
49 26 mo
0.67
0.49 to 0.91
0.008
31.8 vs.44.5 months*
0.82
0.34
0.92
0.58 to 1.16
0.17 to 0.68
0.63 to 1.35
0.68
74% vs.60%
0.56
0.44 to 0.72
0.0001
540
36.6 vs.37.8 months*
1.03
0.80 to 1.34
0.80
458
396
74% vs.78.2%
72% vs. 77.5%
0.6865
0.47 to 0.995
0.047
0.0365
364
Adeno CA
SCC
Mariette et al18
Gastric cancer
Bang et al3
Fuchs et al11
Lee et al12
2010 Esophageal
adenocarcinoma (30%)
and SCC (70%), stage
I/II
2011 Gastric adenocarcinoma
2011 Gastric and GEJ

adenocarcinoma
2012 Gastric adenocarcinoma
CRT (45 Gy) surgery vs.

surgery alone
Post-op
capecitabine/oxaliplatin
vs. surgery alone
Post-op FU/LV FU/RT vs.
post-op ECF FU/RT
Post-op XP vs. XPRTXP
Node-positive patients
195
1035
95% CI
p value
Abbreviations: ECF, epirubicin, cisplatin, and fluorouracil; FU, fluourouracil; GEJ, gastroesophageal junction; LV, leucovorin; post-op, postoperative; RT,
radiotherapy; SCC, squamous cell carcinoma; XP, capecitabine/cisplatin.
* Overall survival.
3-year disease-free survival.
the role of postoperative RT in a patient population that

underwent a standard D2 gastric dissection.12 In this study,
458 patients were randomly assigned to receive postoperative chemotherapy alone (capecitabine/cisplatin [XP]) or
CRT (XP 3 capecitabine/RT 3 XP), with the aim to identify
a 45% improvement in 3-year DFS. These investigators did
observe a modest improvement in the addition of RT, particularly in the large subset of patients who were node
positive (3-year DFS, 72% vs. 77%; p 0.0365; Table 1).
These data support the role of adjuvant RT in this disease,
as suggested initially by INT-0116, and are currently being
prospectively validated in a node-positive resected cohort.
Results of Recent Phase III Studies:
Esophageal Cancer
In esophageal carcinoma, the CROSS study was presented

at the 2011 ASCO Annual Meeting, and examined the
effects of preoperative CRTin advanced esophageal cancer.
This study found that a combination regimen of CRT before
resection is superior to surgery alone.13 In this multicenter
phase III randomized study, 364 patients in the Netherlands
with resectable esophageal adenocarcinoma or squamous
cell carcinoma (SCC) were randomly assigned to receive
combined-modality therapy of CRTfollowed by surgery or
KEY POINTS
Significant heterogeneity in treatment paradigms for

upper GI malignancies.
Several recent studies have attempted to redefine the
standard of care.
Global disease heterogeneity make broad applicability somewhat questionable.
Our challenge is to recognize differences in disease
biology to optimize treatment paradigms.
surgery alone. Preoperative CRT consisted of weekly paclitaxel 50 mg/m2 and carboplatin dosed at area under the
curve (AUC) 2 for 5 weeks with concurrent 41.4 Gy RT
administered in 23 fractions. After CRT, patients underwent
resection within 6 weeks of completion of preoperative
therapy. This study suggests that most patients with T1N1
or T23Nx esophageal carcinoma should consider preoperative CRT as a standard care option. The median survival of
patients who received CRT and surgery was 49 months,
compared with 26 months for those who received surgery
alone (HR 0.67; p 0.011; Table 1). With a median
follow-up of 32 months, 70 patients had died in the CRT
group compared with 97 in the surgery-alone group, and
3-year overall survival was also superior in the CRT arm.
However, although the majority of patients (74% in both
arms) had adenocarcinoma, it appears that the benefit of
CRT was primarily derived in patients with esophageal
SCC. Patients with esophageal SCC observed an HR of 0.34,
representing a dramatic 66% reduction in risk of death with
preoperative CRT, whereas in the subset of patients with
esophageal adenocarcinoma, the HR for survival in patients
receiving CRT was 0.82 (Table 1).
To place these data into context with other randomized
studies that predominantly included distal esophageal and
GEJ carcinoma as well as a recent updated metaanalysis,9,14-17 if surgery is identified as part of the treatment plan for a patient with localized disease, applying
preoperative therapy does confer a survival advantage. In
addition, esophageal SCC seems to be more sensitive to
CRT. There are data that support either combined preoperative CRT or preoperative chemotherapy alone for
esophageal adenocarcinoma, and the data supporting the
superiority of trimodal therapy (CRT surgery) over bimodal therapy (chemotherapy surgery) remains debatable. A recent study by Stahl and colleagues suggested an
improved survival with trimodality therapy in adenocarcinoma of the esophagus/GEJ, although the study was closed
257
MANISH A. SHAH
18
prematurely due to poor accrual. In addition, Mariette and

colleagues demonstrated that for early-stage esophageal
adenocarcinoma, there appears to be no added benefit of
CRT to surgical resection.19
Influence of Disease Heterogeneity on Treatment
Gastric cancer is a heterogeneous disease and subtypes

of gastric cancer exist.20,21 More than 95% of all cancers of
the stomach are adenocarcinomas. A common distinguishing
feature is histopathology, such as the Laurens classification, which distinguishes intestinal and diffuse gastric cancer subtypes. The well-differentiated intestinal type tends
to expand through the stomach wall, whereas the Laurens
diffuse-type is more commonly poorly differentiated and
spreads as individual discohesive cells in an infiltrative
pattern. Diffuse gastric cancer is associated with loss of the
cell-surface protein E-cadherin and germ-line mutations in
CDH1 are associated with the familial form of diffuse gastric
cancer, hereditary diffuse gastric cancer.22 Intestinal gastric
cancers predominate in high-incidence areas (e.g., China),
and this histology is responsible for much of the ethnic
variation across the globe.
Gastric cancers may have different outcomes depending
on disease subtype. More proximal GEJ and cardia tumors
tend to have a worse prognosis compared with distal pyloric,
antral, and curvature cancers.23 Data are also just now
emerging on the potential influence of disease subtype on
treatment outcome. For example, HER2 amplification and
overexpression is far more prevalent in proximal/GEJ adenocarcinoma than in diffuse gastric cancer.24,25 In an exploratory analysis, these proximal/GEJ tumors appeared to be
less sensitive to bevacizumab therapy than are diffuse and
distal nondiffuse gastric cancers.26 Thus, disease biology
may indeed influence patient outcomes with specific treatments.
Implications for the Future
Where do we go from here? Because treatment paradigms

are defined on a global basis, it will be important to understand the global heterogeneity of these diseases. The epidemiology, risk factors, and patterns of care for cancers of the
upper GI tract are substantially different across the globe.
Disease biology in the various regions around the world may
also be different, although this needs to be more thoroughly
investigated. For example, proximal/GEJ adenocarcinomas
seem less frequent in the far East compared with Europe
and the Americas. However, it is not clear whether this
difference solely contributes to differences in patient outcome between these two regions. Could it be practice patterns (screening and early disease identification), use of
second-line therapy,27 or disease biology6? Our challenge is
to recognize the influence of disease biology and heterogeneity on treatment paradigms in this disease. If we are able to
do so, we will be able to provide better treatment guidelines
for specific disease subtypes, and improve patient outcomes
in a more directed approacha laudable goal, indeed!
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Manish A. Shah*
REFERENCES
1. Ajani JA, Barthel JS, Bentrem D, et al. Esophageal and esophagogastric
junction cancers. National Comp Canc Netw. 2011;2.2011:1-97.
2. Ajani JA, Barthel JS, Bentrem D, et al. Gastric cancer. National Comp
Canc Netw. 2011;2.2011:1-84.
3. Bang YJ, Kim YW, Yang HK, et al. Adjuvant capecitabine and oxaliplatin for gastric cancer after D2 gastrectomy (CLASSIC): a phase 3 open-label,
randomised controlled trial. Lancet. 2012;379:315-321.
for gastric cancer with S-1, an oral fluoropyrimidine. N Engl J Med. Nov 1
2007;357:1810-1820.
5. GASTRIC (Global Advanced/Adjuvant Stomach Tumor Research International Collaboration) Group, Paoletti X, Oba K, et al. Benefit of adjuvant
chemotherapy for resectable gastric cancer: a meta-analysis. JAMA. 2010;
303:1729-1737.
6. Strong VE, Song KY, Park CH, et al. Comparison of gastric cancer
survival following R0 resection in the United States and Korea using an
internationally validated nomogram. Ann Surg. 2010;251:640-646.
7. Di Costanzo F, Gasperoni S, Manzione L, et al. Adjuvant chemotherapy
in completely resected gastric cancer: a randomized phase III trial conducted
by GOIRC. J Natl Cancer Inst. 2008;100:388-398.
8. Cunningham D, Allum WH, Stenning SP, et al. Perioperative chemotherapy versus surgery alone for resectable gastroesophageal cancer. N Engl
J Med. 2006;355:11-20.
9. Ychou M, Boige V, Pignon JP, et al. Perioperative chemotherapy compared with surgery alone for resectable gastroesophageal adenocarcinoma: an
FNCLCC and FFCD Multicenter phase III trial. J Clin Oncol. 2011;29:17151721.
10. MacDonald JS, Smalley S, Benedetti J, et al. Chemoradiotherapy after
258
11. Fuchs CS, Tepper JE, Niedzwiecki D, et al. Postoperative adjuvant

chemoradiation for gastric or gastroesophageal junction (GEJ) adenocarcinoma using epirubicin, cisplatin, and infusional (CI) 5-FU (ECF) before and
after CI 5-FU and radiotherapy (CRT) compared with bolus 5-FV/LV before
and after CRT: Intergroup trial CALGB 80101. J Clin Oncol. 2011;29(suppl;
abstr 4003).
12. Lee J, Lim do H, Kim S, et al. Phase III trial comparing capecitabine
plus cisplatin versus capecitabine plus cisplatin with concurrent capecitabine
radiotherapy in completely resected gastric cancer with D2 lymph node
dissection: the ARTIST Trial. J Clin Oncol. 2012;30:268-273.
13. van der Gaast A, van Hagen P, Hulshof M, et al. Effect of preoperative
concurrent chemoradiotherapy on survival of patients with resectable esophageal or esophagogastric junctino cancer: results from a multi-center randomized phase III study. J Clin Oncol. 2010;28(suppl; abstr 4004).
14. Bedenne L, Michel P, Bouche O, et al. Chemoradiation followed by
surgery compared with chemoradiation alone in squamous cancer of the
esophagus: FFCD 9102. J Clin Oncol. 2007;25:1160-1168.
15. Stahl M, Stuschke M, Lehmann N, et al. Chemoradiation with and
without surgery in patients with locally advanced squamous cell carcinoma of
the esophagus. J Clin Oncol. 2005;23:2310-2317.
16. Tepper J, Krasna MJ, Niedzwiecki D, et al. Phase III trial of trimodality therapy with cisplatin, fluorouracil, radiotherapy, and surgery compared
with surgery alone for esophageal cancer: CALGB 9781. J Clin Oncol.
2008;26:1086-1092.
17. Sjoquist KM, Burmeister BH, Smithers BM, et al. Survival after
neoadjuvant chemotherapy or chemoradiotherapy for resectable oesophageal
carcinoma: an updated meta-analysis. Lancet Oncol. 2011;12:681-692.
18. Stahl M, Walz MK, Stuschke M, et al. Phase III comparison of
preoperative chemotherapy compared with chemoradiotherapy in patients
TREATMENT PARADIGMS FOR GASTROESOPHAGEAL ADENOCARCINOMA

with locally advanced adenocarcinoma of the esophagogastric junction. J Clin
Oncol. 2009;27:851-856.
19. Mariette C, Seitz JF, Mailliard E, et al. Surgery alone versus chemoradiotherapy followed by surgery for localized esophageal cancer: analysis of a
randomized controlled phase III trial FFCD 9901. J Clin Oncol. 2010;
28(suppl; abstr 4005).
20. Shah MA, Kelsen DP. Gastric cancer: a primer on the epidemiology
and biology of the disease and an overview of the medical management of
advanced disease. J Natl Compr Canc Netw. 2010;8:437-447.
21. Shah MA, Khanin R, Tang LH, et al. Molecular classsification of gastric
cancer: a new paradigm. Clin Cancer Res. 2011;17:2693-2701.
22. Guilford P, Hopkins J, Harraway J, et al. E-cadherin germline mutations in familial gastric cancer. Nature. 1998;392:402-405.
23. Catalano V, Labianca R, Beretta GD, et al. Gastric cancer. Crit Rev
Oncol Hematol. 2009;19:127-164.
24. Bang YJ, Van Cutsem E, Feyereislova A, et al. Trastuzumab in
combination with chemotherapy versus chemotherapy alone for treatment

of Her2-positive advanced gastric or gastro-oesophageal junction cancer
(ToGA): a phase 3, open-label, randomised controlled trial. Lancet. 2010
28;376:687-97.
25. Tafe LJ, Janjigian YY, Zaidinski M, et al. Human epidermal growth
factor receptor 2 testing in gastroesophageal cancer: correlation between
immunohistochemistry and fluorescence in situ hybridization. Arch Pathol
Lab Med. 2011;135:1460-1465.
26. Shah MA, Van Cutsem E, Kang YK, et al. Survival analysis according
to disease subtype in AVAGAST: first-line capecitabine and cisplatin plus
bevacizumab (bev) or placebo in patients with advanced gastric cancer. J Clin
Oncol. 2012;30:(suppl; abstr 5).
27. Park SH, Lim DH, Park S, et al. A multicenter, randomized phase III
trial comparing second-line chemotherapy (SLC) plus best supportive care
(BSC) with BSC alone for pretreated advanced gastric cancer (AGC). J Clin
Oncol. 2011;29:(suppl; abstr 4004).
259
Adjuvant Treatments for Localized Advanced

Gastric Cancer: Differences among
Geographic Regions
By Yoon-Koo Kang, MD, PhD, and Changhoon Yoo, MD
Overview: After much debate, adjuvant therapy has become

the standard of care worldwide for resected localized gastric
cancer. However, geographic differences exist in standard
adjuvant treatments: postoperative chemoradiation in North
America, perioperative chemotherapy in the United Kingdom,
and postoperative chemotherapy in East Asia. Now that D2
gastrectomy has been recognized as the optimal surgery for
localized gastric cancer in the West as well as in Asia, the
standard adjuvant treatments used in the West may need to be
reconsidered. One of the most important issues in adjuvant
therapy for localized gastric cancer is how to improve the
clinical outcomes of current standard treatments. Recent
LTHOUGH SURGERY is the only curative treatment

option for patients with localized advanced gastric
cancer, many patients experience recurrence even after
complete resection, leading to poor survival.1 To improve
survival outcomes, many clinical trials have evaluated adjuvant treatments over the decades; however, these studies
have produced conflicting results, mainly because of modest
sample size and problematic study design. Meta-analyses
have consistently described a small but significant survival
benefit associated with adjuvant chemotherapy when
compared to surgery alone.2,3 This finding was recently
confirmed by a large patient-level meta-analysis of 17 randomized controlled trials conducted by the Global Advanced/
Adjuvant Stomach Tumor Research International Collaboration (GASTRIC) group.4 Since multiple phase III studies
with large numbers of patients have demonstrated the
survival benefits of adjuvant treatments in localized gastric
cancer compared with surgery alone, there is now global
agreement that adjuvant therapy improves outcomes in
patients with stage IIIV (M0) gastric cancer undergoing
curative surgical resection.
Current Standard Adjuvant Treatments
Even though adjuvant treatment in localized advanced

gastric cancer has become the standard of care worldwide,
no single regimen has been accepted as the global standard.
In addition, geographical differences still exist in the treatment of resectable gastric cancer (see Table 1). The
Intergroup-0116 study revealed that postoperative chemoradiation consisting of bolus 5-fluorouracil/leucovorin
(5-FU/LV) and concurrent radiotherapy significantly prolonged survival compared with surgery alone. Based on the
results of this trial, adjuvant chemoradiotherapy has been
adopted as the standard adjuvant treatment for curatively
resected gastric cancer in North America.5 Perioperative
chemotherapy is currently the standard practice across
Europe for patients with resectable gastric cancer. This
treatment is based on the results of the Medical Research
Council Adjuvant Gastric Infusional Chemotherapy (MAGIC)
trial6 in the United Kingdom. In this study, the perioperative chemotherapy arm consisting of three preoperative
and three postoperative cycles of epirubicin, cisplatin, and
5-fluorouracil (ECF) demonstrated longer survival than the
Cancer and Leukemia Group B (CALGB) and AMC studies

suggest that simply intensifying chemotherapy by adding
more agents or prolonging treatment duration is insufficient.
However, new strategies like early initiation of chemotherapy
and/or intraperitoneal chemotherapy may further improve the
current standard adjuvant therapy. In the era of targeted
therapy, the role of biologic agents for gastric cancer should
also be explored in the adjuvant setting. With a deeper
understanding of the molecular biology of gastric cancer,
adjuvant therapy for patients with localized gastric cancer can
be optimized and individualized.
surgery alone arm. In Asia, postoperative chemotherapy

with fluoropyrimidine-based regimens has been adopted as
standard adjuvant therapy, based on the results of the
Adjuvant Chemotherapy Trial of TS-1 for Gastric Cancer
(ACTS-GC)7 and the recent CLASSIC trials8 conducted in
Japan and Korea, respectively. These trials demonstrated
improved survival in patients with resected gastric cancer
given S-1 for 1 year or the combination of capecitabine and
oxaliplatin for 6 months, compared with surgery alone.
Why Do Standard Treatments Differ among
Geographic Regions?
Geographic differences in the standard adjuvant therapy

for resectable gastric cancer can be explained primarily by
differences in the standard surgery. Extended (D2) lymph
node dissection is well established as the standard of care
in East Asia, whereas Western surgeons have been skeptical
of the benefit of D2 resection over D1 surgery because of
conflicting results reported by previous randomized trials.
However, the 15-year follow-up results of a Dutch D1D2
trial show that D2 surgery is associated with lower rate of
disease-related death than D1 surgery in Western patients.9
This finding suggests that gastric cancer has been surgically
undertreated in the West, which may explain why radiation (in North America) and intensive perioperative chemotherapy (in Europe) have improved outcomes in Western
countries.
Another potential reason for regional differences is the
heterogeneity of study populations in previous clinical trials
for gastric cancer. The ACTS-GC and CLASSIC trials, which
were conducted in East Asia, included only patients with
gastric cancer. However, Western trials, especially in the
United Kingdom, developed therapeutic strategies for localized gastric cancer by including considerable numbers of
From the Department of Oncology, Asan Medical Center, University of Ulsan College of
Medicine, Seoul, Korea.
Address reprint requests to Yoon-Koo Kang, MD, PhD, Department of Oncology, Asan
Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro 43-gil, Songpa-gu,
Seoul, South Korea 138-736;email: ykkang@amc.seoul.kr.
1092-9118/10/1-10
e31
KANG AND YOO

Table 1. Major Pivotal Phase III Trials for Adjuvant Treatments of Gastric Cancer
Study Name
Total
Patients
Treatment Arms
Patients with
EGJ or Lower
Esophageal Cancer
Patients Who
Underwent
D2 Surgery
Intergroup-0116 (United States)
Surgery alone versus postoperative chemoradiation
556
20%
10%
MAGIC (United Kingdom)
Surgery alone versus perioperative chemotherapy
503
26%
38%
ACTS-GC (Japan)
Surgery alone versus postoperative S-1
1059
0%
100%
CLASSIC (East Asia)
Surgery alone versus postoperative capecitabine and oxaliplatin
1035
0%
100%
Hazard Ratio for

OS (95% CI), p
1.35 (1.091.66)*
p .005
0.75 (0.600.93)
p .009
0.68 (0.520.87)
p .003
0.72 (0.521.00)
p .0493
Abbreviations: EGJ, esophagogastric junction; OS, overall survival; CI, confidence interval.
* Hazard ratio of surgery only group.
Overall survival data are not yet mature; a primary endpoint of this study was disease-free survival.
patients with adenocarcinoma of the esophagogastric junction or lower esophagus, primarily because of the increasing
incidence of these cancers and decreasing incidence of gastric cancer. However, unlike gastric cancer, esophageal cancer tends to easily invade surrounding tissue and regional
lymph nodes because of the lack of serosa and abundance
of lymphatics in the esophagus. Thus, long-term survival
rarely exceeds 20% even after successful resection in advanced disease. For this reason, multimodality therapy
including chemotherapy, radiotherapy, and surgery has
been widely investigated for treatment of esophageal cancer,
and Western trials also used this strategy for treatment of
gastric cancer. Given the significant differences between
esophageal cancer and gastric cancer in terms of etiology,
biology, and clinical characteristics, including patients with
adenocarcinoma of the esophagogastric junction or lower
esophagus in clinical trials for gastric cancer does not seem
appropriate.
The standard adjuvant therapies currently used in the
KEY POINTS
e32
After much debate over the past decades, adjuvant

therapy has become the standard of care worldwide
for resected localized gastric cancer.
However, standard adjuvant treatments vary among
geographic regions: postoperative chemoradiation in
North America, perioperative chemotherapy in the
United Kingdom, and postoperative chemotherapy in
East Asia.
Standard adjuvant treatments in the West may need
to be reconsidered as D2 gastrectomy is now the
standard surgical treatment for localized gastric cancer in the West, as well as in the East.
Recent Cancer and Leukemia Group B and AMC
studies suggest that simply intensifying chemotherapy by adding more agents or prolonging treatment
duration is insufficient. However, new strategies like
early initiation of chemotherapy and/or intraperitoneal chemotherapy may further improve the current
standard adjuvant therapy.
The role of targeted agents proven effective in metastatic or recurrent disease should be explored in the
adjuvant setting.
West were established before D2 gastrectomy became the

standard surgery. Although the superiority of D2 surgery
over D0/1 surgery has been consistently demonstrated, it
will take considerable time and effort before D2 surgery is
widely performed in the West, because of the lack of expertise in this procedure and insufficient number of patients to
implement new surgical technique. Nevertheless, we argue
that the standard adjuvant treatments used in the West
need to be reconsidered, because the D2 gastrectomy has
finally become the standard surgery for localized gastric
cancer in the West as well as in the East. Adjuvant therapy
should be determined according to whether the patient
undergoes optimal surgery (D2) or suboptimal surgery (D0/
1), not according to where the patient is treated. Despite the
success of the Intergroup-0116 trial, the role of radiotherapy
was seldom investigated in countries where D2 gastrectomy
is the standard of surgery, because many investigators are
reluctant to add another local therapy to an optimal
surgery (extended lymph node dissection). Based on the
positive results of a retrospective study, the ARTIST trial,10
which evaluated adjuvant chemotherapy with or without
radiation, was conducted in Korea. In contrast to the
Intergroup-0116 study, the control arm in the ARTIST trial
underwent chemotherapy rather than observation, and D2
surgery was mandatory. However, this study failed to show
that adding radiation to adjuvant chemotherapy improved
outcomes for patients who underwent D2 gastrectomy. Although the inclusion of too many patients with early-stage
cancer (approximately 60% in stage IB or II) may have
limited the power of the study, the negative results in the
ARTIST trial suggest that radiation does not improve the
efficacy of adjuvant chemotherapy following optimal surgery. The results of the latest trials strongly suggest that
all patients with localized gastric cancer should undergo
D2 surgery, if technically feasible, and subsequently undergo adjuvant chemotherapy, regardless of ethnicity or
geographic location.
Future Perspective: How to Improve the Current
Standard Adjuvant Treatment?
The most important issue in adjuvant therapy for localized gastric cancer is how to improve the clinical outcomes of
current standard treatments. The results of the following
studies offer timely suggestions. Cancer and Leukemia
Group B (CALGB) 80101 study in North America and the
AMC 0201 study in Korea failed to demonstrate that intensification of adjuvant chemotherapy improves outcomes. The
CALGB 80101 trial11 intensified the regimen used in the
ADJUVANT TREATMENT FOR GASTRIC CANCER
Intergroup-0116 trial by adding two more drugs (epirubicin

and cisplatin) to the bolus 5-FU/LV before and after 5-FU/
radiotherapy for resected gastric or esophagogastric junction
adenocarcinoma. The AMC 0201 study12 increased the duration of oral fluoropyrimidine treatment (12 months) and
added cisplatin to the combination of mitomycin-C and 3
months of oral fluoropyrimidine; the control regimen was
based on prolonged survival of patients with resected stage
III gastric cancer compared with surgery alone in a Spanish
phase III study.13 However, both phase III trials failed to
show increased survival, suggesting that simply intensifying
the adjuvant chemotherapy (with or without radiation) by
adding an agent or prolonging treatment duration does not
always enhance its efficacy in patients with localized gastric
cancer. However, the AMC 0101 study,14 a companion trial
of AMC 0201, evaluated the efficacy of two more strategies
in patients with D2-resected macroscopically serosa-positive
gastric cancer than the AMC 0201 with same control arm:
intraoperative intraperitoneal chemotherapy using cisplatin
and early initiation (day after surgery) of systemic chemotherapy. In this phase III study, which included 521 patients, these strategies significantly improved recurrencefree survival and overall survival compared with the control
regimen (mitomycin-C and 3 months of oral fluoropyrimidine); this finding was verified by follow-up data (median
6.6 years).15 In light of the negative results of AMC 0201,
improved survival in AMC 0101 is attributable to early
initiation of systemic chemotherapy and/or intraperitoneal
cisplatin. These two strategies may enhance the efficacy of
current standard regimens for gastric cancer such as postoperative chemoradiation and perioperative chemotherapy
as well as postoperative chemotherapy. Regarding the potential benefits of early initiation of systemic chemotherapy
in localized gastric cancer, neoadjuvant chemotherapy is, in
a sense, the earliest possible adjuvant chemotherapy. Despite poor compliance in the postoperative phase of the
MAGIC trial, the successful outcome associated with perioperative chemotherapy also suggests the potential benefits
of neoadjuvant chemotherapy. However, in the MAGIC trial,
perioperative chemotherapy was compared with surgery
alone6; therefore, it is not clear whether improved survival
was because of preoperative chemotherapy, postoperative
chemotherapy, or both. Furthermore, only 40% of the patients in the MAGIC trial underwent D2 surgery. Therefore,
the efficacy of neoadjuvant chemotherapy in countries where
D2 surgery and postoperative chemotherapy is the standard
of care remains to be determined. The PRODIGY trial
(preoperative docetaxel, oxaliplatin, and S-1 followed by
postoperative S-1 versus postoperative S-1 for patients with
D2 resection; NCT01515748) aims to answer this question.
In the era of targeted therapy in oncology, biologic agents
have also been investigated for adjuvant treatment of gas-
tric cancer. Based on the success of the MAGIC trial in the

United Kingdom, MAGIC-B trial (MRC-ST03) is ongoing to
determine the efficacy of adding bevacizumab to adjuvant
therapy (perioperative epirubicin, capecitabine, and cisplatin with or without bevacizumab), and recently reported the
feasibility of this regimen. The recent AVAGAST trial16
(conventional chemotherapy with or without bevacizumab)
failed to demonstrate the benefit of bevacizumab in a global
patient population with locally advanced and metastatic
gastric cancer. However, it will be interesting to see whether
adding bevacizumab benefits Western patients with locally
advanced esophagogastric cancer. With the success of the
ToGA trial,17 the door to targeted therapy for gastric cancer
has finally opened. In locally advanced or metastatic gastric
cancer with overexpressed human epidermal growth factor
receptor (HER)-2, trastuzumab in combination with chemotherapy significantly prolonged survival outcomes and is
therefore considered the new standard of care. The efficacy
of adjuvant trastuzumab in HER2-positive breast cancer led
us to expect benefit from adjuvant trastuzumab in HER2positive gastric cancer, although prognostic impact of HER2
expression in resectable gastric cancer was shown to be
limited.18 Multiple biologic agents are currently under investigation for use in gastric cancer, especially in metastatic
or recurrent disease. If the results are promising, these
agents should also be explored in the adjuvant setting. In
the future, we should work to improve our understanding of
the molecular biology of gastric cancer in order to provide
optimized and individualized therapy for patients with localized gastric cancer.
Conclusion
After much debate over the past decades, adjuvant therapy has become the standard of care worldwide for resected
localized gastric cancer. However, geographic differences
exist in standard adjuvant treatments: postoperative chemoradiation in North America, perioperative chemotherapy in
the United Kingdom, and postoperative chemotherapy in
East Asia. Standard adjuvant treatments in the West may
need to be reconsidered as D2 gastrectomy has finally
become the standard surgery for localized gastric cancer in
the West, as well as in the East. Results of the recent
CALGB and AMC studies suggest that simply intensifying
chemotherapy by adding more cytotoxic agents or prolonging duration of treatment is insufficient. Instead, new strategies such as early initiation of chemotherapy and/or
intraperitoneal chemotherapy may further improve the current standard adjuvant therapy. The role of targeted agents
in adjuvant treatment for localized gastric cancer should be
investigated in future based on the experiences in recurrent
or metastatic disease.
Author
Yoon-Koo Kang
Employment or
Leadership
Positions
Consultant or
Advisory Role
Roche
Stock
Ownership
Honoraria
Taiho
Pharmaceuticals;
Roche
Research
Funding
Jeil
Pharmaceutical;
Roche
Expert
Testimony
Other
Remuneration
Changhoon Yoo*
e33
KANG AND YOO
REFERENCES
1. DAngelica M, Gonen M, Brennan MF, et al. Patterns of initial recurrence in completely resected gastric adenocarcinoma. Ann Surg. 2004;240:
808-816.
2. Earle CC, Maroun JA. Adjuvant chemotherapy after curative resection
for gastric cancer in non-Asian patients: Revisiting a meta-analysis of
randomised trials. Eur J Cancer. 1999;35:1059-1064.
3. Nakajima T, Ota K, Ishihara S, et al. [Meta-analysis of 10 postoperative
adjuvant chemotherapies for gastric cancer in CIH]. Gan To Kagaku Ryoho.
1994;21:1800-1805.
4. Paoletti X, Oba K, Burzykowski T, et al. Benefit of adjuvant chemotherapy for resectable gastric cancer: A meta-analysis. JAMA. 2010;303:17291737.
5. Macdonald JS, Smalley SR, Benedetti J, et al. Chemoradiotherapy after
6. Cunningham D, Allum WH, Stenning SP, et al. Perioperative chemotherapy versus surgery alone for resectable gastroesophageal cancer. N Engl
J Med. 2006;355:11-20.
for gastric cancer with S-1, an oral fluoropyrimidine. N Engl J Med. 2007;
357:1810-1820.
8. Bang YJ, Kim YW, Yang HK, et al. Adjuvant capecitabine and oxaliplatin for gastric cancer after D2 gastrectomy (CLASSIC): a phase 3 open-label,
randomised controlled trial. Lancet. 2012;379:315-321.
9. Songun I, Putter H, Kranenbarg EM, et al. Surgical treatment of gastric
cancer: 15-year follow-up results of the randomised nationwide Dutch D1D2
trial. Lancet Oncol. 2010;11:439-449.
10. Lee J, Lim DH, Kim S, et al. Phase III Trial Comparing Capecitabine
Plus Cisplatin Versus Capecitabine Plus Cisplatin With Concurrent Capecitabine Radiotherapy in Completely Resected Gastric Cancer With D2 Lymph
Node Dissection: The ARTIST Trial. J Clin Oncol. 2012;30:268-273.
11. Fuchs CS, Tepper JE, Niedzwiecki D, et al. Postoperative adjuvant
chemoradiation for gastric or gastroesophageal junction (GEJ) adenocarcinoma using epirubicin, cisplatin, and infusional (CI) 5-FU (ECF) before and
after CI 5-FU and radiotherapy (CRT) compared with bolus 5-FU/LV before
e34
and after CRT: Intergroup trial CALGB 80101. J Clin Oncol. 2011;29(suppl;
abstr 4003).
12. Chang H-M, Kang Y-K, Min YJ, et al. A randomized phase III trial
comparing mitomycin-C plus short-term doxifluridine (Mf) versus
mitomycin-C plus long-term doxifluridine plus cisplatin (MFP) after curative
resection of advanced gastric cancer (AMC 0201) (NCT00296335). J Clin
Oncol. 2008;26(suppl; abstr 4531).
13. Cirera L, Balil A, Batiste-Alentorn E, et al. Randomized clinical trial of
adjuvant mitomycin plus tegafur in patients with resected stage III gastric
14. Kang Y-K, Change H-M, Zang DY, et al. Postoperative adjuvant
chemotherapy for grossly serosa-positive advanced gastric cancer: A randomized phase III trial of intraperitoneal cisplatin and early mitomycin-C plus
long-term doxifluridine puls cisplatin (iceMFP) versus mitomycin-C plus
short-term doxifluridine (Mf) (AMC0101) (NCT00296322). J Clin Oncol.
2008;28(suppl; abstr LBA4511).
15. Kang Y-K, Ryoo B-Y, Chang H-M, et al. Update of AMC 0101 study: A
phase III trial of intraperitoneal cisplatin and early mitomycin-C plus
long-term doxifluridine plus cisplatin (iceMFP) versus mitomycin-C plus
short-term doxifluridine (Mf) as adjuvant chemotherapy for grossly serosapositive advanced gastric cancer (NCT00296322). J Clin Oncol. 2012;30(suppl
4; abstr 4).
16. Ohtsu A, Shah MA, Van Cutsem E, et al. Bevacizumab in combination
with chemotherapy as first-line therapy in advanced gastric cancer: A
randomized, double-blind, placebo-controlled phase III study. J Clin Oncol.
2011;29:3968-3976.
17. Bang YJ, Van Cutsem E, Feyereislova A, et al. Trastuzumab in
combination with chemotherapy versus chemotherapy alone for treatment of
HER2-positive advanced gastric or gastro-oesophageal junction cancer
(ToGA): a phase 3, open-label, randomised controlled trial. Lancet. 2010;376:
687-697.
18. Kataoka Y, Okabe H, Yoshizawa A, et al. HER2 expression and its
clinicopathological features in resectable gastric cancer. Gastric Cancer. Epub
2012 Mar 14.
LIVER-DIRECTED THERAPEUTIC OPTIONS FOR

HEPATOCELLULAR CARCINOMA: PATIENT
SELECTION AND CLINICAL OUTCOMES
CHAIR
Laura A. Dawson, MD
Toronto, ON, Canada
SPEAKERS
Kenneth K. Tanabe, MD
Massachusetts General Hospital Cancer Center
Boston, MA
Fred T. Lee, MD
Madison, WI
Stereotactic Body Radiation Therapy for

Hepatocellular Carcinoma
By Laura A. Dawson, MD, Sameh Hashem, MD, and Alexis Bujold, MD
Overview: Stereotactic body radiotherapy (SBRT), in which

highly conformal potent radiation doses are delivered in fewer
fractions than traditional radiation therapy (RT), is an increasingly popular treatment for hepatocellular carcinoma (HCC).
The great majority of HCCs smaller than 6 cm and with
Child-Pugh A liver function are controlled with SBRT with
limited toxicity. Long-term local control is reduced in larger
LOBALLY, HCC IS THE sixth most common cancer

and the fourth most common cause of cancer-related
death. The overall 5-year survival is poor (approximately
5%), and its incidence is increasing.1 While resection and
transplant can cure HCC, only a minority of patients are
suitable for surgery because of multifocal or extrahepatic
cancer, inadequate liver function, and/or involvement of
large vessels. Radiofrequency ablation and percutaneous
ethanol injection are associated with excellent local control
in small HCCs, but outcomes are reduced in HCCs larger
than 4 cm or adjacent to large vessels.
RT is an effective local therapy that has the potential to
benefit patients unsuitable for and/or at high risk of complications following standard local-regional therapies. All of
the following have facilitated the safe delivery of tumorcidal
doses to focal HCCs using conformal RT: advances in imaging, RT planning techniques (to produce three-dimensional
conformal RT plans minimizing dose to surrounding tissues), image-guided radiotherapy (IGRT; to localize the
tumor at time of treatment), tumor immobilization (to account for breathing-related organ motion), and improved
knowledge of what volume of liver is required to be spared
from radiation to preserve function. Protons and carbon
ionsavailable at specialized centers have the ability to
spare more liver parenchyma than photons.
SBRTwhich is shorter radiotherapy schedules (hypofractionation), with higher very conformal doses delivered at
each radiation fraction has more recently been used to
treat focal HCC. This has been used when the majority of the
liver can be spared from irradiation. Initially stereotactic
immobilization body frames were used to aid in patient
SBRT positioning, but with recent advancements in IGRT,
the requirement for such frames has disappeared, as the
liver can be directly visualized before or during RT deliverly.
SBRT is widely available (unlike protons or carbon ions) and
more convenient for patients than conventionally fractionated RT, as it is delivered in far fewer fractions (typically 10
or less) than standard fractionated RT. This article will
focus primarily on SBRT for HCC.
tumors, and toxicity is increased in patients with Child-Pugh B

or C liver function. SBRT is an effective treatment for tumor
vascular thrombi and can lead to sustained vascular recanalization. The first site of recurrence following SBRT is most
often within the liver, away from the high dose volume,
providing rationale for combining SBRT with regional or systemic therapies. Randomized trials of SBRT are warranted.
The classical described toxicity following RT for liver

cancer is a clinical syndrome of anicteric hepatomegaly,
ascites, and elevated liver enzymes (particularly serum
alkaline phosphatase) occurring within three months after
RT, referred to as classic radiation-induced liver disease
(RILD).2 For the most part, this toxicity is preventable, as
long as a sufficient fraction of liver can be spared from RT.
For example, the risk of classic RILD is less than 5% when
the mean liver doses are kept less than 14 Gy (in 6 fractions)
and 28 Gy (in 1.5 Gy fractions).
Other hepatic toxicities, referred to as nonclassic RILD
are more challenging to prevent. Such toxicities include
reactivation of viral hepatitis, elevation of liver enzymes,
and a general decline in liver function. The partial volume
tolerance of the liver following RT is less clearly defined for
nonclassic RILD, with different dose-volume tolerances and
risk factors observed for different types of liver toxicity. In
Taiwan, 17 of 89 patients with HCC treated with up to 66 Gy
(in 1.8 3 Gy per fraction) developed liver toxicity. The risk
was increased in hepatitis B carriers and in Child-Pugh B
liver function.3 Treatment of viral hepatitis B is important
before initiating RT. Other studies have shown that the risk
of toxicity is increased if the spared volume of liver is too
small. For example, in a study of 48 patients with HCC
treated with three-fraction SBRT (30 39 Gy), 11% of patients had a decline in Child-Pugh class, and this was more
likely if less than 800 mL of liver could be spared from 18 Gy
or more.4
Outcomes: Non-SBRT RT
Liver Tolerance of RT
The first experience in RT for HCC was with hyper- or

conventional fractionation, using conformal RT. The median
survival of patients with locally advanced HCC treated with
conventional fractionation ranges from 6 to 19 months.5-9 A
majority of series demonstrated that patients treated with
higher RT doses had better local control and survival than
those treated with lower doses.
Protons or carbon ions produced from highly specialized
treatment units not widely available have also been used to
treat HCC. An advantage of protons is that they are associ-
Just as a portion of the liver may be resected with surgery

or ablated with radiofrequency ablation, portions of the liver
can tolerate high doses of radiation, without liver toxicity.
With long-term follow-up, atrophy of the irradiated portion
of the liver and hypertrophy of the spared portions of the
liver are commonly seen. Objective response assessment is
challenging within 4 months following RT because of
changes in the irradiated liver volume during that time.
From the Department of Radiation Oncology, Princess Margaret Hospital, University of

Toronto, Toronto, Ontario, Canada, and the Departement de Radio-oncologie CliniqueEnseignement-Recherche, Hopital Maisonneuve-Rosemont, Montreal, Quebec, Canada.
Address reprint requests to Laura Dawson, MD, 610 University Ave., Toronto, ON, M5G
2M9, Canada; email: laura.dawson@rmp.uhn.on.ca.
1092-9118/10/1-10
261
DAWSON, HASHEM, AND BUJOLD

Table 1. Selected Outcomes from HCC SBRT Series
Number of Patients
Blomgren, 1998
Choi, 200623
14
Mendez Romero, 200617

Tse, 200818
Louis, 201026
Kwon, 201024
Facciuto, 201122
Andolino, 201120
Dose/Fraction
Median Follow-up
(Months)
Tumor Size
9 pts HCC, 1 pt IHC

20 pts HCC
515 Gy/13 #
50 Gy/510 #
NR
23
NR
3.8 cm (26.5 cm)
11
CP
31
All
25
CP
42
CP
27
All
60
CP
25 Gy/5 #
30 37.5 Gy/3 #
36 Gy (median)/6 #
Range 2454 Gy
45 Gy/3 #
NR
pts HCC
A and B
pts HCC
CP A
pts HCC
A and B
pts HCC
A 90%
pts HCC
CP A
pts HCC
A/B: 36/24
Overall
Response Rate
Survival
7 cm
70%
80% overall
20% CR/60% PR
1-yr LC: 82%
NR
OS 1 yr: 70%
OS 2 yr: 43.1%
OS 1 yr: 75%
18
Median: 173 cm3
1-yr infield LC: 65%
OS 1 yr: 48%
13
Median: 150 cm3
86% overall
3039 Gy/3#
29
15.4 cm3 (382 cm3)
2436/24 #
22
2.0 cm 0.8 cm
Overall: 86%
3-yr LC: 68%
Overall: 37%
OS 1 yr: 79%
OS 2 yr: 52%
OS 3 yr: 59%
44 Gy (median)/3 # CP A
40 Gy (median)/5 # CP B
27
Median: 3 cm
Overall: 90%
OS 2 yr: 82%
OS 2 yr: 67%
Abbreviations: HCC, hepatocellular carcinoma; SBRT, stereotactic body radiotherapy; Pts, patients; IHC, immunohistochemical; #, fractions; NR, not reported; OS,
overall survival; CR, complete response; PR, partial response; LC, local control; CP, Child-Pugh.
ated with unique dose distributions that allow rapid falloff

in dose and substantial sparing of dose to tissues adjacent to
tumors. Proton or carbon RT has the best reported outcomes
postradiation in patients with HCC.10,11 In one prospective
study, patients with Child-Pugh A liver disease and potentially resectable single HCCs, had a 5-year survival of 56%
following proton therapy.10 In another series, in six patients
with HCC who went on to have liver transplantation 6 to
18 months after proton therapy (63 Gy equivalent in 15
fractions), two complete pathologic responses were observed,
demonstrating proof that high-dose RT may ablate HCC.12
Proton and photon therapy have been used to successfully treat HCC with portal vein or inferior vena cava
thrombosis.11
Outcomes: SBRT
More recently, SBRT has been used to treat patients with

HCC, with a summary of outcomes shown in Table 1.
Blomgren et al first reported on the use of SBRT on extracranial sites in 1995 and 1998.13,14 In this series that included
patients with HCC, 15 to 45 Gy was delivered over one to
five fractions. The majority of patients had objective responses. Subsequently, Herfarth et al treated 60 liver tumors (which included 3 primary HCCs) in 37 patients with
KEY POINTS
262
Stereotactic body radiation therapy (SBRT) highly

conformal, potent-dose radiation therapy delivered in
fewer fractions than usual (usually 10)is being
used more commonly in hepatocellular carcinoma
(HCC).
SBRT is most effective in HCC smaller than 6 cm,
with local control rates from 70% to 95% at 2 years.
SBRT can also lead to sustained control of HCCs
larger than 6 cm and the recanalization of vascular
tumor thrombi from HCC.
Toxicity risks are increased in patients with ChildPugh B and C baseline liver function.
Randomized trials of SBRT for HCC are required.
14 to 26 Gy SBRT in one fraction. The treatment was well

tolerated.15 Wulf et al then reported on 39 patients treated
with three-fraction SBRT to the liver, five of whom had
HCC. No more than 50% of the total functional liver tissue
could receive more than 5 Gy, and no more than 30% could
receive more than 7 Gy. The doses delivered ranged from 30
Gy to 37.5 Gy in three fractions. No local failures were seen
in two HCC patients after 15 and 48 months. Three of the
HCC patients died of disease progression in the liver outside
the RT field at 2, 7, and 17 months. No acute or late serious
toxicity was reported in any patient.16
Mendez Romero et al treated eight patients with 11 HCCs
of Child-Pugh A or B (in addition to 17 patients with liver
metastases) with 25 Gy in five fractions, 30 Gy in three
fractions, or 37.5 Gy in three fractions over 5 to 10 days. The
maximum tumor size was 7 cm. Two of the patients with
HCC who received 25 Gy in five fractions failed locally at 4
and 7 months and were retreated with 24 Gy in three
fractions. The local control rate of the patients with HCC
was reported as 82%, and the 1- and 2-year actuarial
survival rates were 75% and 40%, respectively. One patient
in the HCC group with Child-Pugh B liver function developed grade 5 toxicity because of liver failure and infection.17
In Toronto, a prospective dose-escalation study of sixfraction SBRT was conducted in 31 patients with locally
advanced HCC unsuitable for standard therapies. A typical
plan is shown in Fig. 1. The dose per fraction was determined based on the effective volume of normal liver irradiated (Veff). When the effective liver volume irradiated was
low (Veff 25%), doses of 54 Gy (9 Gy 6) were delivered
safely to HCCs, with excellent local control. For patients
requiring moderate volume liver irradiation (Veff 25% to
80%), doses from 24 to 54 Gy (4 to 9 Gy 6) were delivered
safely. All patients had Child-Pugh liver function A, and the
majority of patients had portal vein thrombosis. No classic
RILD was seen. Five patients had a decline in their ChildPugh class at 3 months. These patients had extensive
tumors, and three had tumor progression that likely contributed to the decline in liver function. The median survival
was 11.7 months (95% CI, 9.221.6 months).18 One hundred
and two eligible patients were included in an update of the
Toronto phase I study and a subsequent phase II study,
using the same dose-allocation approach. Underlying liver
SBRT FOR HEPATOCELLULAR CARCINOMA
Fig. 1. Axial slice of HCC SBRT plan showing a

conformal isodose distribution. The red and blue
solid lines represent the gross target volume (GTV)
and planning target volume (PTV) to be irradiated,
to account for uncertainties, respectively. The HCC
was treated with 42 Gy in six fractions (pink line),
which tightly surrounds the PTV, with a rapid
falloff in dose.
Abbreviations: HCC, hepatocellular carcinoma;
SBRT, stereotactic body radiotherapy.
disease included hepatitis B, 39%; hepatitis C, 40%; and

alcohol, 25%. Prior therapies were delivered in 50% of
patients. Baseline Barcelona Clinic Liver Cancer stage was
class C in 66%. Multiple lesions were present in 59% of
patients, and the median sum of liver lesion diameters was
10 cm (1.8 43 cm). Tumor vascular thrombosis was present
in 55% and extrahepatic disease in 14%. At one year, the
local control of irradiated HCC was 79% (CI 95%, 66 87%).
Median overall survival was 17.0 months (CI 95%, 10.6 21.8
months).19
The Indiana University School of Medicine has also conducted prospective studies of HCC SBRT.20,21 In a phase I
study by Cardenes et al, in 17 patients with HCC with 25
lesions (median diameter 3.2 cm), dose was escalated in
patients with Child-Pugh A from 36 Gy in three fractions to
48 Gy in three fractions, without dose-limiting toxicity. In
patients with Child-Pugh B, two developed grade 3 hepatic
toxicity at 42 Gy in three fractions. Subsequently, five
patients with Child-Pugh B were treated with 40 Gy in five
fractions, and one patient developed liver failure.21 Overall,
20% of patients experienced an increase in Child-Pugh class:
seven of 36 patients with Child-Pugh A experienced progression to Child-Pugh B, and five of 24 patients with ChildPugh B progressed to Child-Pugh C, demonstrating an
increased risk of any toxicity in patients with worse ChildPugh class at baseline. After a median follow-up of 27
months, 2-year local control was 90%, progression-free survival was 48%, and overall survival was 67%.
Retrospective series of SBRT have also been published.22,23 Kwon et al treated 42 patients with HCC with 30
to 39 Gy in three fractions. With a median follow-up of 29
months, the response rate was 86% (60% complete response
and 26% partial response).24 Seo et al treated 38 patients
with HCC with 33 to 57 Gy in three to four fractions, with a
61% 2-year survival and 79% local control rate.25 Doses of
more than 42 Gy in three fractions were associated with
improved local control. Twenty-five European patients with
HCC have also been treated with SBRT (45 Gy in 3 fractions), with a 1-year local control rate of 95%.26
SBRT has been used to effectively treat HCC tumor
thrombi and has been used as a bridge to liver transplant.21,27,28 In most series, following SBRT, the first site of
recurrence is in the liver outside the irradiated volume,
providing rationale for studies combining regional or systemic therapies with SBRT.
Toxicity: SBRT
In addition to the potential for liver toxicity (including

classic and nonclassic RILD), as previously described, SBRT
is associated with the possibility for other late toxicity.
Gastric, duodenum, and small and large bowel late toxicity
(e.g., ulcer, fistula, bleed) are more likely to occur following
SBRT potent doses than conventional radiation fractionations. Therefore, HCC tumors best suited for SBRT are
located at least 1 cm away from luminal gastrointestinal
structures. Proton pump inhibitors may reduce the risk of
luminal gastrointestinal toxicity, and strategies to move
gastrointestinal structures away from the tumor itself may
be beneficial to patients.
A potential toxicity that is unique to SBRT is chest pain
and rib fracture. This has not commonly been reported but is
a possible late sequelae from therapy for peripherally located HCCs. The risk of toxicities can be reduced by ensuring that radiation hot spots are within the target volume
and not near adjacent critical normal tissues.
Another potential toxicity is biliary. For caudate lesions
and HCC invading the biliary track, edema may occur
following SBRT, so care is required to stent the patient
before therapy and/or to premedicate patients with steroids
to reduce this risk. In the long term, there is a risk of late
biliary stenosis, which has been rarely reported following
proton therapy and not yet reported following SBRT. Nonetheless, being cautious to reduce the dose per fraction and/or
overall maximal dose is reasonable for caudate lesions to
reduce the risk of this potential late toxicity.
Conclusion
Technical advances in radiation oncology have made it

possible for SBRT to be used safely for the treatment of
263
DAWSON, HASHEM, AND BUJOLD
HCC, with encouraging outcomes in early and locally

advanced HCC. The risk of liver toxicity is increased
in patients with Child-Pugh B or C cirrhosis, hepatitis B
carrier status, and large tumor size relative to the
liver. Improved tumor control and survival are seen in
patients who can be treated with higher doses. Randomized
trials are required to better understand the benefits

and toxicities of SBRT. An international phase III study
(RTOG 1112, in development) of sorafenib versus SBRT
followed by sorafenib in locally advanced HCC should
provide some insight into the benefits of SBRT in this
setting.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Laura A. Dawson
Research
Funding
Expert
Testimony
Other
Remuneration
Bayer
Sameh Hashem*
Alexis Bujold*
REFERENCES
2010;60:277-300.
2. Pan CC, Kavanagh BD, Dawson LA, et al. Radiation-associated liver
injury. Int J Radiat Oncol Biol Phys. 2010;76:S94-S100.
3. Cheng JC, Liu HS, Wu JK, et al. Inclusion of biological factors in
parallel-architecture normal-tissue complication probability model for
radiation-induced liver disease. Int J Radiat Oncol Biol Phys. 2005;62:11501156.
4. Son SH, Choi BO, Ryu MR, et al. Stereotactic body radiotherapy for
patients with unresectable primary hepatocellular carcinoma: Dosevolumetric parameters predicting the hepatic complication. Int J Radiat
Oncol Biol Phys. 2010;78:1073-1080.
5. Ben-Josef E, Normolle D, Ensminger WD, et al. Phase II trial of
high-dose conformal radiation therapy with concurrent hepatic artery floxuridine for unresectable intrahepatic malignancies. J Clin Oncol. 2005;23:
8739-8747.
6. Mornex F, Girard N, Beziat C, et al. Feasibility and efficacy of high-dose
three-dimensional radiotherapy in cirrhotic patients with small-size hepatocellular carcinoma non-eligible for curative therapiesmature results of the
French phase II RTF-1 trial. Int J Radiat Oncol Biol Phys 2006;66:1152-1158.
7. Seong J, Shim SJ, Lee IJ, et al. Evaluation of the prognostic value of
Okuda, Cancer of the Liver Italian Program, and Japan Integrated Staging
systems for hepatocellular carcinoma patients undergoing radiotherapy. Int J
8. Seong J, Park HC, Han KH, et al. Clinical results of 3-dimensional
conformal radiotherapy combined with transarterial chemoembolization for
hepatocellular carcinoma in the cirrhotic patients. Hepatol Res. 2003;27:3035.
9. McIntosh A, Hagspiel KD, Al-Osaimi AM, et al. Accelerated treatment
using intensity-modulated radiation therapy plus concurrent capecitabine for
unresectable hepatocellular carcinoma. Cancer. 2009;115:5117-5125.
10. Fukumitsu N, Sugahara S, Nakayama H, et al. A prospective study of
hypofractionated proton beam therapy for patients with hepatocellular carcinoma. Int J Radiat Oncol Biol Phys. 2009;74:831-836.
11. Mizumoto M, Tokuuye K, Sugahara S, et al. Proton beam therapy for
hepatocellular carcinoma with inferior vena cava tumor thrombus: Report of
three cases. Jpn J Clin Oncol. 2007;37:459-462.
12. Bush DA, Hillebrand DJ, Slater JM, et al. High-dose proton beam
radiotherapy of hepatocellular carcinoma: Preliminary results of a phase II
trial. Gastroenterology. 2004;127:S189-S193.
13. Blomgren H, Lax I, Naslund I, et al. Stereotactic high dose fraction
radiation therapy of extracranial tumors using an accelerator. Clinical
experience of the first thirty-one patients. Acta Oncol. 1995;34:861-870.
14. Blomgren H, Lax I, Goranson II, et al. Radiosurgery for tumors in the
body: Clinical experience using a new method. J Radiosurg. 1998;1:63-74.
264
15. Herfarth KK, Debus J, Lohr F, et al. Stereotactic single-dose radiation

therapy of liver tumors: Results of a phase I/II trial. J Clin Oncol. 2001;19:
164-170.
16. Wulf J, Guckenberger M, Haedinger U, et al. Stereotactic radiotherapy
of primary liver cancer and hepatic metastases. Acta Oncologica. 2006;45:
838-847.
17. Mendez Romero A, Wunderink W, Hussain SM, et al. Stereotactic body
radiation therapy for primary and metastatic liver tumors: A single institution phase I-II study. Acta Oncologica. 2006;45:831-837.
18. Tse R, Hawkins M, Lockwood G, et al. Phase I study of individualized
stereotactic body radiotherapy for hepatocellular carcinoma and intrahepatic
cholangiocarcinoma. J Clin Oncol. 2008;26:657-664.
19. Bujold A, Massey C, Kim JJ, et al. Outcomes of stereotactic body
radiotherapy (SBRT) for hepatocellular carcinoma (HCC). Int J Radiat Oncol
Biol Phys. 2011;81:S70-S71.
20. Andolino DL, Johnson CS, Maluccio M, et al. Stereotactic body radiotherapy for primary hepatocellular carcinoma. Int J Radiat Oncol Biol Phys.
2011;81:447-453.
21. Cardenes HR, Price TR, Perkins SM, et al. Phase I feasibility trial of
stereotactic body radiation therapy for primary hepatocellular carcinoma.
Clin Transl Oncol. 2010;12:218-225.
22. Facciuto ME, Singh MK, Rochon C, et al. Stereotactic body radiation
therapy in hepatocellular carcinoma and cirrhosis: evaluation of radiological
and pathological response. J Surg Oncol. Epub 2011 Sep 29.
23. Choi BO, Jang HS, Kang KM, et al. Fractionated stereotactic radiotherapy in patients with primary hepatocellular carcinoma. Jpn J Clin Oncol.
2006;36:154-158.
24. Kwon JH, Bae SH, Kim JY, et al. Long-term effect of stereotactic body
radiation therapy for primary hepatocellular carcinoma ineligible for local
ablation therapy or surgical resection. Stereotactic radiotherapy for liver
cancer. BMC Cancer. 2010;10:475.
25. Seo YS, Kim MS, Yoo SY, et al. Preliminary result of stereotactic body
radiotherapy as a local salvage treatment for inoperable hepatocellular
carcinoma. J Surg Oncol. 2010;102:209-214.
26. Louis C, Dewas S, Mirabel X, et al. Stereotactic radiotherapy of
hepatocellular carcinoma: Preliminary results. Technol Cancer Res Treat.
2010;9:479-487.
27. Zeng ZC, Fan J, Tang ZY, et al. A comparison of treatment combinations with and without radiotherapy for hepatocellular carcinoma with portal
vein and/or inferior vena cava tumor thrombus. Int J Radiat Oncol Biol Phys.
2005;61:432-443.
28. Sandroussi C, Dawson LA, Lee M, et al. Radiotherapy as a bridge to
liver transplantation for hepatocellular carcinoma. Transpl Int. 2010;23:299306.
Patient Selection, Resection, and Outcomes

for Hepatocellular Carcinoma
By Claudius Conrad, MD, PhD, and Kenneth K. Tanabe, MD
Overview: Hepatocellular carcinoma (HCC) is an aggressive

malignancy of the liver that most often arises in patients with
cirrhosis and other chronic liver diseases. Worldwide, it is the
sixth most common cancer and the third most common cause
of cancer-related death. Median survival is poor, ranging from
6 to 20 months. Definitive treatment options for HCC are
surgical resection, ablation, or transplantation. The selection
of patients for surgical resection is based on clinical findings,
laboratory data, and imaging. Although a number of staging
systems exist, all have their limitations. A multidisciplinary
approach to patient selection for surgery that includes the
input of an experienced liver surgeon assures optimal outcomes. Sound understanding of liver segmentation, modern
surgical techniques, and the use of intraoperative ultrasound
have led to a reported perioperative mortality rate below 3%,
HE INCIDENCE of HCC is rising. HCC represents the

fastest growing cause of cancer-related death in men in
the United States, with reported overall survival rates of
20% to 60% for resected patients.1 In contrast to other
malignancies, the resectability of this tumor is not only
affected by its anatomic location, the extent of disease, and
the overall medical condition of the patient but also by the
degree of underlying chronic cirrhosis of the liver that is
present in more than 80% of patients. The etiology of the
liver cirrhosis is chronic hepatitis B and C in the majority of
cases, although alcoholic liver disease; cryptogenic cirrhosis;
and, increasingly, nonalcoholic steatohepatitis secondary to
the increasing incidence of obesity are clinically relevant as
well.2 The increased incidence of HCC in the United States
has been primarily attributed to the concomitant increase in
hepatitis C infection.3 However, one must keep in mind that
15% to 20% of patients with HCC in the United States
develop HCC without any known risk factors.4 HCC is now
identified earlier with screening of high-risk patients.2 For
patients with cirrhosis, portal hypertension, and tumors
confined to the liver, orthotopic liver transplantation has
been considered the most effective treatment. However,
there are no prospective randomized controlled trials that
directly compare liver transplantation with liver resection
for HCC. Most studies compare either transplantation or
resection with historical controls and do not analyze according to intent to treat, and a significant selection bias of
these studies cannot be excluded. Transplantation for HCC
is limited by the donor organ shortage, which results in
disease progression and death among patients with HCC
awaiting a donor liver.5 In addition to the limitations of liver
transplantation, surgical resection for HCC has its challenges as well. Only 20% to 30% of all patients with HCC in
the United States will be candidates for either resection or
even locoregional therapies, including radiofrequency and
cryoablation or transarterial chemoembolization. This article will focus on patient selection for surgical resection,
which is mainly determined by extent of disease and liver
function, advances in operative technique, and prognostic
factors that determine outcome.
blood transfusion requirements of less than 10%, and 5-year

survival rates of at least 50%. Advances in laparoscopic
technique and technology have expanded the indications for a
safe and oncologically appropriate minimally invasive resection. Deciding which treatment option to employ depends on
tumor resectability and the degree of underlying liver disease,
which is present in 80% to 85% of patients with HCC; however,
despite these surgical advances, a high recurrence rate of
70% in patients with cirrhosis and a survival rate of 65% to
80% in well-selected transplant patients are expected. This
article will focus on the evaluation and selection of patients
for surgical intervention, considerations in selecting the appropriate type of resection, and expected outcomes following
liver resection.
Patient Selection and Assessment of Hepatic Reserve
The existence of a multitude of scoring and staging systems (e.g., Okuda, Cancer of the Liver Italian Program, and
American Joint Committee on Cancer) suggests that the
ideal one has yet to be identified. Thorough clinical, laboratory, and imaging assessment and careful preoperative
patient selection by experienced liver surgeons are necessary for optimal surgical outcomes. Assessment of HCC
resectability and extent of resection requires evaluation of
patient functional status and comorbidities, tumor location,
and underlying liver function. Routine grayscale ultrasound
has value for screening patients with cirrhosis but rarely
has value in surgical planning. Multidetector computed
tomography (CT) with three-dimensional reconstruction and
magnetic resonance imaging (MRI) are superior for surgical
planning, and they provide information on the morphologic
characteristics of the tumor, presence of intrahepatic metastasis and secondary lesions, extent of chronic liver disease,
and vascular involvement. Special focus should be given to
the number and location of suspicious lesions and any
suspicious regional nodes, as well as any signs of advanced
liver disease (ascites, nodular hepatic contour, enlarged
caudate lobe, splenomegaly, gastrosplenic and umbilical
venous collateral vessels, and fatty hepatic infiltration).1 In
addition, the anatomic relationship of the tumor to important vascular structures as well as the presence of a portal or
hepatic vein thrombus are important findings on CT and
MRI. A vascular thrombus that is bulging in appearance and
enhances with contrast is suspicious for tumor thrombus,
whereas nonenhancing thrombus may represent venous clot
rather than tumor. Macrovascular involvement portends a
poor prognosis, unlikely to be improved with surgical resection, and represents a contraindication to liver transplanta-
From the Massachusetts General Hospital, Department of Surgery, Division of Surgical

Oncology, Harvard Medical School, Boston, MA.
Address reprint requests to Kenneth K. Tanabe, MD, Massachusetts General Hospital, 55
Fruit St., Yawkey 7.924, Boston, MA 02114; email: ktanabe@partners.org.
1092-9118/10/1-10
265
CONRAD AND TANABE
tion. Advanced computer-based image reconstruction

provides a virtual three-dimensional model for accurate
quantitative assessment of areas at risk for devascularization or venous congestion, thus influencing the extent of
resection or need for vascular reconstruction. Liver volumetry using axial images from two-dimensional CT scans is
used to build a virtual model of the liver to measure the
future liver remnant (FLR) as a predictor of postoperative
hepatic dysfunction.6 The following formula for the estimation of the total liver volume (TLV) in adults was found to be
the most precise7:
TLV (cm3) 794.41 1267.28 body-surface area (m2)

Preoperative portal vein embolization of the branches
supplying the portion of the liver to be resected may be used
to induce hypertrophy to increase the FLR and reduce the
risk of morbidity and mortality. This is particularly helpful
if the future remnant volume is below 40% in a noncirrhotic
liver or below 60% in a cirrhotic liver.8 This technique allows
liver regeneration (and therefore an increase in FLR) to take
place before definitive major liver resection. Failure of the
liver to undergo hypertrophy and regenerative hyperplasia
in response to portal vein embolization should be a warning
sign of a decreased ability of the liver to regenerate and an
increased risk of liver failure after liver resection. Portal
vein ligation (rather than percutaneous embolization) can
KEY POINTS
266
A multidisciplinary approach to patient selection for

surgery that includes the input of an experienced
liver surgeon is necessary for optimal surgical outcomes, which are a perioperative mortality rate below
3%, blood transfusion requirements in less than 10%
of cases, and 5-year survival rates of 50%.
Underlying liver disease is present in more than 80%
of patients with hepatocellular carcinoma. Thorough
preoperative clinical, laboratory, and imaging assessment is necessary to optimize patient selection and
avoid small-for-size future liver remnant leading to
liver failure.
Intraoperative ultrasound with full liver mobilization
is an essential component of every liver cancer operation. Anterior approach, hanging maneuver, and
diverse parenchymal transaction devices have improved surgical outcome.
Laparoscopic resection is a viable alternative to open
resection with an improved perioperative period and
similar oncologic outcomes. A laparoscopic approach
may decrease morbidity of salvage liver transplantation.
Risk factors associated with early recurrence are
tumor size, microvascular invasion, satellite nodules,
alpha-fetoprotein levels, and nonanatomical resection. Risk factors associated with late recurrence
include presence of cirrhosis, active hepatitis, vascular invasion, moderate or poor differentiation, and
multinodularity.
Table 1. Overall and Disease-Free Survival Rates after

Resection According to Prognostic Factors a
Prognosticator
Cirrhosis
Tumor Size
Number of lesions
Factor
5-yr OS (%)
5-yr DFS (%)
HCC and cirrhosis

HCC, no cirrhosis
HCC 3 cm
HCC 5 cm
HCC 5 cm
Single
Multiple
2348
4458
5578
4167
2956
3568
2158
2236
2445
3051
2144
2223
1946
625
Abbreviations: DFS, disease-free survival; HCC, hepatocellular carcinoma;

OS,overall survival.
a
Adapted from Rahbari and colleagues. 30 The three most important prognosticators for long-term survival after liver resection for HCC are the absence or
degree of liver cirrhosis; size of the lesion below 3 cm, below 5 cm, or above 5
cm; and the presence of single or multiple lesions.
also be performed laparoscopically in a safe manner and

may be particularly useful in patients undergoing staging
laparoscopy to rule out disseminated disease.9 Before any
regional therapy including resection or transplantation,
chest imaging should be performed to rule out distant
metastases.
In addition to imaging and clinical evaluation, laboratory
studies (total bilirubin, albumin, International normalized
ratio) provide information required to determine Child-Pugh
classification. Patients with class A cirrhosis may be good
surgical candidates, whereas almost no patients with class B
cirrhosis are appropriate for resection. Thrombocytopenia,
especially when combined with splenomegaly, is generally a
sign of portal hypertension and excludes a patient from
consideration of resection (though liver transplantation remains a potential option). Preoperative percutaneous biopsy
to confirm a diagnosis of HCC is not typically required for
lesions that meet radiographic criteria for HCC in the
setting of underlying liver disease.
Advances in Operative Techniques
It is well-accepted that HCC resection should be performed with at least 1-cm margins and sound oncologic
principles (avoidance of tumor spillage). Controversy exists
as to whether there is a survival benefit of anatomic resections according to the Couinaud classification of liver segmentation over nonanatomic resection. A series from an
Asian center demonstrated significantly improved survival
of the group that underwent resection according to
Couinaud classification over nonanatomic resection: 66%
compared with 35%, p 0.01 for 5-year survival, and 34%
compared with 16%, p 0.006 for disease-free survival.10 A
comparison of the outcome with anatomic and nonanatomic
resection for HCC, using a nationwide Japanese database of
72,744 patients, demonstrated an improved disease-free
survival rate with anatomic resection but no difference in
the overall survival rate. When survival was stratified by
tumor size, the disease-free survival rate was significantly
improved with an anatomic resection for HCC with a diameter of 2 to 5 cm (p 0.0005).11 This finding was not
confirmed by Western studies, where only tumor size and
presence of vascular invasion affected survival.12
Full liver mobilization and ultrasound examination allow
full ultrasonic inspection of the liver and delineation of the
intrahepatic anatomy and may spare the patient from undergoing a potentially noncurative operation or one in which
a tumor is unknowingly left behind. Depending on the
SURGICAL MANAGEMENT OF HCC
quality of preoperative imaging, new lesions are detected in

up to 30% of cases by intraoperative ultrasound, of which
one-third will turn out to be malignant, underscoring the
importance of this diagnostic tool.13 Historically, resection of
the right liver has involved mobilization of the right liver
with extreme left displacement to expose the retrohepatic
inferior vena cava followed by extrahepatic control of the
right hepatic vein. Potential deleterious consequences of this
approach are iatrogenic rupture of the tumor and hemodynamic compromise of the left lobe. The so-called anterior
approach entails initial vascular inflow control and parenchymal transection before mobilization of the right liver.
This technique minimizes manipulation of the tumorbearing liver and spillage of tumor cells. Results from a
prospective randomized controlled study of anterior compared with a conventional approach for patients with HCC
measuring 5 cm or more demonstrated significantly less
major operative blood loss of 2 L or more (28.3% vs. 8.3%, p
0.005) and superior overall survival (median 68.1 months vs.
22.6 months, p 0.006) with the anterior approach.14
The liver-hanging maneuver originally described by Belighiti and colleagues15 consists of passing a tape in the
avascular retrohepatic, precaval space to suspend the liver
during parenchymal transaction. Elevation of the tape compresses the liver and reduces bleeding of the deeper parenchymal plane while simultaneously guiding the plane of the
parenchymal transection. The hanging maneuver approach,
which greatly facilitates the anterior approach described
above, has been modified based on the three Glissons
pedicles and hepatic veins facilitating right and left anatomic liver resections.16
The best technique of parenchymal transection remains a
matter of debate. A Cochrane Collaborative meta-analysis of
seven trials evaluated 556 randomly selected patients who
had undergone liver resection using the most common liver
parenchymal transection devices available today.17 The
comparisons include cavitron ultrasound surgical aspirator
(CUSA) compared with the clamp-crush technique (two
trials); radiofrequency dissecting sealer (RFDS) compared
with the clamp-crush technique (two trials); sharp dissection
compared with the clamp-crush technique (one trial); and
hydrojet compared with CUSA (one trial).15 The clampcrush technique is a liver parenchymal transsection technique that involves crushing the liver parenchyma with a
clamp, which leaves blood vessels and bile ducts behind.
Those structures can subsequently be ligated or clipped. One
trial compared CUSA, RFDS, hydrojet, and the clamp-crush
technique. The report found that the clamp-crush technique
was 2 to 6 times less expensive than the other methods
depending on the number of surgeries performed each
year and therefore was favored by the authors. An additional technique that is now commonly used for parenchymal liver transection is stapling with a vascular stapler.
Currently, there is an open prospective trial comparing
the clamp-crush technique to vascular stapler hepatectomy
for parenchymal transection in elective hepatic resection
(CRUNSH, NCT01049607). Weber and colleagues reported
on liver tumor resections using radiofrequency energy in 15
patients between January 2000 and June 2001.18 The device
creates zones of necrosis that are subsequently transected
with a scalpel, thereby rendering this approach suitable for
laparoscopic liver resections.19
A growing body of literature has confirmed the safety and
good long-term outcome of laparoscopic liver resection. The

largest series of laparoscopic liver resection for HCC was a
recent multicenter European trial in which 163 laparoscopic
liver resections were performed in a population of 74%
cirrhotics.20 Inclusion criteria were predefined and included
patients with compensated cirrhosis, esophageal varices of
grade 1 or less, a platelet count of at least 80,000/mm3,
tumors less than 10 cm in size, an absence of major vascular
invasion, and an American Society of Anesthesiologists
(ASA) score of 3 or less, as well as those without evidence of
cirrhosis. Median follow-up was 30.4 months after resection
with 1- and 3-year overall survival rates of 93% and 69%,
respectively. Since this series was reported, operative time,
blood loss, number of transfused packed red blood cells, and
open conversion rates have continued to decline, suggesting
a learning curve of this relatively novel technique.21 A small
study has suggested patients who had previously undergone
laparoscopic resection compared with those who had undergone open surgery for HCC had decreased morbidity following salvage liver transplantation.22 In this study, 24 total
patients underwent salvage liver transplant after either
prior laparoscopic (12 patients) or open resection (12 patients). The laparoscopy group demonstrated shorter resection and total operative time, less blood loss, and reduced
need for blood transfusions. However, the most important
question in comparing open resection with laparoscopic
resection is whether oncologic outcomes are the same. A
recent meta-analysis evaluated 10 nonrandomized controlled studies with 494 subjects, of whom 213 underwent
laparoscopic and 281 underwent open resection for HCC. In
addition to the improved morbidity among patients undergoing laparoscopic compared with open resection, there was
no difference between the groups with respect to surgical
margins, overall survival rates, and disease-free survival
rates.23 These findings were similar to those observed in a
recent meta-analysis of 10 studies looking at 627 patients
from China.24 Despite the absence of higher-level evidence,
laparoscopic liver resection for HCC is rapidly becoming
preferable to open resection in well-selected patients.
Outcomes and Prognostic Factors
Hepatic resection for HCC has become a safer operation,

with a reported in-hospital mortality rate of around 2% and
a 90-day mortality rate of 5%, largely due to advances in
surgical technique and improved patient selection, as demonstrated in a recent series of 129 patients with HCC from
Toronto.25 These marked improvements in outcomes can, in
part, be attributed to increased utilization of segmental and
parenchymal-sparing resections and decreased intraoperative blood loss. Nevertheless, morbidity for liver resections
rates remain high at 20% to 50% and include complications
such as pleural effusion (9%), perihepatic abscess (6%), ileus
(6%), sterile perihepatic fluid collections (5%), wound infection (5%), urinary tract infection (4%), bile leak/biloma (3%),
pneumonia (3%), and deep venous thrombosis (2%), as
reported in a series of 1,803 patients consisting of 21%
patients with HCC.26 Expected 1-, 3-, 5-, and 10-year survival rates following HCC resection are 85%, 64%, 45%, and
21%, respectively, as reported by the Liver Cancer Study
Group in Japan. These data are derived from the largest
report to date, which includes 6,785 patients with cirrhosis
operated on between 1988 and 1999.27 Patients without
significant portal hypertension and normal bilirubin achieve
267
CONRAD AND TANABE
a 5-year survival rate of 70% following liver resection for

HCC. In those with portal hypertension, however, the rate of
5-year survival is 50% and even lower in the setting of an
abnormal bilirubin. Similar results have been reported from
Western and other Asian centers. Survival rates as high as
60% at 5 years may be achieved in Child class A patients
with well-encapsulated tumors of 2 cm in diameter, which
are equivalent to results following liver transplantation.
Unfortunately, few patients ( 10%) meet these selection
criteria.28 Furthermore, recurrence after resection is common, occurring in up to 80% of the patients at 5 years.
Approximately two-thirds of recurrences occur within 2
years after treatment; these are considered early recurrence
by convention.29
The factors associated with early recurrence are vascular
invasion, tumor size (only 25% of patients with HCC of less
than 2 cm have vascular invasion), satellite nodules, alphafetoprotein levels, nonanatomic resection, and extent of
underlying disease. There has been some debate as to
whether tumor recurrence 2 or more years following resection represents a true recurrence of the initial lesion or a de
novo HCC in the oncogenic cirrhotic liver. Risk factors
associated with late recurrence include presence of cirrhosis,
active hepatitis, vascular invasion, moderate or poorly differentiated HCC, and multinodularity.29 Genomic analyses
to better define recurrence risk based on primary tumor
characteristics and molecular evaluation of the cirrhotic
liver may be valuable for prognostication in the future,30
and future research will show whether neoadjuvant therapy
with radiation, retinoids, chemoembolization, interferon, or
sorafenib (an inhibitor of several tyrosine protein kinases
[VEGFR, PDGFR, Raf]) leads to delayed recurrence. Despite
the fact that recurrence following resection of HCC is a poor
prognosticator, there is evidence that some patients will
benefit from aggressive surgical approaches if recurrence
occurs. It is our practice to offer patients regional therapy

occasionally including resectionfor recurrent HCC when
the recurrence is limited to the liver. If recurrence occurs, it
has been reported that multimodality therapy including
transarterial chemoembolization, percutaneous ablations,
and surgery results in an overall 5-year survival rate of 20%.
Conclusion
Surgical resection, ablation, and liver transplantation

offer the highest likelihood of long-term survival or cure in
carefully selected patients. A multidisciplinary approach by
a team consisting of surgical oncologists, transplant surgeons, medical oncologists, interventional radiologists, radiation oncologists, social workers, and other ancillary staff
provides best outcomes. Advances in surgical techniques,
technology, and understanding of liver anatomy allow experienced liver surgeons to perform resections that are safer
than before, and resections are available to more patients
than before. Whereas neoadjuvant therapy has not been
shown to convert unresectable tumors to resectable ones,
preoperative portal vein embolization can increase the FLR
and convert patients from unresectable to resectable. Ablative techniques can be valuable adjuncts in the surgical
therapy of patients with HCC. And in some instances
ablation can be the sole local treatment modality during an
open or laparoscopic operation, recognizing that ablation is
similarly effective as resection for early HCC. Improvements
in surgical techniques have contributed to decreased perioperative morbidity and mortality. Important prognostic factors associated with survival after resection include vascular
invasion, satellite nodules, and extent of underlying liver
disease. Further improvements in survival after hepatectomy for HCC will depend on strategies for early detection as
well as effective neoadjuvant or adjuvant therapies.
Author
Claudius Conrad*
Kenneth K. Tanabe
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Best Doctors;
Covidien; Health
Advances; LEK
Consulting;
UpToDate
Honoraria
Springer
Science and
Business Media
Research
Funding
Expert
Testimony
Other
Remuneration
AstraZeneca
REFERENCES
1. Duffy JP, Hiatt JR, Busuttil RW. Surgical resection of hepatocellular
carcinoma. Cancer J. 2008;14:100-110.
2. Bruix J, Hessheimer AJ, Forner A, et al. New aspects of diagnosis and
therapy of hepatocellular carcinoma. Oncogene. 2006;25:3848-3856.
3. Davila JA, Morgan RO, Shaib Y, et al. Hepatitis C infection and the
increasing incidence of hepatocellular carcinoma: a population-based study.
Gastroenterology. 2004;127:1372-1380.
4. El-Serag HB, Mason AC. Risk factors for the rising rates of primary liver
cancer in the United States. Arch Intern Med. 2000;160:3227-3230.
5. Duffy JP, Vardanian A, Benjamin E, et al. Liver transplantation criteria
for hepatocellular carcinoma should be expanded: a 22-year experience with
467 patients at UCLA. Ann Surg. 2007;246:502-509; discussion 509-511.
6. Schindl MJ, Redhead DN, Fearon KC, et al. The value of residual liver
volume as a predictor of hepatic dysfunction and infection after major liver
resection. Gut. 2005;54:289-296.
7. Johnson TN, Tucker GT, Tanner MS, et al. Changes in liver volume from
birth to adulthood: a meta-analysis. Liver Transpl. 2005;11:1481-1493.
268
8. Makuuchi M, Sano K. The surgical approach to HCC: our progress and

results in Japan. Liver Transpl. 2004;10:S46-S52.
9. Are C, Iacovitti S, Prete F, et al. Feasibility of laparoscopic portal vein
ligation prior to major hepatectomy. HPB (Oxford). 2008;10:229-233.
10. Hasegawa K, Kokudo N, Imamura H, et al. Prognostic impact of anatomic resection for hepatocellular carcinoma. Ann Surg. 2005;242:252-259.
11. Eguchi S, Kanematsu T, Arii S, et al. Comparison of the outcomes
between an anatomical subsegmentectomy and a non-anatomical minor
hepatectomy for single hepatocellular carcinomas based on a Japanese
nationwide survey. Surgery. 2008;143:469-475.
12. Eltawil KM, Kidd M, Giovinazzo F, et al. Differentiating the impact of
anatomic and non-anatomic liver resection on early recurrence in patients
with Hepatocellular Carcinoma. World J Surg Oncol. 2010;8:43.
13. Takigawa Y, Sugawara Y, Yamamoto J, et al. New lesions detected by
intraoperative ultrasound during liver resection for hepatocellular carcinoma.
Ultrasound Med Biol. 2001;27:151-156.
14. Liu CL, Fan ST, Cheung ST, et al. Anterior approach versus conven-
SURGICAL MANAGEMENT OF HCC

tional approach right hepatic resection for large hepatocellular carcinoma: a
prospective randomized controlled study. Ann Surg. 2006;244:194-203.
15. Belghiti J, Guevara OA, Noun R, et al. Liver hanging maneuver: a safe
approach to right hepatectomy without liver mobilization. J Am Coll Surg.
2001;193:109-111.
16. Kim SH, Park SJ, Lee SA, et al. Various liver resections using hanging
maneuver by three Glissons pedicles and three hepatic veins. Ann Surg.
2007;245:201-205.
17. Gurusamy KS, Pamecha V, Sharma D, et al. Techniques for liver
parenchymal transection in liver resection. Cochrane Database Syst Rev.
2009;(1):CD006880.
18. Weber JC, Navarra G, Jiao LR, et al. New technique for liver resection
using heat coagulative necrosis. Ann Surg. 2002;236:560-563.
19. Jiao LR, Ayav A, Navarra G, et al. Laparoscopic liver resection assisted
by the laparoscopic Habib Sealer. Surgery. 2008;144:770-774.
20. Dagher I, Belli G, Fantini C, et al. Laparoscopic hepatectomy for hepatocellular carcinoma: a European experience. J Am Coll Surg. 2010;211:16-23.
21. Reddy SK, Tsung A, Geller DA. Laparoscopic liver resection. World
J Surg. 35:1478-1486.
22. Gigot JF, Glineur D, Santiago Azagra J, et al. Laparoscopic liver
resection for malignant liver tumors: preliminary results of a multicenter
European study. Ann Surg. 2002;236:90-97.
23. Zhou YM, Shao WY, Zhao YF, et al. Meta-analysis of laparoscopic versus
open resection for hepatocellular carcinoma. Dig Dis Sci. 2011;56:1937-1943.
24. Li N, Wu YR, Wu B, et al. Surgical and oncologic outcomes following

laparoscopic versus open liver resection for hepatocellular carcinoma: a
meta-analysis. Hepatol Res. 2012;42:51-59.
25. Greco E, Nanji S, Bromberg IL, et al. Predictors of peri-opertative
morbidity and liver dysfunction after hepatic resection in patients with
chronic liver disease. HPB (Oxford). 2011;13:559-565.
26. Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative
outcome after hepatic resection: analysis of 1,803 consecutive cases over the
past decade. Ann Surg. 2002;236:397-406; discussion 406-407.
27. Ikai I, Itai Y, Okita K, et al. Report of the 15th follow-up survey of
primary liver cancer. Hepatol Res. 2004;28:21-29.
28. Cha CH, Ruo L, Fong Y, et al. Resection of hepatocellular carcinoma in
patients otherwise eligible for transplantation. Ann Surg. 2003;238:315-321;
discussion 321-323.
29. Imamura H, Matsuyama Y, Tanaka E, et al. Risk factors contributing
to early and late phase intrahepatic recurrence of hepatocellular carcinoma
after hepatectomy. J Hepatol. 2003;38:200-207.
30. Hoshida Y, Villanueva A, Kobayashi M, et al. Gene expression in fixed
tissues and outcome in hepatocellular carcinoma. N Engl J Med. 2008;359:
1995-2004.
31. Rahbari NN, Mehrabi A, Mollberg NM, et al. Hepatocellular carcinoma:
current management and perspectives for the future. Ann Surg. 2011;253:
453-469.
269
THE MANAGEMENT OF LESS COMMON BUT

COMPLEX UPPER GASTROINTESTINAL
MALIGNANCIES: HEPATOCELLULAR CARCINOMA,
PANCREATIC NEUROENDOCRINE TUMORS, AND
BILIARY TRACT TUMORS
CHAIR
Pamela L. Kunz, MD
Stanford University Medical Center
Stanford, CA
SPEAKERS
Andrew X. Zhu, MD, PhD
Boston, MA
Jordan D. Berlin, MD
Nashville, TN
A Renaissance in Therapeutic Options for

Pancreatic Neuroendocrine Tumors
By Pamela L. Kunz, MD
Overview: The field of pancreatic neuroendocrine tumors

(NETs) has seen a remarkable renaissance in recent years with
exponential increases in published research, clinical trials,
and U.S. Food and Drug Administration (FDA)-approved treatments. Surgical resection remains the foundation for management of locoregional disease. However, for patients with
advanced disease, novel therapeutic options have emerged.
ETS ARISE from neuroendocrine cells throughout

the body, most commonly in the lungs, gastrointestinal
tract, and pancreas. The field of pancreatic NETs has seen
a remarkable renaissance in recent years, with exponential
increases in published research, clinical trials, and FDAapproved treatments (Fig. 1). Pancreatic NETs have an
estimated incidence of 0.32 cases/100,000 people1 and account for 1% of all pancreatic cancers by incidence and 10%
of pancreatic cancers by prevalence.2 The median age at
diagnosis is age 60, with a white male predominance. The
majority of NETs are functionally inactive but some produce
hormones that lead to the clinical syndromes associated
with hormone excess. Though most pancreatic NETs are
indolent, the spectrum of clinical behavior is quite variable
and sometimes difficult to predict. The nomenclature, classification, and grading systems for NETs have historically
been inconsistent. Through critical evaluations of these
systems, common principles have emerged.3,4 NETs are
generally divided into two main categories: welldifferentiated (low and intermediate grade) and poorly differentiated (high grade). The proliferative rate (mitotic
index and Ki67) is considered a critical component of pathologic evaluation. Well-differentiated tumors usually have
less than 20 mitoses/high powered field (hpf ) and a Ki67
index of less than 20%, and poorly differentiated tumors
more than 20 mitoses/hpf and a Ki67 more than 20%.
This discussion will focus on the management of welldifferentiated pancreatic NETs.
Approach to Treatment
Locoregional Disease
At the time of diagnosis, approximately 40% of patients

with pancreatic NETs have locoregional disease.2 Surgical
resection remains the mainstay of treatment for these patients. There is currently no role for adjuvant treatment
following resection of a primary pancreatic NET, as little is
known about patient and tumor characteristics that predict
for recurrence.
Unresectable and Metastatic Disease
Nonsurgical therapies play a role in the setting of unresectable and metastatic disease, which account for approximately 50% of newly diagnosed patients.2 Patient selection
is a critical first step in the treatment algorithm. For
patients with functional tumors, somatostatin analogs
should be considered to alleviate symptoms of peptide release. For patients with nonfunctional tumors, systemic
treatments are generally indicated for patients with pro-
Two separate randomized placebo-controlled studies have

shown prolonged progression-free survival (PFS) with everolimus or sunitinib. Future studies are designed to answer
questions about the role of somatostatin analogs as antiproliferative agents, combinations of biologic therapies, and new
cytotoxic chemotherapy backbones.
gressive, bulky, or symptomatic disease. Otherwise, patients

with low-volume, stable, and asymptomatic disease may be
monitored closely without treatment.
Mediating Symptoms of Hormone Excess
An estimated 40% to 53% of pancreatic NETs are functional.5,6 Functional tumors are defined as having inappropriate elevation of serum hormone markers combined with
clinical evidence of hormone oversecretion, including insulinomas, gastrinomas, VIPomas, glucagonomas, and somatostatinomas. Somatostatin analogs are the foundation of
symptom management for patients with functional pancreatic NETs and can both decrease the secretion of such
hormones and inhibit their end-organ effects. Somatostatin
is a naturally occurring polypeptide produced by paracrine
cells that are scattered throughout the gastrointestinal tract
and inhibits gastrointestinal endocrine and exocrine function. Its effects are mediated through G-coupled protein
somatostatin receptors (SSTR15). Short- and long-acting
octreotide (with high affinity for SSTR2) is available in the
United States. Lanreotide, available in Europe, is a longacting analog with similar binding affinity to octreotide.
Pasireotide, a novel somatostatin analog with a different
binding affinity profile compared to octreotide or lanreotide,
is currently in development. Note that somatostatin analogues should be used with caution for patients with insulinomas, as it may precipitate or worsen hypoglycemia.
Oncologic Control
Biologic Agents
Recent studies with inhibitors of signaling pathways

that target vascular endothelial growth factor (VEGF) and
the mammalian target of rapamycin (mTOR) have demonstrated considerable activity in pancreatic NETs.
RADIANT-3 was a randomized phase III study that evaluated the efficacy of everolimus, an mTOR inhibitor, in
advanced pancreatic NETs.7 In this international, multisite
study, 410 patients with low- or intermediate-grade progressive advanced pancreatic NETs were randomly selected to
From the Stanford University School of Medicine, Stanford Cancer Institute, Stanford,
CA.
Address reprint requests to Pamela L. Kunz, MD, Stanford University School of Medicine, Stanford Cancer Institute, 875 Blake Wilbur Drive, Stanford, CA 94305; email:
pkunz@stanford.edu.
1092-9118/10/1-10
271
PAMELA L. KUNZ
RRs in the sunitinib and placebo arms were 9.3% and 0%,
respectively. A high proportion of patients in the placebo
arm received sunitinib at progression through a continuation study, and although an early survival analysis appeared to favor the sunitinib arm, this difference did not
remain significant with additional follow-up (Table 1).
Cytotoxic Chemotherapy
Fig 1. PubMed publications and clinical trial postings by year for

pancreatic NETs.
receive everolimus 10 mg orally daily or placebo. The primary endpoint was PFS; secondary endpoints were overall
survival (OS), response rate (RR), and safety. The median
PFS was 11.0 months with everolimus as compared to 4.6
months with placebo (hazard ratio [HR] 0.35; 95% CI 0.27 to
0.45; p 0.001). The RR was 5% in the everolimus arm
compared to 2% in the placebo arm. Updated OS showed no
difference between arms; however, patients in the placebo
arm crossed over to treatment at progression, likely reducing the chance of observing a survival difference (Table 1).
Another randomized study evaluated the efficacy of
sunitinib, an inhibitor of the VEGF pathway, in advanced
pancreatic NETs. One hundred seventy one patients with
advanced, well-differentiated, progressive pancreatic NETs
were randomly selected to receive sunitinib 37.5 mg orally
daily or placebo.8 Primary endpoints were PFS; secondary
endpoints were RR, OS, and safety. The study was discontinued before the preplanned interim efficacy analysis, after
an independent data and safety monitoring committee observed more serious adverse events and deaths in the
placebo arm and a difference in PFS that favored the
sunitinib arm. At the time of study closure, median PFS was
11.4 months in the sunitinib arm compared with 5.5 months
in the placebo arm (HR 0.42; 95% CI 0.26 to 0.66; p 0.001).
KEY POINTS
272
Surgical resection continues to be the mainstay of

treatment for patients with locoregional disease.
Somatostatin analogs are indicated for patients experiencing symptoms of hormone excess; there are
currently no prospective data to support the use of
somatostatin analogs as antitumor agents for pancreatic neuroendocrine tumors.
For patients with advanced disease, single-agent
everolimus and sunitinib have recently been shown to
improve progression-free survival.
Temozolomide-based chemotherapy regimens are
emerging as an acceptable alternative to streptozocin; prospective randomized studies are in
development.
Patients with advanced pancreatic NETs currently have

few treatment options that yield objective radiographic
tumor regression. Recent studies evaluating everolimus7
and sunitinib8 in this patient population have demonstrated
prolongation of PFS compared to placebo, but overall tumor
RRs (by RECIST) with these agents are less than 10%.
Historical studies reporting the highest RRs include regimens with cytotoxic chemotherapy. In an initial randomized study of 106 patients, Moertel and colleagues reported
activity associated with the combination of streptozocin and
doxorubicin in patients with advanced islet-cell tumors
(Table 1).9 The RR associated with this regimen was reported to be 69%; however, because of the use of nonstandard response criteria, the true objective radiologic RR was
likely much lower. Retrospective series have reported overall RRs of 6% to 39% associated with streptozocin-based
regimens in pancreatic NETs.10,11 Although clearly associated with activity, the combination of streptozocin and
doxorubicin is also associated with considerable toxicity,
including myelosuppression, asthenia, and renal insufficiency, precluding its routine use in this disease. Additionally, recent issues with drug availability have made routine
use difficult.
Recent retrospective series and small, prospective phase
II studies suggest that temozolomide is similarly active but
less toxic than streptozocin-based therapy when treating
patients with pancreatic NETs. In a retrospective series of
36 patients treated with temozolomide monotherapy, tumor
regression was observed in 31% of bronchial carcinoid tumors and 8% of pancreatic NETs.12 In a series of 97 patients,
18 of 53 patients with pancreatic NETs (34%) achieved a
partial or complete response to temozolomide-based therapies.13 A third retrospective series of 21 patients treated
with temozolomide monotherapy demonstrated responses in
25% of pancreatic NETs.14
Temozolomide has also been investigated prospectively in
phase II studies of patients with NETs, though always in
combination with other agents. In an initial study, 29
patients with metastatic NETs were treated with a combination of temozolomide, administered at a dose of 150 mg/m2
Table 1. Studies Leading to FDA Approval of Agents in

Advanced Pancreatic NETs
Regimen
Number of
Patients
RR
(%)
TTP or PFS
(mo)
OS
(mo)
Strep/dox vs.
Strep/5FU vs.
Chlorotozocin
Everolimus vs.
Placebo
Sunitinib vs.
Placebo
36
33
33
207
203
86
85
69
45
30
5
2
9
0
20
6.9
6.9
11.0
4.6
11.4
5.5
26.4
16.8
16.8
NR
36.6
30.5
24.4
Reference
Moertel et al.9
Yao et al.7
Raymond et al.8
Abbreviations: mo, months; NET, pancreatic neuroendocrine tumor; NR, not

reached; OS, overall survival; PFS, progression-free survival; RR, response rate;
TTP, time to progression.
THERAPEUTIC OPTIONS FOR PANCREATIC NETS
for 7 days, every other week, and thalidomide at doses of

50 to 400 mg daily; this combination was associated with
objective tumor responses in 5 of 11 patients (45%).15 A
subsequent phase II study evaluated the combination of
temozolomide and bevacizumab in the treatment of 34
patients with NETs (15 pancreatic, 19 carcinoid).16 Patients
received 150 mg/m2/day of temozolomide orally for 7 days,
every other week, and 5 mg/kg of bevacizumab intravaneously every other week. Because of anticipated lymphopenia, patients also received prophylaxis with trimethoprim/
sulfamethoxazole (1 double strength tablet every Monday,
Wednesday, Friday). Objective tumor responses were observed in 33% of patients with pancreatic NETs, but not in
patients with carcinoid tumors. In a third, prospective
study, a regimen of everolimus and temozolomide was associated with an overall tumor RR of 35% for patients with
advanced pancreatic NETs.17 Taken together, these prospective and retrospective studies suggest that temozolomidebased therapy is comparable to streptozocin-based regimens
and might reasonably be associated with an overall tumor
RR of 30% to 40% for patients with advanced pancreatic
NETs.
Interestingly, preclinical and early clinical evidence suggests that capecitabine may be synergistic with temozolomide.18,19 In a series of 17 patients with pancreatic NETs,
combination therapy with temozolomide and capecitabine
was associated with a tumor RR of 59%.20 A recent singleinstitution retrospective study by Strosberg and colleagues
reported RRs of 70% and median PFS of 18 months for 30
patients with advanced pancreatic NETs.21 In this study,
temozolomide was administered at a dose of 200 mg/m2
on days 10 to 14, and capecitabine was administered at a
dose of 750 mg/m2 twice daily on days 1 to 14. The combination and specific dosing schedules were well-tolerated,
with only four patients experiencing grade 3 or 4 adverse
events (anemia, thrombocytopenia, elevated aspartate aminotransferase, and elevated alanine aminotranferease). The
most common grade 1 and 2 adverse events were fatigue,
nausea, myelosuppression, and hand-foot syndrome. Based
on this study, the combination of temozolomide and capecitabine has become popular and commonly used in patients
with advanced pancreatic NETs.
Other Therapies for Advanced Disease
Hepatic metastases commonly occur in patients with

pancreatic NETs and adversely affect overall prognosis and
quality of life. Therapies directed at locoregional control of
hepatic disease may be necessary to decrease symptoms
associated with hormone excess. Surgery for hepatic metastases should be considered whenever the metastases are
considered resectable and when there is no evidence of
extrahepatic disease. Thermal ablation or cryoablation may
be considered as an adjunct to surgery or in settings where
extrahepatic disease or comorbidities might favor a less
aggressive intervention. Selective catheterization of the
hepatic artery and embolization of vessels perfusing the
tumors can result in clinically significant responses. Embolization options include bland embolization, chemoembolization, and radioembolization. Retrospective studies report
benefit, though no prospective studies have been conducted.22
Radiolabeled somatostatin analogs with therapeutic doses
of the radioactive isotope can also provide disease control in

patients with advanced disease and are used routinely in
Europe.23 The most commonly used radionuclides are indium (111I), yttrium (90Y), and lutetium (177Lu) and are only
available in Europe. The largest retrospective study of
patients with gastroenteropancreatic NETs demonstrated
clinical responses in 46% of patients at 3 months (complete
2%, partial 28%, and minor 16%) and stable disease in 36%;
the minority had progressive disease (20%). Median time to
progression (TTP) was 40 months; median OS was 128
months. The bone marrow and kidneys are the most important dose-limiting organs in peptide receptor radionuclide
therapy.24 Prospective randomized studies are needed to
confirm these observed benefits.
Ongoing and Future Directions
The renaissance of clinical research in the field of NETs

has only just begun. Ongoing and proposed studies are
poised to answer important questions.
First, do somatostatin analogs have antitumor activity in
pancreatic NETs? Though many physicians extrapolate the
findings from the PROMID study,25 there is currently no
prospective data to support the use of somatostatin analogs
as antitumor agents for pancreatic NETs. An ongoing international study was designed to answer this question by
randomly selecting patients with advanced nonfunctioning
pancreatic NETs to receive lanreotide autogel versus placebo (CLARINET, NCT00353496). Primary endpoints are
TTP and death. The study opened in June 2006, and accrual
is ongoing.
Second, does the combination of two biologic agents improve efficacy outcomes in pancreatic NETs? A CALGB
study (80701, NCT01229943) addresses this question and
randomly selects patients with advanced pancreatic NETs to
receive everolimus and octreotide long-acting release (LAR)
versus everolimus, bevacizumab, and octreotide LAR. The
primary endpoint is PFS. The study opened in October 2010,
and accrual is ongoing.
The development of thoughtful future studies is critical
in moving the field forward. In fact, a National Cancer
Institute Neuroendocrine Tumor Clinical Trials Planning
Meeting26 held in 2009 identified key unmet needs, recommended appropriate study endpoints and inclusion criteria,
and formulated priorities for future NET studies. The prospective evaluation of single-agent temozolomide and a
comparision to temozolomide-based combinations for pancreatic NETs was a key recommendation of this meeting. ECOG 2211 is a proposed study in which patients with
advanced pancreatic NETs will be randomly selected to
receive temozolomide versus temozolomide and capecitabine. The principal objective of the study will be to assess
whether the addition of capecitabine to temozolomide improves RRs when compared to temozolomide alone. It is
anticipated that the superior arm will serve as a building
block in future studies.
Conclusion
The recent advances in the field of pancreatic NETs

are truly exciting. We now have additional agents in our
arsenal of therapeutic options for patients with this disease,
including everolimus and sunitinib, both of which have
been shown to prolong PFS. Given the increasing number
273
PAMELA L. KUNZ
of therapeutic options, patient selection and treatment

sequence can be complicated. A number of tools are available for the treating physician, including the National
Comprehensive Cancer Network Neuroendocrine Tumor
Guidelines27 and guidelines from the North American
Neuroendocrine Tumor Society.28 Additionally, a multidis-
ciplinary approach can be helpful as modalities from medical

oncology, interventional radiology, and surgery need to be
weighed. Future studies are aimed at answering questions
about the utility of somatostatin analogs for tumor control,
use of combined biologics, and temozolomide-based chemotherapy regimens.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Pamela L. Kunz
Research
Funding
Expert
Testimony
Other
Remuneration
Roche/Genentech
REFERENCES
1. Yao JC, Hassan M, Phan A, et al. One hundred years after carcinoid:
epidemiology of and prognostic factors for neuroendocrine tumors in 35,825
cases in the United States. J Clin Oncol. 2008;26:3063-3072.
2. Yao JC, Eisner MP, Leary C, et al. Population-based study of islet cell
carcinoma. Ann Surg Oncol. 2007;14:3492-3500.
3. Klimstra DS, Modlin IR, Adsay NV, et al. Pathology reporting of
neuroendocrine tumors: application of the Delphic consensus process to the
development of a minimum pathology data set. Am J Surg Pathol. 2010;34:
300-313.
4. Klimstra DS, Modlin IR, Coppola D, et al. The pathologic classification
of neuroendocrine tumors: a review of nomenclature, grading, and staging
systems. Pancreas. 2010;39:707-712.
5. Ito T, Tanaka M, Sasano H, et al. Preliminary results of a Japanese
nationwide survey of neuroendocrine gastrointestinal tumors. J Gastroenterol. 2007;42:497-500.
6. Panzuto F, Nasoni S, Falconi M, et al. Prognostic factors and survival in
endocrine tumor patients: comparison between gastrointestinal and pancreatic localization. Endocr Relat Cancer. 2005;12:1083-1092.
7. Yao JC, Shah MH, Ito T, et al. Everolimus for advanced pancreatic
neuroendocrine tumors. N Engl J Med. 2011;364:514-523.
8. Raymond E, Dahan L, Raoul JL, et al. Sunitinib malate for the
treatment of pancreatic neuroendocrine tumors. N Engl J Med. 2011;364:501513.
9. Moertel CG, Lefkopoulo M, Lipsitz S, et al. Streptozocin-doxorubicin,
streptozocin-fluorouracil or chlorozotocin in the treatment of advanced isletcell carcinoma. N Engl J Med. 1992;326:519-523.
10. Kouvaraki M, Ajani J, Hoff P, et al. Fluorouracil, doxorubicin, and
streptozocin in the treatment of patients with locally advanced and metastatic pancreatic endocrine carcinomas. J Clin Oncol. 2004;22:4762-4771.
11. Cheng P, Saltz L. Failure to confirm major objective antitumor activity
of streptozocin and doxorubicin in the treatment of patients with advanced
islet cell carcinoma. Cancer. 1999;86:944-948.
12. Ekeblad S, Sundin A, Janson ET, et al. Temozolomide as monotherapy
is effective in treatment of advanced malignant neuroendocrine tumors. Clin
Cancer Res. 2007;13:2986-2991.
13. Kulke MH, Hornick JL, Frauenhoffer C, et al. O6-methylguanine DNA
methyltransferase deficiency and response to temozolomide-based therapy in
patients with neuroendocrine tumors. Clin Cancer Res. 2009;15:338-345.
14. Maire F, Hammel P, Faivre S, et al. Temozolomide: a safe and effective
treatment for malignant digestive endocrine tumors. Neuroendocrinology.
2009;90:67-72.
15. Kulke M, Stuart K, Enzinger P, et al. Phase II study of temozolomide
and thalidomide in patients with metastatic neuroendocrine tumors. J Clin
Oncol. 2006;24:401-406.
274
16. Kulke M, Stuart K, Earle C, et al. A phase II study of temozolomide and

bevacizumab in patients with advanced neuroendocrine tumors. J Clin Oncol.
2006;24:189s (suppl; abstr 4044).
17. Kulke M, Blaszkowsky L, Zhu A, et al. Phase I/II study of everolimus
(RAD001) in combination with temozolomide (TMZ) in patients (pts) with
advanced pancreatic neuroendocrine tumors (NET). Gastrointestinal Cancers
Symposium Proceedings. 2010 (abstr 223).
18. Fine R, Fogelman D, Schreibman S. Effective treatment of neuroendocrine tumors with temozolomide and capecitabine. J Clin Oncol. 2005;23:361s
19. Murakami J, Lee YJ, Kokeguchi S, et al. Depletion of O6methylguanine-DNA methyltransferase by O6-benzylguanine enhances 5-FU
cytotoxicity in colon and oral cancer cell lines. Oncol Rep. 2007;17:1461-1467.
20. Isacoff W, Moss R, Pecora A, et al. Temozolomide/capecitabine therapy
for metastatic neuroendocrine tumors of the pancreas: a retrospective review.
J Clin Oncol. 2006;24:625s (suppl; abstr 14023).
21. Strosberg JR, Fine RL, Choi J, et al. First-line chemotherapy with
capecitabine and temozolomide in patients with metastatic pancreatic endocrine carcinomas. Cancer. 2011;117:268-275. Epub 2010 Sep 7.
22. Kvols LK, Turaga KK, Strosberg J, et al. Role of interventional
radiology in the treatment of patients with neuroendocrine metastases in the
liver. J Natl Compr Canc Netw. 2009;7:765-772.
23. van Vliet EI, Teunissen JJ, Kam BL, et al. Treatment of gastroenteropancreatic neuroendocrine tumors with peptide receptor radionuclide therapy. Neuroendocrinology. Epub 2012 Jan 10.
24. Kwekkeboom DJ, de Herder WW, Kam BL, et al. Treatment with the
radiolabeled somatostatin analog [177Lu-DOTA0,Tyr3] octreotate: toxicity,
efficacy, and survival. J Clin Oncol. 2008;26:2124-2130.
25. Rinke A, Muller HH, Schade-Brittinger C, et al. Placebo-controlled,
double-blind, prospective, randomized study on the effect of octreotide LAR
in the control of tumor growth in patients with metastatic neuroendocrine
midgut tumors: a report from the PROMID Study Group. J Clin Oncol.
2009;27:4656-4663.
26. Kulke MH, Siu LL, Tepper JE, et al. Future directions in the treatment
of neuroendocrine tumors: consensus report of the National Cancer Institute
Neuroendocrine Tumor clinical trials planning meeting. J Clin Oncol. 2011;
29:934-943.
27. Kulke M. NCCN Clinical Practice Guidelines in Oncology (NCCN
Guidelines): Neuroendocrine Tumors. Version 1. National Comprehensive
Cancer Network; 2011.
28. Kulke MH, Anthony LB, Bushnell DL, et al. NANETS treatment
guidelines: well-differentiated neuroendocrine tumors of the stomach and
pancreas. Pancreas. 2010;39:735-752.
A Review of the Management of

Hepatocellular Carcinoma: Standard Therapy
and a Look to New Targets
By Andrew X. Zhu, MD, PhD
Overview: Management of hepatocellular carcinoma (HCC)

continues to be challenging, but new treatment options are
evolving. A multidisciplinary evaluation will help make the best
treatment decisions for each patient. Although we continue to
improve the outcomes of curative treatment with resection,
liver transplant, and radiofrequency ablation (RFA), many new
liver-directed regional therapies including drug-eluting beads,
CC IS a malignancy of global importance: it is the sixth

most common cancer and the third most common
cause of cancer-related mortality worldwide.1 In the United
States, the incidence of HCC has tripled, and the 5-year
survival rate has remained below 12% during the past two
decades.2 HCC is the most common primary liver cancer,
accounting for 90% of all liver malignancies and has become
the fastest-rising cause of cancer-related death in the United
States. Management of HCC continues to present with many
challenges. First, despite the well-recognized risk factors,
including hepatitis B infection (HBV), hepatitis C infection
(HCV), alcoholic liver disease, and nonalcoholic fatty liver
disease, as well as the increased use of surveillance programs, most patients still present with unresectable or
advanced-stage disease. Unfortunately, these patients do
not benefit from curative treatment options including surgical resection or liver transplantation. Second, despite the
availability of many local-regional therapies, including radiofrequency ablation (RFA) and transarterial chemoembolization (TACE), there is a paucity of data from definitive
phase III studies. It is often confusing to the community
oncologist how to select the most appropriate treatment for
HCC patients, underscoring the importance of multidisciplinary evaluation. Third, most patients with HCC have
underlying cirrhosis, which adversely affects the overall
survival (OS) of these patients and greatly limits the treatment options. Fourth, despite the approval of sorafenib
based on phase III trials demonstrating OS benefits3,4 and
its extensive application in clinical practice during the past
few years, it is increasingly clear that the benefits of
sorafenib remain modest. More importantly, the mechanisms of sorafenibs efficacy, toxicity, and resistance remain
elusive. On the other hand, we have witnessed an unparalleled time period of active drug development in HCC. Many
molecularly targeted agents that inhibit different pathways
of hepatocarcinogenesis are under various phases of clinical
development. This review will attempt to summarize briefly
the current status of management of HCC with a focus on
the advanced stage of disease and potential new targeted
agents under investigation in HCC.
Staging and General Principles for the Management
of HCC
Once HCC is diagnosed, clinicians would aim to assess the

extent of disease, the presence and severity of underlying
cirrhosis, the performance status, and the access to institutional or local-regional expertise in formulating a treatment
radioembolization, and radiation are emerging. Sorafenib remains the only approved agent for advanced HCC, and its
role in the adjuvant setting following resection or RFA, with
transarterial chemoembolization, or in combination with other
targeted agents or chemotherapy in the advanced stage is
under investigation. Many molecularly targeted agents with
novel mechanisms of action are under active development.
decision. Every effort should be made to ensure a careful

evaluation of whether the patient is a candidate for surgical
resection or liver transplant. A referral to a tertiary medical
center is often helpful to render a definitive decision. A
multidisciplinary team evaluation is critical to assess the
pros and cons of each treatment option for individual patients, particularly in the setting of assessing local-regional
therapy. The clinicians should be familiar with the indication of sorafenib and how to manage the sorafenib-related
side effects. It is important to be aware of the ongoing
clinical trials available for HCC and to support the enrollment of HCC-specific clinical trials. When the patients have
advanced HCC with poor performance status or worsening
cirrhosis, symptomatic management and preserving the
quality of life of these patients should be the primary
considerations.
Numerous staging systems for HCC have been developed.
These systems are useful for stratification of patients
based on their prognosis, allocating specific treatment
based on the stage, and allowing comparison of clinical
outcomes from different clinical studies. Many different
staging systems have been developedBarcelona Clinic
Liver Cancer (BCLC),5 Cancer of the Liver Italian Program
(CLIP),6 tumornodemetastasis (tumor, node, metastasis
staging system),7 Groupe dEtude et de Traitement du
Carcinome Hepatocellulaire (GRETCH),8 Chinese University Prognostic Index (CUPI),9 and Japanese Integrated
Staging (JIS)10and there is currently no universally accepted staging system. The BCLC staging classification
is increasingly used in western regions, and it tries to
capture the tumor features, severity of cirrhosis, performance status, and a recommended treatment algorithm for
each stage. However, because of the geographic variation of
different risk factors, one staging system may perform
better than others in certain regions. In addition, depending
on the stage of the disease, certain staging system may
be more prognostic, as suggested by one report comparing
the various staging systems for patients with advanced
disease.11 In this study, the BCLC system was found to be
From the Massachusetts General Hospital Cancer Center, Harvard Medical School,
Boston, MA.
Address reprint requests to Dr. Andrew X. Zhu, Massachusetts General Hospital Cancer
Center, 55 Fruit Street, LH/POB 232, Boston, MA 02114; email: azhu@partners.org.
1092-9118/10/1-10
275
ANDREW X. ZHU
less informative than the GRETCH and CLIP classifications.11 Clinicians should become familiar with some of
these staging systems, their limitations, and controversies
in the assessment of HCC. Although genomic analysis and
molecular markers have been used to identify possible
sub-classification of HCC, these findings would require additional validation and have not been incorporated in clinical staging.
Management of Early-Stage HCC
Despite some variation of practice patterns worldwide, it

is clear that early-stage HCC can be cured by several
treatment options including surgical resection, liver transplantation, and ablative therapies (Table 1).
The aim of surgical resection is to remove the entire portal
territory of the neoplastic segment(s) with a clear margin,
while preserving maximum liver parenchyma to avoid hepatic failure. Because of the presence of extrahepatic disease, severe underlying cirrhosis, anatomic location of
tumor, and vascular invasion, less than 20% of patients with
HCC are suitable for surgical resection. Resection and
transplantation achieve the best outcomes in well-selected
candidates (5-year survival of 60% to 80%) and compete as
the first option in patients with early tumors on an
intention-to-treat perspective. Hepatic resection is the treatment of choice for HCC in noncirrhotic patients,12 where
major resections can be performed with low rates of lifethreatening complications and acceptable outcome (5-year
survival: 30% to 50%).
Orthotopic liver transplantation is considered to be the
first-line treatment option for patients with single tumors
less than 5 cm or 3 nodules or less with each measuring 3 cm
or less (Milan criteria) and advanced liver dysfunction not
suitable for resection. In a landmark study by Mazzaferro
KEY POINTS
276
Management of hepatocellular carcinoma (HCC) requires a multidisciplinary approach with careful

evaluation of the extent of the tumor, the presence
and severity of underlying cirrhosis, and performance
status before a treatment decision can be made.
Surgical resection, liver transplantation, and radiofrequency represent curative treatment options for
HCC. There is no established role of adjuvant therapy.
Transarterial chemoembolization and other localregional therapies including drug-eluting beads, radioembolizaton, and radiation are evolving treatment
options for unresectable HCC. The role of sorafenib,
when given in combination with these treatments, is
under investigation.
Sorafenib remains the only agent approved for advanced HCC. Careful selection of patients and timely
management of side effects are important to optimize
the efficacy of sorafenib.
Molecularly targeted agents are under active development and hold promise to improve the outcomes in
patients with HCC.
Table 1. Management of Hepatocellular Carcinoma

Early stage:
Surgical resection
Transplantation
RFA, PEI
Intermediate stage:
TACE
drug-eluting beads
radioembolization
radiation
Advanced stage:
Sorafenib
Clinical trials
Best supportive care
Abbreviations: RFA, radiofrequency ablation; PEI, percutaneous ethanol injection; TACE, transarterial chemoembolization.
and colleges, patients with HCC who meet the Milan criteria
have an expected 4-year overall survival rate of 85% and a
recurrence-free survival rate of 92%.13 Therefore, Milan
criteria have been adopted by the United Network for Organ
Sharing (UNOS) and widely around the world. Efforts have
been made to expand the transplant criteria, for example,
the University of San Francisco criteria (solitary HCC measuring up to 6.5 cm in diameter or up to three lesions, each
measuring no more than 4.5 cm in diameter, with a total
combined measurement of less than 8 cm) has been proposed
and used in select centers.14
For patients with small HCCs who are poor surgical
candidates because of impaired liver function or serious
comorbid medical conditions, local ablative therapy represents another attractive treatment option. RFA has become
the most commonly used local ablation therapy, as recent
randomized trials have shown RFA to be more effective than
percutaneous ethanol injection (PEI) in treating patients
with small HCC (23 cm in diameter), with lower rates of
local recurrence and higher rates of overall and disease-free
survival.15 In addition, a randomized controlled trial has
compared RFA with surgical resection and shown no significant differences in overall or recurrence-free survival, with
lower rates of complications and hospitalization associated
with RFA.16
An ongoing study is assessing the benefits of sorafenib
in the adjuvant setting following surgical resection and
RFA. Various bridging therapies including RFA and TACE
are continuing to be used while patients are waiting for
transplant, although the definitive benefits remain to be
defined (p 0.05).
Management of Intermediate-Stage HCC
For patients with intermediate-stage disease with multifocal lesions and without vascular invasion, TACE has
been the default treatment option (Table 1). TACE takes
advantage of the fact that HCCs derive their blood supply
almost entirely from the hepatic artery. The experience
with TACE has been mixed, leaving many unanswered
questions and controversies. Although several studies
have shown negative survival benefits, two studies have
demonstrated improved OS compared with best supportive
care (BSC) alone in highly selective patient populations.
In a randomized controlled trial, Llovet and colleagues
demonstrated that patients (more than 80% with underlying
HCV-related cirrhosis) who received doxorubicin-based
EVOLVING TREATMENT OF HCC

Table 2. Phase III Studies of Sorafenib in Hepatocellular
Carcinoma
SHARP Sorafenib
versus Placebo3
Endpoint
OS
TTSP
TTP
RR
Hazard Ratio
(95% CI)
10.7 vs. 7.9 mo
0.69 (0.550.87)
1.08 (0.881.31)
5.5 vs. 2.8 mo
0.58 (0.450.74)
2% vs. 1%
P value
0.001
0.768
0.001
Asia-Pacific Sorafenib
versus Placebo4
Hazard Ratio
(95% CI)
6.5 vs. 4.2 mo
0.68 (0.500.93)
0.90 (0.671.22)
2.8 vs. 1.4 mo
0.57 (0.420.79)
3.3% vs. 1.3%
P value
0.014
0.50
0.001
Abbreviations: OS, overall survival; TTSP, time to symptomatic progression;

TTP, time to tumor progression; RR, response rate.
TACE had improved OS compared with those who received

BSC (p 0.009).17 Survival probabilities at 1 year and 2
years were 82% and 63% for chemoembolization and 63%
and 27% for control. In another single-center study conducted in Hong Kong where the majority of patients had
underlying HBV infection, Lo and colleagues showed that
patients with unresectable HCC who received cisplatinbased TACE had improved survival (1 year, 57%; 2 years,
31%; 3 years, 26%) compared with those who received only
symptomatic control (1 year, 32%; 2 years, 11%; 3 years,
3%; p 0.002).18 A meta-analysis of randomized, controlled
trials assessing the use of arterial embolization, chemoembolization, or both as primary palliative treatment for
HCC showed that these procedures were associated with an
improved 2-year survival rate as compared with conservative treatment.19
In addition to conventional TACE, several novel regional
therapies, including drug-eluting beads TACE, radioembolization, and external beam radiation, are under active investigation in HCC. Whether certain techniques will perform
better than others and how to incorporate regional therapy
for patients with portal vein invasion requires additional
studies. Current clinical studies are also assessing the role
of sorafenib given concurrently or following TACE or other
regional treatment modalities. Despite the early studies
demonstrating the tolerability of combining sorafenib with
TACE,20,21 this approach has not definitively shown clinical
benefits of sorafenib either following TACE or concurrently
with TACE.22,23
Management of Advanced-Stage HCC
Following the initial experience in a phase II study,24 two

randomized phase III trials definitively demonstrated the
improved OS benefit of sorafenib in advanced HCC (Table 2).
The approval and wide application of sorafenib has changed
the treatment paradigm for HCC. However, as sorafenib is
gaining more clinical experience, several important findings
and related questions have emerged. First, the clinical
benefits are modest and only seen in certain patients. This
highlights the importance of understanding the mechanism
of action of sorafenib and identifying potential predictive
markers. Second, as sorafenib-related toxicities, including
hand and foot skin reaction, diarrhea, and fatigue can be
challenging to manage and affect the quality of life of
patients, many clinicians and investigators have asked the
relevant question: what is the optimal dose of sorafenib?
Although the targeted dose of sorafenib should be 400 mg
twice daily based on phase III data, many clinicians have

adopted a dose titration step-up approach, starting at 400
mg daily and increase the dosage by 200 mg every 12 weeks
to the targeted 800 mg daily as tolerated. Experience form
the observational GIDEON study showed that 34% of patients in the United States started sorafenib at 400 mg daily.
Third, because the agent was tested only in patients with
underlying Child A cirrhosis in the phase III trials, the
benefits of sorafenib in patients with worsening hepatic
dysfunction remains uncertain. Several institutional-based
retrospective studies have examined the use of sorafenib
in patients with Child B cirrhosis. Although these studies
have their inherent limitation, these findings suggest that
sorafenib can be safely given to most patients with underlying Child B cirrhosis. However, increased hyperbilirubinemia and other side effects can be encountered at higher
frequency. The pharmacokinetic (PK) parameters are similar or modestly different in patients with Child B in comparison with those with Child A. The treatment duration
and OS are generally shorter in Child B than those with
underlying Child A cirrhosis. Based on the available data,
the starting dose for patients with good performance status
and compensated hepatic function should be 800 mg daily.
However, in patients with borderline performance status
and compromised hepatic function, a reduced dose of
sorafenib at 200 400 mg can be started with the goal of
escalating to full dosage if tolerated. The PK study of
sorafenib in patients with hepatic and renal dysfunction also
provides some guideline for dosing for these patients.
New Targets and Regimens under Development for
Advanced HCC
The approval of sorafenib has greatly stimulated research

in drug development in HCC. Many molecularly targeted
agents that inhibit different pathways of hepatocarcinogenesis are under various phases of clinical development, and
novel targets are being assessed in HCC. There are three
main strategies in this area: (1) identifying and testing
targeted agents with novel mechanisms of action; (2) combining various targeted agents that blocks specific targets in
different or same pathways; (3) combining targeted agents
with chemotherapy. Despite the excitement for these extensive efforts in the past few years, we have observed a few
worrisome trends. First, all ongoing phase III studies with
targeted agents are conducted in unselected populations.
Second, most agents that are in phase III testing are not
based on robust data from randomized phase II studies.
Third, there is a high rate of failure of phase III trials in
HCC. Table 3 lists some of the key studies in various phases
of clinical development.
Antiangiogenic Agents
HCCs are vascular tumors, and increased levels of vascular endothelial growth factor (VEGF) and microvessel density (MVD) have been observed.25 VEGF is one of the main
inducers of liver tumor angiogenesis. High VEGF expression
has been associated with lower survival. Inhibition of angiogenesis represents a potential therapeutic strategy and has
been extensively tested in HCC.
Several VEGF-R inhibitors have moved to phase III development based on the initial phase II data.25 Sunitinib is
an oral multikinase inhibitor that targets receptor tyrosine
277
ANDREW X. ZHU
Table 3. New Agents/Regimens under Development in Advanced
Hepatocellular Carcinoma 25
Agents/Regimen
Phase III, first-line

Phase III, second-line
Sorafenib/erlotinib
Brivanib (failed)
Sorafenib/doxorubicin
Everolimus
Sunitinib (failed)
Ramucirumab
Brivanib
ADI-PEG 20
Linifanib
Phase I-II
Single-agent study:
Antiangiogenic agents
Sunitinib, brivanib, bevacizumab, ramucirumab, TSU-68, linifanib,
cediranib, pazopanib, lenvatinib, lenalidomide, and axitinib
Epidermal growth factor receptor inhibitors
Erlotinib, gefitinib, lapatinib, cetuximab
mTOR inhibitors
Everolimus, temsirolimus, and sirolimus
c-Met inhibitors
Tivantinib (ARQ197), and cabozantinib (XL184)
MEK inhibitors
Selumetinib (AZD6244)
Histone deacetylase inhibitor
Belinostat
HSP-90 inhibitor
STA-9090
Combined targeted agents
Bevacizumaberlotinib, sorafenibeverolimas, sorafenibAZD6244,
sorafenibbevacizumab, sorafenib temsirolimus, sorafenib and vorinostat,
sorafenib GC33, sorafenib OSI-906
results from an open label, multicenter phase II study of

linifanib in advanced HCC were reported. Linifanib was
given at 0.25 mg/kg daily in Child-Pugh A. For all 34
patients, median TTP was 112 days and median OS was
295 days (95% CI, 182333). A phase III study comparing
linifanib with sorafenib should complete in the near future.
Ramucirumab (IMC-1121B), a recombinant human monoclonal antibody against VEGFR-2, has been examined in a
phase II study. This demonstrated a response rate (RR) of
10%, PFS of 4.0 months, and OS of 12.0 months in patients who have not received prior systemic therapy. Remucirumab is undergoing phase III development in the secondline setting against placebo in patients where sorafenib
failed.
Several other antiagiogenic agents are at early stages of
clinical development in HCC. These include bevacizumab,29
cediranib, pazopanib, lenvatinib, and axitinib. The abundance of VEGF inhibitors that entered HCC clinical trials
and the failure of sunitinib and brivanib in phase III trials
have prompted us to reconsider the relevance of targeting
VEGF receptors in HCC. The challenge of developing additional antiangiogenic agents in HCC is to understand the
mechanisms of action and develop potential surrogate and
predictive markers to identify patients likely to benefit from
treatment. Although we aim to select the most potent
inhibitors, the safety profiles of these agents including
bleeding, thromboembolic events, skin rashes, and hypertension should be carefully evaluated.
Epidermal Growth Factor Receptor (EGFR) Inhibitors
kinases (RTKs) including VEGFR-1, VEGFR-2, PDGFRalpha/beta, c-KIT, FLT3, and RET kinases. Following the
initial experience of sunitinib from several single-arm phase
II studies that used three different dose schedules, a randomized phase III study comparing sunitinib at 37.5 mg
continuous daily dosing with sorafenib at 400 mg twice daily
in advanced HCC was conducted. Unfortunately, in this
large study of 1,073 patients, sunitinib failed to demonstrate
either superiority or noninferiority in OS when compared
with sorafenib.26 Toxicities including thrombocytopenia and
neutropenia were seen with sunitinib leading to discontinuation. Brivanib alaninate is a dual inhibitor of VEGFR and
fibroblast growth factor receptor (FGFR) signaling pathways
that can induce tumor growth inhibition in mouse HCC
xenograft models. A phase II study was conducted to assess
the efficacy and safety of brivanib in patients with unresectable, locally advanced, or metastatic HCC who had received
either no prior systemic therapy (Cohort A, 55 patients) or
one prior regimen of angiogenesis inhibitor (Cohort B, 46
patients). Treatment schedule consisted of continuous daily
dosing of brivanib at 800 mg. Both schedules reported
preliminary evidence of antitumor activity. Median
progression-free survival (PFS) and OS were 2.7 (95% CI,
1.4 3.0) and 10 months (95% CI, 6.8 15.2) in the first-line
study.27 Median OS and time to progression (TTP) as assessed by study investigators following second-line treatment with brivanib were 9.79 and 2.7 months,
respectively.28 Despite the ambitious phase III development
program with brivanib in HCC, a recent press release
reported that brivanib failed to demonstrate improved OS
when compared with placebo in patients with advanced
HCC who failed sorafenib. Linifanib (ABT-869) is a potent
and selective inhibitor of VEGFR and PDGFR. Preliminary
278
Increasing evidence has highlighted the importance of

EGFR/HER1 and its ligands EGF and transforming growth
factor-alpha (TGF-alpha) in hepatocarcinogenesis. The expression of several EGF family members, specifically EGF,
TGF-alpha, and heparin-binding epidermal growth factor,
as well as EGFR, has been described in several HCC cell
lines and tissues. Multiple strategies to target EGFR signaling pathways have been developed, and two classes of
anti-EGFR agents are tested in HCC: monoclonal antibodies
that competitively inhibit extracellular endogenous ligand
binding and small molecules that inhibit the intracellular
tyrosine kinase (TK) domain. With the exception of erlotinib
showing modest activity in single-arm studies, the other
EGFR inhibitors (gefinitib, lapatinib, and cetuximab) have
not demonstrated convincing antitumor activity as single
agents. Only erlotinib is being examined in an ongoing
phase III study in combination with sorafenib.
mTOR Inhibitors
Mammalian target of rapamycin (mTOR) functions to

regulate protein translation, angiogenesis, and cell cycle
progression in many cancers including HCC. Preclinical
data have demonstrated that mTOR inhibitors were effective in inhibiting cell growth and tumor vascularity in HCC
cell lines and HCC tumor models. Aberrant mTOR signaling
was present in half of the HCC cases and correlated with
HCC recurrence following resection.
A number of mTOR inhibitors (sirolimus, temsirolimus,
and everolimus) are under clinical development in HCC.
Early evidence of tolerability and efficacy has emerged from
phase I and II studies with everolimus, which has led to the
ongoing randomized phase III study comparing everolimus
EVOLVING TREATMENT OF HCC
against placebo for patients who failed or could not tolerate

sorafenib.
Hepatocyte Growth Factor (HGF)/c-Met Inhibitors
Dysregulation of c-Met is seen in HCC, and silencing the

expression of c-Met inhibits HCC growth in HCC cell lines
and tumor models. Tivantinib (ARQ197), a selective, nonATP-competitive inhibitor targeting MET tyrosine kinase, is
under early clinical evaluation. In a press release, tivantinib
reportedly met the primary endpoint of improving TTP in a
randomized phase II study comparing tivantinib with placebo in previously treated patients. Cabozantinib (XL184), a
dual c-Met/VEGFR-2 inhibitor, also demonstrated early evidence of antitumor activity in a randomized discontinuation
phase II study with a median PFS of 4.2 months.
MEK Inhibitors
HCC is characterized by frequent MEK/ERK activation in

the absence of RAS or RAF mutation. A multicenter, singlearm study with a two-stage design was conducted using
selumetinib (AZD6244), a specific inhibitor of MEK, in
advanced HCC.30 The primary endpoint was RR. No radiographic responses were seen and TTP was only 8 weeks
suggesting minimal single-agent activity despite evidence of
inhibition of ERK phosphorylation.
Other Molecularly Targeted Agents under

Development in HCC
Multiple genetic and epigenetic changes occur during

hepatocarcinogenesis. These pathways include the PI3K/Akt
pathway, insulin growth factor (IGF) and its receptor
(IGFR), as well as the Wnt/beta-catenin pathway. Multiple
agents targeting these key pathways are under early-stage
evaluation in HCC. The major challenge for future development of these targeted agents is the identification of predictive markers for specific drugs so that the targeted
population can be enriched.
Conclusion
Management of HCC continues to be challenging, but

potential treatment options are evolving. A multidisciplinary evaluation will allow physicians to make the best
treatment decisions for patients. As we continue to improve
the outcomes of curative treatment with resection, liver
transplant, and RFA, many new liver-directed regional
therapies are emerging. Sorafenib remains the only approved agent for advanced HCC, and its role in the adjuvant
setting following resection or RFA, with TACE, or in combination with other targeted agents or chemotherapy in the
advanced stage is under investigation. Many targeted
agents with novel mechanisms of action are under active
development.
Author
Andrew X. Zhu
Employment or
Leadership
Positions
Consultant or
Advisory Role
Bristol-Myers
Squibb; Pfizer;
Sanofi
Stock
Ownership
Honoraria
Research
Funding
Bayer/Onyx;
ImClone Systems
Expert
Testimony
Other
Remuneration
REFERENCES
1. Parkin DM, Bray F, Ferlay J, et al. Global cancer statistics, 2002. CA
Cancer J Clin. 2005;55:74-108.
2. El-Serag HB. Hepatocellular carcinoma. N Engl J Med. 365:1118-1127.
3. Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med. 2008;359:378-390.
4. Cheng AL, Kang YK, Chen Z, et al. Efficacy and safety of sorafenib in
patients in the Asia-Pacific region with advanced hepatocellular carcinoma: A
phase III randomised, double-blind, placebo-controlled trial. Lancet Oncol.
2009;10:25-34.
5. Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: The
BCLC staging classification. Semin Liver Dis. 1999;19:329-338.
6. A new prognostic system for hepatocellular carcinoma: A retrospective
study of 435 patients: The Cancer of the Liver Italian Program (CLIP)
investigators. Hepatology. 1998;28:751-755.
7. Sobin LH. TNM, sixth edition: New developments in general concepts
and rules. Semin Surg Oncol. 2003;21:19-22.
8. Chevret S, Trinchet JC, Mathieu D, et al. A new prognostic classification
for predicting survival in patients with hepatocellular carcinoma. Groupe
dEtude et de Traitement du Carcinome Hepatocellulaire. J Hepatol. 1999;
31:133-141.
9. Leung TW, Tang AM, Zee B, et al. Construction of the Chinese University Prognostic Index for hepatocellular carcinoma and comparison with the
TNM staging system, the Okuda staging system, and the Cancer of the Liver
Italian Program staging system: A study based on 926 patients. Cancer.
2002;94:1760-1769.
10. Kudo M, Chung H, Osaki Y. Prognostic staging system for hepatocellular carcinoma (CLIP score): its value and limitations, and a proposal for
a new staging system, the Japan Integrated Staging Score (JIS score).
J Gastroenterol. 2003;38:207-215.
11. Huitzil-Melendez FD, Capanu M, OReilly EM, et al. Advanced hepa-
tocellular carcinoma: Which staging systems best predict prognosis? J Clin

Oncol. 2010;28:2889-2895.
12. Lang H, Sotiropoulos GC, Domland M, et al. Liver resection for
hepatocellular carcinoma in non-cirrhotic liver without underlying viral
hepatitis. Br J Surg. 2005;92:198-202.
13. Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the
treatment of small hepatocellular carcinomas in patients with cirrhosis.
N Engl J Med. 1996;334:693-699.
14. Yao FY, Ferrell L, Bass NM, et al. Liver transplantation for hepatocellular carcinoma: Expansion of the tumor size limits does not adversely impact
survival. Hepatology. 2001;33:1394-1403.
15. Cho YK, Kim JK, Kim MY, et al. Systematic review of randomized
trials for hepatocellular carcinoma treated with percutaneous ablation therapies. Hepatology. 2009;49:453-459.
16. Chen MS, Li JQ, Zheng Y, et al. A prospective randomized trial
comparing percutaneous local ablative therapy and partial hepatectomy for
small hepatocellular carcinoma. Ann Surg. 2006;243:321-328.
17. Llovet JM, Real MI, Montana X, et al. Arterial embolisation or
chemoembolisation versus symptomatic treatment in patients with unresectable hepatocellular carcinoma: A randomised controlled trial. Lancet. 2002;
359:1734-1739.
18. Lo CM, Ngan H, Tso WK, et al. Randomized controlled trial of
transarterial lipiodol chemoembolization for unresectable hepatocellular carcinoma. Hepatology. 2002;35:1164-1171.
19. Bruix J, Sala M, Llovet JM. Chemoembolization for hepatocellular
carcinoma. Gastroenterology. 2004;127:S179-188.
20. Dufour JF, Hoppe H, Heim MH, et al. Continuous administration of
sorafenib in combination with transarterial chemoembolization in patients
with hepatocellular carcinoma: Results of a phase I study. Oncologist.
2010;15:1198-1204.
21. Pawlik TM, Reyes DK, Cosgrove D, et al. Phase II trial of sorafenib
279
ANDREW X. ZHU
combined with concurrent transarterial chemoembolization with drug-eluting
beads for hepatocellular carcinoma. J Clin Oncol. 2011;29:3960-3967.
22. Kudo M, Imanaka K, Chida N, et al. Phase III study of sorafenib after
transarterial chemoembolisation in Japanese and Korean patients with
unresectable hepatocellular carcinoma. Eur J Cancer. 2011;47:2117-2127.
23. Lencioni R LJ, Han G, Tak WY, et al. Sorafenib or placebo in
combination with transarterial chemoembolization (TACE) with doxorubicineluting beads (DEBDOX) for intermediate-stage hepatocellular carcinoma
(HCC): Phase II, randomized, double-blind SPACE trial.. J Clin Oncol.
2012;30(suppl 4; abstr LBA154).
24. Abou-Alfa GK, Schwartz L, Ricci S, et al. Phase II Study of Sorafenib in
Patients With Advanced Hepatocellular Carcinoma. J Clin Oncol. 2006;10:
4293-4300.
25. Zhu AX, Duda DG, Sahani DV, et al. HCC and angiogenesis: Possible
targets and future directions. Nat Rev Clin Oncol. 2011;8:292-301.
280
26. Cheng A, Kang YK, Lin D, et al. Phase III trial of sunitinib (Su) versus
sorafenib (So) in advanced hepatocellular carcinoma (HCC). J Clin Oncol.
2011;29(suppl; abstr 4000).
27. Park JW, Finn RS, Kim JS, et al. Phase II, open-label study of brivanib
as first-line therapy in patients with advanced hepatocellular carcinoma. Clin
Cancer Res. 2011;17:1973-1983.
28. Finn RS, Kang YK, Mulcahy M, et al. Phase II, Open-label Study of
Brivanib as Second-line Therapy in Patients with Advanced Hepatocellular
Carcinoma. Clin Cancer Res. 2012; Epub ahead of print.
29. Siegel AB, Cohen EI, Ocean A, et al. Phase II trial evaluating the
clinical and biologic effects of bevacizumab in unresectable hepatocellular
carcinoma. J Clin Oncol. 2008;26:2992-2998.
30. ONeil BH, Goff LW, Kauh JS, et al. Phase II study of the mitogenactivated protein kinase 1/2 inhibitor selumetinib in patients with advanced
hepatocellular carcinoma. J Clin Oncol. 2011;29:2350-2356.
Advanced Biliary Tract Cancers

By Laura Williams Goff, MD, and Jordan D. Berlin, MD
Overview: Single-agent management of metastatic biliary

tract cancers with 5-fluorouracil (5-FU) or gemcitabine has
shown limited efficacy, although 5-FU has been shown to be
more effective than best supportive care alone. An analysis
of phase II trials has suggested that platinums enhanced
the efficacy of single-agent fluoropyrimidines. In a phase III
randomized trial comparing single-agent gemcitabine with
gemcitabine plus cisplatin, the gemcitabine/cisplatin combination significantly improved median overall survival (OS)
ILIARY TRACT cancers consist of a somewhat heterogeneous group of tumors that include gallbladder cancer, extrahepatic biliary tract cancer, and intrahepatic
cholangiocarcinoma. The exact incidence of each of these
cancers is difficult to discern from annual cancer statistics
because, for example, intrahepatic cholangiocarcinoma is
still combined with hepatocellular carcinoma despite the
fact that these are biologically distinct entities. In 2011,
9,250 new cases and 3,300 deaths from biliary tract cancers
and gallbladder cancer (excluding intrahepatic cholangiocrcinoma) were anticipated.1 Cancers of the biliary tract are
often found at late stages when resection is not feasible and
treatment options are limited. Overall, 5-year survival rates
are estimated to be approximately 15%.
and progression-free survival (PFS), which established a new

option for standard of care. However, the future of cancer
medicine lies in newer, targeted agents. In the management of
biliary tract cancers, preliminary evidence with epidermal
growth factor receptor inhibitors has already demonstrated
activity. This article reviews systemic therapies for metastatic
biliary tract cancers as they relate to current and emerging
standards of care.
KEY POINTS
Single-agent management of advanced biliary cancers with 5-fluorouracil or gemcitabine yields short
survival times.
Adding cisplatin to gemcitabine significantly improves progression-free survival and overall survival
for patients with biliary tract cancers.
Targeted therapies are being evaluated in patients
with biliary tract cancers, showing some preliminary
evidence of activity for epidermal growth factor receptor inhibition.
The systemic treatment of biliary tract cancers has largely

been based on cytotoxic chemotherapy. Data comparing
5-FU based chemotherapy with best supportive care demonstrated that median survival times are significantly prolonged (6.0 months vs. 2.5 months) with treatment.2 In
addition, the time before declines in quality of life was
prolonged with 5-FU based chemotherapy. Rates of response to either single-agent 5-FU or 5-FU with leucovorin
range up to approximately 20% with survival times up to
8 months. Similarly, single-agent gemcitabine has been
explored as an alternative treatment option for biliary tract
cancers. Response rates for single-agent gemcitabine have
ranged from 16% to 36% and survival times from 6 months
to 16 months.3 Phase II study of combinations of gemcitabine with either 5-FU or its oral prodrug, capecitabine, has
not clearly improved on the results of single-agent phase II
data.4 The most promising results came from the combination of gemcitabine and capecitabine with a response rate
of 31% in 45 patients and an OS of 14 months.5
However, studies combining 5-FU or gemcitabine with
platinums have yielded very promising response rates.4
Therefore, a study was undertaken to evaluate the effects of
gemcitabine plus cisplatin compared to gemcitabine alone.6
This study had an initial randomized phase II portion
(ABC-01) and, when its endpoints were met, the trial proceeded to phase III (ABC-02). The ABC-02 trial demonstrated a significant benefit in both response rate and PFS
favoring the gemcitabine/cisplatin arm. Most importantly,
OS increased from 8.1 months for those treated with singleagent gemcitabine to 11.7 months for those treated with
gemctiabine/cisplatin. These results have established the
standard of care for biliary tract cancers at the current time.
Targeted Agents
Because of the limited likelihood that cytotoxic therapy

will substantially change the natural history of biliary tract
cancers, investigation is now focusing on the use of targeted
agents. One of the first agents studied in biliary tract
cancers was erlotinib, an oral tyrosine kinase inhibitor of
the epidermal growth factor receptor (EGFR). EGFR is
overexpressed in most biliary tract cancers.7 In a singleagent, multi-institutional study, erlotinib produced rare
responses (8%) in 42 patients.8 Although the median time to
tumor progression was only 2.6 months, 17% of patients
were progression-free at 6 months and OS was 7.5 months.
Because adding EGFR inhibition to chemotherapy has
worked in other diseases such as colorectal cancer, investigators have added cetuximab to gemcitabine and oxaliplatin in a phase II study. A small randomized trial
comparing the combination of gemcitabine plus oxaliplatin
with gemcitabine/oxaliplatin plus erlotinib has been reported.9 In the 133 randomly assigned patients, there was a
trend toward a PFS benefit from adding erlotinib, but OS
was identical for both arms at 9.5 months. However, the
EGFR pathway is actually a complex network and has
interactions with several other pathways. Others have tried
From the Vanderbilt Ingram Cancer Center, Nashville, TN.

Address reprint requests to Jordan D. Berlin, MD, Vanderbilt Ingram Cancer Center, 777
Preston Research Building, 2220 Pierce Ave., Nashville, TN 37232-6307, email: jordan.
berlin@vanderbilt.edu.
1092-9118/10/1-10
281
BERLIN AND GOFF
to optimize the use of erlotinib by adding other targeted

agents. One phase II study of combined EGFR and vascular
endothelial growth factor inhibition with erlotinib and bevacizumab showed promising results.10 In 53 patients, six
(12%) patients had confirmed partial responses. Median
time to progression was 4.4 months and OS was 9.9 months.
Other investigators have evaluated targeting different proteins downstream of EGFR for better results. Some evidence
of activity has been seen with MEK inhibition and biliary
tract cancers.11 However, better understanding of these
pathways in biliary tract cancer will help us to better use the
new, targeted therapies for patients with biliary tract cancers.
Conclusion
Management of biliary tract cancers has previously been

limited to fluoropyrimidines; however, over the last several
years, new two-drug combinations have been developed with
evidence of better activity than single-agent regimens. Although gemcitabine/cisplatin have been established as a
standard option for these cancers, other two-drug regimens
appear to have similar results. Now that the era of targeted
agents for biliary tract cancers is starting, it is important
that we improve our understanding of the biology of the
disease and how best to use these new classes of agents to
improve the outcomes for patients with these deadly diseases.
Author
Jordan D. Berlin
Laura Williams Goff
Employment or
Leadership
Positions
NCI (L)
Consultant or
Advisory Role
Amgen; Arno
Therapeutics;
AstraZeneca;
Celgene; Clovis
Oncology;
Genentech;
Novartis; Otsuka;
Roche; Sanofi
Abbott
Laboratories
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Bristol Myers
Squibb; OSI
Pharmaceuticals;
Sanofi
REFERENCES
CA: Cancer J Clin. 2011;61:212-236.
2. Glimelius B, Hoffman K, Sjoden PO, et al. Chemotherapy improves
survival and quality of life in advanced pancreatic and biliary cancer. Ann
Oncol. 1996;7:593-600.
3. Anderson CD, Pinson CW, Berlin J, et al. Diagnosis and treatment of
cholangiocarcinoma. Oncologist. 2004;9:43-57.
4. Eckel F, Schmid RM. Chemotherapy in advanced biliary tract carcinoma: a pooled analysis of clinical trials. Br J Cancer. 2007;96:896-902.
5. Knox JJ, Hedley D, Oza A, et al. Combining gemcitabine and capecitabine in patients with advanced biliary cancer: a phase II trial. J Clin Oncol.
2005;23:2332-2338.
6. Valle JW, Wasan HS, Palmer DD, et al. Gemcitabine with or without
cisplatin in patients (pts) with advanced or metastatic biliary tract cancer
(ABC): Results of a multicenter, randomized phase III trial (the UK ABC-02
trial). J Clin Oncol. 2009;27:15s (suppl; abstr 4503).
282
7. Lee CS, Pirdas A. Epidermal growth factor receptor immunoreactivitiy

in gallbladder and extrahepatic biliary tract tumours. Pathol Res Pract.
1995;191:1087-1091.
8. Philip PA, Mahoney MR, Allmer C, et al. Phase II study of erlotinib in patients with advanced biliary cancer. J Clin Oncol. 2006;24:30693074.
9. Lee J, Park SH, Chang HM, et al. Gemcitabine and oxaliplatin
with or without eroltinib in advanced biliary-tract cancer: A multicentre, open-label, randomised, phase 3 study. Lancet Oncol. 2012;13:181188
10. Lubner SJ, Mahoney MR, Kolesar JL et al. Report of a multicenter
phase II trial testing a combination of biweekly bevacizumab and daily
erlotinib in pateints with unresectable biliary cancer: A phase II consortium
11. Bekaii-Saab T, Phelps M, Li X, et al. A multi-institutional study of
AZD6244 in patients with advanced biliary cancers. Cancer Res. 2009; (suppl;
abstr LB 129)
BEST USE OF IMAGING TECHNIQUES, OLD AND

NEW, FOR GENITOURINARY CANCERS
IN CLINICAL PRACTICE
CHAIR
Walter M. Stadler, MD
Chicago, IL
SPEAKERS
Charles J. Ryan, MD
San Francisco, CA
Josef J. Fox, MD
New York, NY
Optimal Use of Imaging to Guide Treatment

Decisions for Kidney Cancer
By Walter M. Stadler, MD
Overview: Treatment monitoring for solid tumors in general

and for metastatic renal cancer in particular has been dominated by assessment of tumor burden via cross-sectional
imaging. This poses a special problem for the mammalian
target of rapamycin and vascular endothelial growth factor
pathway-directed agents used in this disease. The standard
RECIST metrics used to categorize response and progression are arbitrary and do not adequately capture the effect of
these agents. Other approaches, including use of relative
RECIST measures as a continuous variable, volumetric measurements, and functional assessments, such as dynamic
HE USE of imaging for therapeutic decision making in

oncology has traditionally been limited to assessment of
changes in tumor burden with treatment. The underlying
presumption is that cancer will continue to grow in the
absence of treatment and any tumor shrinkage is thus a
bona fide treatment effect. The original challenges were thus
considered to simply be a determination of the degree of
tumor shrinkage that can be reliably detected. These tumor
burden changes were then codified and categorized into
complete response, partial response, and progressive disease, with the remaining category labeled stable disease.
Over the past decade, however, this rather simplistic view
has been challenged by changes in both available therapies
and available imaging technologies. Renal cancer is a microcosm of these challenges, and the issues are not only important for management of patients with this disease but are
also illustrative of the challenges in oncology therapeutic
monitoring in general.
We discuss tumor burden monitoring in the context of
available renal cancer therapies, novel approaches to tumor
burden assessment, the additional data available from contrast enhanced imaging, and radioisotope imaging with an
emphasis on technetium bone scans and fluorodeoxyglucosepositron emission tomography(FDG-PET). In the context of
these discussions, we also briefly mention data on use of
baseline imaging data to select among various therapeutic
options.
Two-Dimensional Tumor Burden Assessment
In the early days of medical oncology and oncologic therapeutics, reliable detection of tumor shrinkage in the context of limited technology was a major challenge. In a now
classic study by Mortel and colleagues,1 the investigators
asked 16 experienced oncologists at an early ASCO meetings
to palpate 12 solid spheres of various sizes, two of which
were the same, that were beneath a thin mattress. The
principal objective was to determine the size difference that
could be reliably detected. The authors demonstrated that
the product of the orthogonal bi-dimensional diameter of a
sphere had to be at least 50% smaller in order for the error
in detecting a difference to be 10%. They thus proposed
that this 50% shrinkage would be necessary before a therapeutic response could be declared. As cross-sectional imaging was developed, this metric was arbitrarily translated
to these new modalities. The metric was further simplified
284
contrast-enhanced magnetic resonance imaging-based quantitative variables and fluorodeoxyglucose-positron emission

tomography, have been proposed as alternatives, but the data
do not support their routine clinical use. Even fewer data are
available on the use of baseline imaging characteristics to
choose a specific therapy. Therefore, until further research on
imaging predictive and intermediate biomarkers matures, a
combination of standard cross-sectional imaging and clinical
judgment is the most pragmatic option for treatment decision
making for patients with metastatic renal cancer.
by the RECIST guidelines to translate the original required

bi-dimensional measurements to uni-dimensional measurements.2
Interestingly, these standards were promulgated without
any formal assessment as to the degree of tumor shrinkage
that could be reliably detected by these newer methods.
Even more importantly, there were essentially no formal,
statistically rigorous studies at the time that were designed
to determine whether any specified degree of tumor shrinkage could be used as an intermediate endpoint in trials.
More recently, studies have demonstrated the variability of
assessing uni-dimensional changes in the size of nodules in
well-aerated lung. In one such study, approximately 84% of
the measures were within a range of 10% to 10%, with
much greater relative variability among smaller tumors.3
Few if any studies have been performed with tumors in
other organs, and no studies have been done, to my knowledge, to specifically evaluate metastatic renal cancer. Perhaps more concerning is the fact that the clinical trial metric
of a partial response, which is based on a reliability
measurement of simulated palpation, has become a metric
for clinical decision making. Thus, a patient with 31% tumor
shrinkage is deemed to have a response and continues
therapy, whereas one with 29% shrinkage is deemed as
having stable disease and as such, discontinuation of
therapy might be considered.
The development of agents directed at the vascular endothelial growth factor (VEGF) pathway and the mammalian
target of rapamycin (mTOR) in renal cancer furthermore
made it clear that inhibition of tumor growth is not only a
therapeutic outcome but is also an outcome with likely
clinical relevance. To this end, it is notable that the standard
tumor burden criteria for progressive disease (20% increase
in uni-dimensional measurements) are completely arbitrary
and are not based on any experimental derivation. It is
perhaps thus not surprising that progression-free survival,
despite being the endpoint for the majority of phase III renal
cancer trials, and the basis of U.S. Food and Drug Admin-
From the University of Chicago, Chicago, IL.

Address reprint requests to Walter M. Stadler, MD, University of Chicago, 5814 S.
Maryland MC2115, Chicago, IL 60637; email: wstadler@medicine.bsd.uchicago.edu.
1092-9118/10/1-10
IMAGING BIOMARKERS IN RENAL CANCER
istration (FDA) approval for several drugs, has never been

demonstrated to be a surrogate endpoint for survival in this
disease. One of the few studies in which this progression
metric was formally evaluated (which has been reported in
abstract form only) demonstrated that there is a correlation
between progression-free survival and overall survival in
renal cancer but that the Kendalls tau statistic was 0.50,
which is considered a modest correlation and not sufficient
for general use as a surrogate endpoint.4
A number of new immune therapy regimens being developed might have activity in renal cancer as well. An important issue to consider with regard to therapeutic monitoring
of response to these drugs is that tumor shrinkage and
inhibition of tumor growth might be delayed, presumably
while the immune function is activated. As such, one needs
to allow for a period of growth before declaring such an agent
ineffective. Whether the criteria suggested for melanoma in
the context of ipilimumab therapy will apply in renal cancer
as well remains to be determined.5
reliably detected with a sample of approximately 150 patients.6

Although a detectable difference between an experimental
and a control group using tumor size measurements as a
continuous variable could be reliably ascribed to the intervention with use of standard statistical tests, such a difference clearly does not demonstrate that the difference has
clinical relevance. Thus, a continuous tumor size metric
could theoretically be used as a phase II clinical trial
endpoint but not a phase III trial endpoint without additional data elucidating the relationship between changes
in tumor size and clinically relevant endpoints. To this end,
we have conducted additional simulations on data from
phase III renal cancer trials of sorafenib and pazopanib and
suggest that such an approach, if used in a phase II trial,
would be highly predictive for the phase III results.7 Obviously, further work assessing the relationship between tumor burden and clinical outcome is necessary.
Use of Tumor Size Measurements as a

Continuous Variable
Challenges with analysis and interpretation of unidimensional and bi-dimensional measurements of tumor
size have led some authors to contend that volumetric
measurements would be more predictive of clinical outcome.
This belief is based on the observation that many tumors,
including metastatic renal cancer, do not always grow or
shrink in a spherical fashion. This observation raises the
hypothesis that the poor correlation between measurements
of tumor burden and clinically relevant outcomes such as
survival are due to the fact that the uni-dimensional and
bi-dimensional measurements do not accurately reflect tumor burden. Just as important, the manual extraction of
tumor burden data is cumbersome and, thus, the RECIST
guidelines exclude many lesions. More recently, a number of
algorithms have been developed to more accurately determine tumor volume from standard cross-sectional images.8
It is now becoming technically possible to address this
underlying hypothesis, and several studies are attempting
to do so.
Traditional tumor burden measurements using either

bi-dimensional or uni-dimensional criteria, and which were
derived from reliability measurements, categorized changes
in tumor size into variables labeled complete response,
partial response, stable disease, and progressive disease. The questionable clinical relevance of these categories
has already been noted. Just as importantly, categorization
of an otherwise continuous variable leads to loss of statistical information. Using actual RECIST-based measurements
rather than their aforementioned categorization, one can
compare two groups (e.g., a treated and an untreated group)
with use of standard statistical approaches for group comparisons. Because RECIST measurements are not normally
distributed, an initial transformation of the data (typically
log-transformed) and formal rules for how to manage missing measurements and difficult-to-measure lesions need to
be constructed. We undertook this exercise, and our simulations suggest that differences between an experimental
and a control group in terms of changes in tumor size can be
KEY POINTS
RECIST-based metrics of response, stable disease,

and progression are arbitrary.
In the context of the vascular endothelial growth
factor pathway and mammalian target of rapamycindirected therapy in renal cancer, the correlation
between RECIST-based disease progression and survival is modest.
Volumetric tumor measurements, dynamic contrastenhanced imaging quantitative parameters, and
functional imaging with radionuclides have all been
proposed as novel predictive and intermediate biomarkers for renal cancer therapy.
Standard cross-sectional imaging and clinical judgment remain the best pragmatic approach to therapeutic monitoring in renal cancer.
Novel Approaches to Tumor Burden Assessments
Contrast-Enhanced Imaging
With the advent of therapies directed at the VEGF pathway in renal cancer, it became quickly apparent that
contrast-enhanced imaging provides additional information
beyond tumor size alone. A number of investigators realized
that these targeted agents lead to a hypoperfused appearance of lesions, consistent with the mechanism of action of
the agents. It has thus been proposed that more formal
measurements of this hypoperfusion could not only be a
measure of drug effect (a pharmacodynamic endpoint) but
also an indication of clinical benefit (an intermediate or
surrogate endpoint). A number of criteria for assessing these
changes has been used, of which the most common are the
Choi Criteria, which classify a response as tumor shrinkage of at least 10% or a decrease in tumor density of at least
15%.9 Some studies have suggested a correlation of Choi
Response to VEGF pathway inhibitors with clinical outcome in patients with renal cancer.10 A major challenge with
these criteria, however, is that computerized tomography
(CT) scans are not always well standardized with regard
to the timing of image acquisition in relation to contrast
administration or in the contrast administration rate, which
are both critical variables in determining the degree of
285
WALTER M. STADLER
enhancement seen. Furthermore, the criteria are categoric

in nature, raising the same issues as discussed earlier for
categorization of tumor size metrics.
Investigators have thus attempted to conduct a more
formal analysis of uptake, distribution, and washout of
contrast agent in lesions using pharmacokinetic approaches.
Although this can be done with both contrast-enhanced CT
and magnetic resonance imaging (MRI), the requirement for
multiple images and associated radiation exposure, as well
as the limited resolution traditionally associated with CT
scans, has led to MRI being used in most of these studies. A
number of quantitative variables describing uptake, distribution, and washout of contrast agent can be derived, with
Ktrans, a complex variable dependent on both tumor blood
flow and perfusion, being the most commonly used variable.
Studies in patients with renal cancer have demonstrated
that VEGF pathway inhibitors decrease Ktrans in a reliably
detectable manner, but that there is little to no correlation
between this decrease and clinical outcome.11,12 Some have
suggested that baseline Ktrans might be predictive of outcome in the context of these therapies, but that hypothesis
remains to be proven.11,12
MRI-based measurements also have a number of additional limitations, including imaging time required to obtain
the parameters, difficulty in determining these quantitative parameters in more than one or a few lesions, heterogeneity in the quantitative parameter even within one
tumor, (which is technically challenging to measure), and
the nonlinear relationship between contrast enhancement
and contrast concentration, which is required for assessing
Ktrans and other parameters.
Alternative approaches using blood flow determined by
nuclear medicine techniques and molecular probes are being
developed, but the technology is currently immature, and
correlations with clinical outcome remains incompletely
determined.
Technetium Bone Scans
Bone scans have been used extensively in oncology imaging to determine the presence of metastases in bone. Given
the limitations of standard CT-based cross-sectional imaging in x-ray dense bone and the ability to image the entire
skeleton with a bone scan, a bone scan has been a valuable
modality for clinical decision making. Nevertheless, it needs
to be remembered that technetium is taken up at sites of
bone remodeling and thus the bone scan detects bone remodeling not the tumor per se. Other pathologic conditions,
including inflammation, Pagets disease, and fractures, can
all lead to increased bone remodeling and positive bone
scans. Furthermore, in renal cancer, sensitivity of bone
scintigraphy has been reported to be 62%.13 Lastly, dramatic antitumor effects might actually lead to an increase in
bone modeling and a worse-appearing bone scan. This has
been most prominently demonstrated in prostate cancer
(see article in the ASCO 2012 Educational Book by Biting
and Armstrong). Given these limitations, bone scans are a
poor modality for monitoring response to therapy in renal
cancer.
FDG-PET
The Warburg effect is one of the most ubiquitous and

earliest recognized features of malignancy. In this phenomenon, malignant cells preferentially utilize an aerobic glyco-
286
lysis pathway, which in turn leads to dramatic upregulation

of glucose transport. Fluorodeoxyglucose is a nonmetabolizable radioactive glucose analog that is taken up in cells by
these same glucose transporters. It is therefore a useful
imaging diagnostic marker for many cancers. Interestingly,
renal cancers, and especially clear cell renal cancers, do not
uniformly have elevated glucose uptake; in fact the sensitivity of FDG-PET for small pulmonary nodules that were
subsequently histologically demonstrated to be metastatic
renal cancer is 64%.14
The mTOR inhibitors are known to lead to hyperglycemia
in part because of a more general inhibition of glucose
uptake, and glucose uptake has been proposed to be a
pharmacodynamic biomarker for mTOR therapy. Perhaps
more interesting and related to the variable glucose uptake
in renal cancer specifically, it has been proposed that renal
cancers that have elevated glucose uptake would be preferentially sensitive to mTOR inhibition.15 We tested this
concept in a small single-arm trial and demonstrated that
treatment with everolimus reliably leads to a decrease in
FDG uptake, and is thus a pharmacodynamic biomarker.
However, baseline FDG measurements did not correlate
with patient outcome, and the correlation between the
degree of decrease in FDG uptake and patient outcome was
quite modest as well.16
Summary and Clinical Implications
Many of the discussions here are more directly applicable

to the use of imaging and imaging biomarkers in clinical
trial monitoring. Although their findings may not be directly
applicable to the clinical care scenario, the studies conducted to date provide some guidance. Most importantly,
these studies have strongly suggested that more sophisticated imaging methods such as FDG-PET or quantitative
parameters derived from measurements of contrast uptake
and washout are unlikely to be useful for monitoring response to treatment directed at the VEGF pathway or
mTOR or in selecting specific therapies. In addition, bone
scans, although useful from a diagnostic perspective, have
rather limited utility in therapeutic monitoring as well.
Thus, for clinical monitoring of response to therapies directed at VEGF and mTOR in renal cancer, we are left
essentially with standard cross-sectional imaging.
Because agents directed at the VEGF pathway and mTOR
lead to tumor shrinkage that does not always meet standard
RECIST criteria, or lead only to tumor growth inhibition,
lack of growth is typically used as a clinical indicator of
benefit. The obvious clinical challenge is that renal cancer
can be indolent, and it is often impossible to determine
whether any observed lack of growth is because of or despite
of the administered agent. Conversely, growth exceeding
the standard 20% above nadir used for determining progressive disease in clinical trials might still reflect a certain
degree of growth inhibition by the agent. Clinical management therefore requires an understanding of the patients
history, clinical disease symptoms, and clinical toxicities,
and a thorough review of the images themselves. Careful
clinical judgment is required to provide the patient with the
maximum treatment duration of an agent from which he or
she is benefiting, but no longer. Such judgment may become
even more challenging if immune-stimulating agents, from
which delayed responses may occur, become clinically available.
IMAGING BIOMARKERS IN RENAL CANCER
Author
Walter M. Stadler
Employment or
Leadership
Positions
UpToDate
Consultant or
Advisory Role
Stock
Ownership
AVEO; Caremark;
Genentech;
Millennium;
Novartis; Pfizer
Abbott
Laboratories (I)
Honoraria
Bayer/Onyx;
Clinical Care
Options; CME
Innovations;
Imedex; Pfizer;
Research-toPractice
Research
Funding
Expert
Testimony
Other
Remuneration
Active Biotech;
Bayer;
Boehringer
Ingelheim;
Bristol-Myers
Squibb;
Genentech;
GlaxoSmithKline;
ImClone
Systems; Infinity;
Lilly; Medarex;
Medivation;
Millennium;
Novartis; Pfizer;
Solvay; Takeda
REFERENCES
1. Moertel CG, Hanley JA. The effect of measuring error on the results of
therapeutic trials in advanced cancer. Cancer. 1976;38:388-394.
2. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to
evaluate the response to treatment in solid tumors. European Organization
for Research and Treatment of Cancer, National Cancer Institute of the
United States, National Cancer Institute of Canada. J Natl Cancer Inst.
2000;92:205-216.
3. Oxnard GR, Zhao B, Sima CS, et al. Variability of lung tumor measurements on repeat computed tomography scans taken within 15 minutes. J Clin
Oncol. 2011;29:3114-3119.
4. Halabi S, Rini BI, Stadler WM, et al: Use of progression-free survival
(PFS) to predict overall survival (OS) in patients with metastatic renal cell
carcinoma. J Clin Oncol. 2010;28;15s (suppl; abstract 4525).
5. Wolchok JD, Hoos A, ODay S, et al. Guidelines for the evaluation of
immune therapy activity in solid tumors: immune-related response criteria.
Clin Cancer Res. 2009;15:7412-7420.
6. Karrison TG, Maitland ML, Stadler WM, et al. Design of phase II cancer
trials using a continuous endpoint of change in tumor size: application to a
study of sorafenib and erlotinib in non small-cell lung cancer. J Natl Cancer
Inst. 2007;99:1455-1461.
7. Sharma MR, Karrison TG, Jin Y, et al. Resampling phase III data to
assess phase II trial designs and endpoints. Clin Cancer Res. Epub 2012 Jan
27.
8. Zhao B, Oxnard GR, Moskowitz CS, et al. A pilot study of volume
measurement as a method of tumor response evaluation to aid biomarker
development. Clin Cancer Res. 2010;16:4647-4653.
9. Choi H, Charnsangavej C, Faria SC, et al. Correlation of computed
tomography and positron emission tomography in patients with metastatic

gastrointestinal stromal tumor treated at a single institution with imatinib
mesylate: proposal of new computed tomography response criteria. J Clin
Oncol. 2007;25:1753-1759.
10. Krajewski KM, Guo M, Van den Abbeele AD, et al. Comparison of four
early posttherapy imaging changes (EPTIC; RECIST 1.0, tumor shrinkage,
computed tomography tumor density, Choi criteria) in assessing outcome to
vascular endothelial growth factor-targeted therapy in patients with advanced renal cell carcinoma. Eur Urol. 2011;59:856-862.
11. Flaherty KT, Rosen MA, Heitjan DF, et al. Pilot study of DCE-MRI to
predict progression-free survival with sorafenib therapy in renal cell carcinoma. Cancer Biol Ther. 2008;7:496-501.
12. Hahn OM, Yang C, Medved M, et al. Dynamic contrast-enhanced
magnetic resonance imaging pharmacodynamic biomarker study of sorafenib
in metastatic renal carcinoma. J Clin Oncol. 2008;26:4572-4578.
13. Sohaib SA, Cook G, Allen SD, et al. Comparison of whole-body MRI and
bone scintigraphy in the detection of bone metastases in renal cancer. Br J
Radiol. 2009;82:632-639.
14. Majhail NS, Urbain JL, Albani JM, et al. F-18 fluorodeoxyglucose
positron emission tomography in the evaluation of distant metastases from
renal cell carcinoma. J Clin Oncol. 2003;21:3995-4000.
15. Thomas GV, Tran C, Mellinghoff IK, et al. Hypoxia-inducible factor
determines sensitivity to inhibitors of mTOR in kidney cancer. Nat Med.
2006;12:122-127.
16. Chen JL, Appelbaum DE, Kocherginsky M, et al. FDG-PET as a
predictive marker for therapy with everolimus in metastatic renal cell cancer.
J Clin Oncol. 2011;29 (suppl; abstr e1504).
287
CASTRATION-RESISTANT PROSTATE CANCER: NEW

INSIGHTS INTO BIOLOGY AND TREATMENT (eQ&A)
CHAIR
Oliver Sartor, MD
Tulane Cancer Center
New Orleans, LA
SPEAKERS
Robert Evan Reiter, MD
Los Angeles, CA
Andrew J. Armstrong, MD, MSc
Duke Cancer Institute
Durham, NC
State-of-the-Art Management for the Patient

with Castration-Resistant Prostate Cancer
in 2012
By Oliver Sartor, MD
Overview: Much progress has been made in metastatic

castration-resistant prostate cancer (CRPC), and multiple new
U.S. Food and Drug Administration (FDA)-approved survivalprolonging drugs are now available. In 2004, docetaxel/prednisone was the first therapy shown to prolong survival. In 2010
and 2011, sipuleucel-T, cabazitaxel/prednisone, and abiraterone/prednisone were FDA approved. Two new agents,
radium-223 and MDV-3100, have recently reported large phase
III trials prolonging overall survival and will be submitted for
regulatory approval in 2012. One can now begin to ask, is there
an optimal sequence for therapies in metastatic CRPC? Despite the recent progress, there is much we do not know and
virtually no information on this important question. We know
that abiraterone/prednisone and cabazitaxel/prednisone are
appropriate choices for a patient after receiving docetaxel,
but we do not know what, if anything, represents the optimal
sequence for abiraterone and cabazitaxel. In fact we do not
understand how one therapy may affect the response to a
HE LAST several years have seen extraordinary progress in the management of patients with CRPC, with
multiple FDA approvals for agents that extend patients
overall survival. In addition there are several new anticipated FDA approvals that might occur in 2012. From a
perspective of understanding the current state of affairs, I
would first like to review the history of FDA approvals for
patients with metastatic CRPC. As seen in Fig. 1, all the
FDA drug approvals that occurred before 2004 involved
endpoints other than overall survival. In 2004, the combination of docetaxel/prednisone was approved on the basis of
a prolongation in survival (as compared with mitoxantrone/
prednisone), using data from the pivotal TAX327 study.1 A
separate study with docetaxel in 2004 (SWOG 9916), using
docetaxel/estramustine, also demonstrated superior survival in comparison to mitoxantrone/prednisone.2 Because of
the importance of these findings, from both a pragmatic and
regulatory perspective, the post-2004 world of metastatic
CRPC began to be divided into patients who had or had not
received prior docetaxel.
For the next 6 years, no substantial progress was made in
prolonging survival in metastatic CRPC. Since 2010, rather
remarkably, there have been four new FDA approvals and
three of these have involved agents that prolong survival.
Unlike a number of other diseases, the underlying mechanisms for these agents are distinct with a novel immunotherapy termed sipuleucel-T, a novel taxane called
cabazitaxel, and novel hormone therapy in the form of
abiraterone prolonging survival in various pivotal phase III
randomized trials.3-5 Two of these agents were FDA approved in the postdocetaxel setting (cabazitaxel/prednisone
and abiraterone/prednisone). No comparisons between any
of these agents have been performed and no direct comparison between docetaxel/prednisone plus an additional agent
has ever proven superior to docetaxel/prednisone alone.
A number of trials have tried to improve on docetaxel/
prednisone but none have succeeded. Combinations of do-
subsequent therapy. We are also aware that the pre- and

postdocetaxel spaces represent regulatory rather than biologic divisions. In addition, despite the proven role of docetaxel/prednisone, many patients with CRPC are not
considered to be suitable for chemotherapy, and worldwide
many never receive any form of chemotherapy. What is the
optimal management for these patients? Taken together it is
reasonable to assess patient preferences, prior therapies and
response/tolerance to prior therapies, burden of disease,
comorbidities, current symptoms, drug toxicities, out-ofpocket costs, etc., in clinical decision making. Given the many
factors we do not know, it is hard to be dogmatic in approaching the therapeutic options for the patient with CRPC. We will
likely soon move beyond the current sequencing paradigm and
begin to assess new combinations in a systematic and rational
fashion. Perhaps one day, in the not too distant future, we will
develop molecular stratification systems to better guide
therapeutic choices in CRPC.
cetaxel and DN-101 (calcitriol), GVAX, bevacizumab,

atrasentan, zibotentan, and lenalidomide all have failed in
phase III trials against docetaxel/prednisone. Dasatinib,
aflibercept, and custirsen are in current combination trials
with docetaxel/prednisone, and in each case the control arm
is docetaxel/prednisone. These agents target the src kinase
(dasatinib), vascular endothelial growth factor (aflibercept),
and clusterin (custirsen), and each of these targets are
supported by preclinical data in various systems. Whether
or not these combinations will be an improvement on the
current standard of care is unknown at the time this was
written, but some results may be available during the 2012
ASCO Annual Meeting. Both the dasatinib and aflibercept
trials are fully accrued and awaiting maturation of survival
data at this time. The trial with custirsen is still accruing as
of spring 2012.
One trial is going head to head with docetaxel/prednisone.
That trial will compare cabazatizel/prednisone (two doses of
cabazitaxel are being tested, 20 or 25 mg/m2) compared with
docetaxel/prednisone. The primary endpoint is overall survival. This trial (FIRSTANA) is currently accruing patients.
In 2011 and early 2012, two new agents have resulted in
prolonged survival in large phase III trials. These agents
include the novel targeted alpha-particle emitter radium223 and a novel antiandrogen MDV-3100.6,7 Taken together
there are now six agents that have prolonged survival in
phase III trials conducted in metastatic CRPC (Fig. 2).
As noted above, the disease state for these therapies has
varied. For instance, sipuleucel-T is indicated for use in the
asymptomatic or minimally symptomatic metastatic CRPC
From the Tulane Cancer Center, Tulane University School of Medicine, New Orleans, LA.
Address reprint requests to Oliver Sartor, MD, Tulane School of Medicine, 1430 Tulane
Ave., Box SL-42, New Orleans, LA 70112; email: osartor@tulane.edu.
1092-9118/10/1-10
289
OLIVER SARTOR
Fig. 1.
States.
FDA regulatory approvals in metastatic CRPC in the United
state. The eligibility requirements for the phase III trials

with MDV3100, cabazitaxel, and abiraterone all required
docetaxel pretreatment. The radium-223 study was unique,
though the majority of patients had received prior docetaxel,
the trial was also open to patients with bone-metastatic
CRPC, patients who were not deemed ineligible for docetaxel, and for those patients who refused docetaxel.6 This
is the first trial to evaluate this group of patients. Because
not all patients with metastatic CRPC go on to receive
docetaxel, this is an important group of patients to include in
prospective studies.8
There are several questions about the management of the
patients with metastatic CRPC. The first is a very practical
one: is there an optimal sequence for the FDA-approved
therapies? Is there a proper way to pick the right therapy for
the right patient at the right time? Is there any way to risk
stratify patients to maximize their opportunity for response?
The short answer to these reasonable questions is that we
are poorly informed on what might constitute an optimal
sequencing strategy for using new agents.
KEY POINTS
290
Patients with metastatic castration-resistant prostate cancer (CRPC) have more options than ever
before, including new agents that have been shown to
prolong survival.
New Food and Drug Administrationapproved agents
for CRPC that prolong survival include sipuleucel-T,
cabazitaxel, and abiraterone, and more are on the
way.
Optimal choices and sequences are much discussed in
CRPC but comparative data are absent; thus, dogmatic views of sequences are inappropriate.
Patient preferences, current symptoms, responses
and tolerance of prior therapies, pace of the disease,
burden of disease, drug toxicities, performance status, comorbidities, out-of-pocket costs, compliance,
and various clinical trial options are some of the
choices that guide practical considerations in treatment management.
Combination therapies are beginning to be explored
in a systematic fashion.
Fig. 2.
tages.
Phase III trials in metastatic CRPC with survival advan-
One can perhaps imagine that sipuleucel-T could be administered to a patient with small-volume asymptomatic
disease, followed by docetaxel at progression, followed by
cabazitaxel and/or abiraterone at progression. This would be
a logical series. Despite the excellent logic of such an
approach, there is very limited clinical data on patients that
might be treated in this manner. We have reports of docetaxel following sipuleucel-T, and certainly that appears to
be a safe combination, but we have no real data on abiraterone
response rates postcabazitaxel or cabazitaxel response rates
postabiraterone. Though MDV-3100 has yet to be approved by
regulatory authorities, it is not too soon to ask what will be the
activity of this agent postabiraterone (and vice versa). Will
resistance to one of the newer hormonal therapies result in
resistance to the other? We do not know.
At the same time that therapeutic decisions are made, it is
an important reminder that a bone-targeted therapy, such
as zoledronic acid or denosumab, might be integrated into
the treatment regimen for appropriate patients. External
beam radiation could be used to palliate focal pain on an as
needed basis. Growth factors and various other supportive
care options are also considerations in selected settings.
Taken together, there is little that we are certain of when
it comes to optimal drug sequencing in this disease state.
Further, the best choice of therapieswhen choices clearly
exist (such as the current postdocetaxel setting)is not at
all clear and one is simply left to conjecture. It is reasonable
to assess patient preferences, prior therapies and response
to prior therapies, performance status, pace of progression,
burden of disease, comorbidities, tolerance/intolerance of
prior therapies, drug toxicities, neuroendocrine status, and
current symptoms in clinical decision making. It is also
critical to understand the availability of clinical trials,
patient compliance, travel distances, and out-of-pocket
costs. Taken together, at this time, no simple algorithm can
suffice when it comes to making clinical decisions, and any
dogmatic approach to this problem cannot be justified.
We all realize that none of the therapies for CRPC to date
are curative, and in my practice, I try to ensure that patients
have the opportunity to be treated with as many of the active
therapies as possible during the course of their disease. Our
ability to predict who will and who will not respond to
various therapies is poor, and many of my grateful patients
have benefited from a treatment plan that was not judged a
priori to have a high rate of success. Perhaps one day our
molecular markers will guide us to the right choice of
THERAPEUTIC OPTIONS FOR PROSTATE CANCER
therapy at each juncture, but today we are often humbled by

how poorly we predict patient outcomes.
What about combination therapy? Is the metastatic CRPC
state to be restricted to a sequencing therapeutic paradigm?
Though combination therapy has a certain attraction in
CRPC, other than the agents designed to inhibit skeletalrelated events (zoledronic acid and denosumab), there are
minimal data on combination therapies using the survivalprolonging agents. Until new data exist, the current sequencing paradigm may be optimal for patients in nonclinical trial
settings.
It is clear from phase I/II studies that patients with
nondocetaxel-pretreated metastatic CRPC are quite
responsive to the newer hormonal agents, such as abiraterone and MDV-3100.9,10 This emphasizes that our division of metastatic CRPC into pre- and postdocetaxel spaces
is one based on regulatory and not biologic concerns. Trials

are ongoing to further assess both abiraterone and MDV3100 in the predoctaxel space and perhaps there will be
results from one of these trials by the 2012 ASCO Annual
Meeting.
All of the FDA-approved agents to date have been approved in a metastatic CRPC setting, yet many patients
present to a physician with a rising prostate-specific antigen, a castrate level of testosterone, and no radiographic
evidence of metastatic disease. What do you do in this case?
The answer is not yet clear and there are no FDA-approved
treatments. In my practice, a variety of secondary hormonal
agents, such as nilutamide, bicalutamide, low-dose diethylstilbestrol, and ketoconazole are used despite no phase III
evidence to support their use. Good clinical trials are always
an important consideration as well.
Author
Oliver Sartor
Employment or
Leadership
Positions
Consultant or
Advisory Role
Algeta; Amgen;
Bayer; Bellicum;
Bristol-Myers
Squibb; Celgene;
Dendreon;
Exelixis;
GlaxoSmithKline;
Johnson &
Johnson;
Medivation;
Oncogenex;
Pfizer; Sanofi;
Takeda
Stock
Ownership
Honoraria
Sanofi
Research
Funding
Algeta;
AstraZeneca;
Cougar
Biotechnology;
Exelixis;
GlaxoSmithKline;
Johnson &
Johnson; Sanofi
Expert
Testimony
Other
Remuneration
REFERENCES
1. Tannock IF, de Wit R, Berry WR, et al. Docetaxel plus prednisone or
mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med.
2004;351:1502-1512.
2. Petrylak DP, Tangen CM, Hussain MH, et al. Docetaxel and estramustine compared with mitoxantrone and prednisone for advanced refractory
prostate cancer. N Engl J Med. 2004;351:1513-1520.
3. Kantoff PW, Higano CS, Shore ND, et al. Sipuleucel-T immunotherapy
for castration-resistant prostate cancer. N Engl J Med. 2010;363:411-422.
4. de Bono JS, Oudard S, Ozguroglu M, et al. Prednisone plus cabazitaxel
or mitoxantrone for metastatic castration-resistant prostate cancer progressing after docetaxel treatment: A randomised open-label trial. Lancet. 2010;
376:1147-1154.
5. de Bono JS, Logothetis CJ, Molina A, et al. Abiraterone and increased
survival in metastatic prostate cancer. N Engl J Med. 2011;364:1995-2005.
6. Parker C, Heinrich D, OSullivan JM, et al. Overall survival benefit of
radium-223 chloride (Alpharadin) in the treatment of patients with symptom-
atic bone metastases in castration-resistant prostate cancer (CRPC): a phase

III randomized trial (ALSYMPCA). Eur J Cancer. 2011;47 (suppl 2).
7. Scher HI, Fizazi K, Saad F, et al. Effect of MDV3100, an androgen
receptor signaling inhibitor (ARSI), on overall survival in patients with
prostate cancer postdocetaxel: Results from the phase III AFFIRM study.
J Clin Oncol. 2012;30 (suppl 5; abstr LBA1).
8. Perlroth DJ, Thompson SF, Luna Y, et al. Time to ADT and chemotherapy initiation for treatment of metastatic prostate cancer (mPC). J Clin
Oncol. 2012;30 (suppl 5; abstr 41).
9. Attard G, Reid AH, Yap TA, et al. Phase I clinical trial of a selective
inhibitor of CYP17, abiraterone acetate, confirms that castration-resistant
prostate cancer commonly remains hormone driven. J Clin Oncol. 2008;26:
4563-4571.
10. Scher HI, Beer TM, Higano CS, et al. Antitumour activity of MDV3100
in castration-resistant prostate cancer: A phase 1-2 study. Lancet. 2010;375:
1437-1446.
291
Prognostic, Predictive, and Surrogate Factors

for Individualizing Treatment for Men with
Castration-Resistant Prostate Cancer
By Rhonda L. Bitting, MD, and Andrew J. Armstrong, MD, MSc
Overview: With the surge in therapeutic options for castrationresistant prostate cancer (CRPC) comes increasingly complicated treatment decision making, highlighting the need for
biomarkers that can identify appropriate patients for specific
ITH THE increasing number of U.S. Food and Drug

Administration (FDA) approvals of new systemic
agents for men with metastatic CRPC over the past 2 years,
the optimal treatment strategy and sequence of therapies
are evolving quickly. With novel immunotherapies, hormonal therapies, bone-targeted radioisotopes, bone microenvironmenttargeted agents, and chemotherapies emerging
in both the predocetaxel and postdocetaxel space, it becomes
imperative to develop rational combinations and sequences
of these agents, as well as to identify groups of men that are
most likely to benefit from a particular treatment. Thus,
there is an increasing need for biomarkers to help guide
clinical decision making, especially considering the number
of emerging agents currently in development and the high
cost of these therapies. Here we discuss the role, context of
use, and the limitations of existing biomarkers in CRPC. We
also examine approaches for the evaluation of novel biomarkers in clinical trials depending on the clinical or research context of use.
A biomarker is defined as a characteristic that is objectively measured and evaluated as an indicator of normal
biologic processes, pathogenic processes, or pharmacologic
responses to a therapeutic intervention.1 Biomarkers are
intended to guide the treatment of patients, and they can be
prognostic, predictive, pharmacodynamic, correlative, and/or
surrogate in nature. A prognostic biomarker reflects a patients disease outcome independent of therapy (natural
history), whereas a predictive biomarker identifies the likelihood of benefit from a specific therapy.2 A number of
prognostic biomarkers have been reported in CRPC (Table
1); however, there are no validated predictive biomarkers in
this disease. Although predictive biomarkers exist in oncology (e.g., HER2 overexpression in breast cancer predicts for
benefit with anti-HER2 agents, BRAF mutations predict for
improved outcomes with BRAF inhibitors in melanoma,
anaplastic lymphoma kinase [ALK] fusions predict for benefit to ALK tyrosine kinase inhibitors in lung adenocarcinoma), prostate cancer has lagged behind in the clinical
application of predictive biomarkers in drug development.
These predictive biomarkers offer the hope of individualized
approaches to therapy, breaking down a generally heterogeneous disease into more homogeneous molecularly defined
subsets of patients in order to maximize benefit and minimize harm.
In addition, surrogate biomarkers of overall survival (OS)
remain uncertain in CRPC, in which measures such as
progression-free survival (PFS), radiographic or pain responses, and prostate-specific antigen (PSA) declines have
generally fallen short of a generalizable surrogate. A surrogate biomarker is a biomarker that can substitute for a
292
treatments and accurately assess disease response. Here we

discuss existing and potential prognostic, predictive, and
surrogate biomarkers in CRPC.
clinically important endpoint such as OS but must meet

several statistical criteria (such as Prentices criteria) and be
validated across multiple trials examining agents with a
range of mechanisms and clinical contexts.3 Ongoing work
in CRPC to identify optimal surrogate biomarkers, particularly in the fields of circulating tumor cell (CTC) characterization and optimal classifications of radiographic
progression in CRPC, holds some promise in providing early
evidence of activity for novel therapies and thus accelerated
drug approvals and delivery of active agents into the clinic.
However, much work is to be done on this front.
Certain biomarkers are entirely used for determining
drug mechanism or metabolism (pharmacogenomic or pharmacogenetic) and will not be further discussed in this
review. These biomarkers are essential, however, in identifying pathophysiology and drug mechanism and in suggesting predictive and surrogate measures of clinical efficacy,
and they are typically explored in the initial characterization, validation, and qualification of a biomarker. As biomarkers emerge in clinical settings, these predictive or
surrogate factors can potentially guide treatment decision
making (e.g., when to start a therapy, who to offer a therapy
to, when to stop a treatment and declare failure or response),
underscoring the need for a rigorous approach to the development of predictive and surrogate biomarkers in parallel
with drug development in CRPC.
Prognostic Biomarkers in CRPC
The sections below provide advantages, limitations, and

evidence for use of common serologic and urinary biomarkers in CRPC. Imaging response, pain, and quality-of-life
changes are not addressed here, though Table 1 provides a
comprehensive list of prognostic biomarkers in CRPC.
PSA
Serum PSA often reflects the disease burden in men with

CRPC and has been included in prognostic models as an
independent risk factor for disease-related mortality.4
Changes in PSA over time during treatment may be informative regarding a patients response to therapy; however,
this response is variable depending on the mechanism of
action of the therapy. For example, treatment with the
immunotherapy sipuleucel-T improves survival without
From the Duke Cancer Institute and the Duke Prostate Center, Durham, NC.
Address reprint requests to Andrew Armstrong, MD, MSc, Duke University Medical
Center 102002, Durham, NC 27710; email: andrew.armstrong@duke.edu..
1092-9118/10/1-10
INDIVIDUALIZING TREATMENT FOR MEN WITH CRPC
changing PSA levels, whereas a 30% or greater PSA decline

within 3 months of treatment initiation with the cytotoxic
agent docetaxel correlates with a survival benefit in retrospective analyses.5-7 However, there is no percentage of PSA
decline that consistently correlates with survival across
trials of a range of drug mechanisms, and therefore PSA
decline is not a surrogate for survival in itself and should not
be a primary endpoint for systemic therapy trials in CRPC
or for regulatory approval. Prospective evaluation is needed
for PSA declines in certain contexts where the correlation
appears strongest, particularly for hormonal therapies.8
Likewise, PSA progression during therapy conveys a poor
prognosis but is not a surrogate for survival.9,10 Early but
transient rises in PSA occur in up to 20% of men during
chemotherapy but do not affect survival.11 PSA declines can
often lag or be slow to occur. Therefore, the Prostate Cancer
Working Group (PCWG2) recommends that PSA change not
be used in isolation to define disease progression or to stop
therapy, particularly during the first several months of
systemic therapy. Regular PSA evaluation is recommended
in clinical trials, however, so that it can be studied independently but in the context of other disease-related outcomes.8
Reporting of PSA changes in the form of descriptive water-
KEY POINTS
The utility of a biomarker depends on the clinical

question and its context of use.
Prostate-specific antigen (PSA), circulating tumor
cell (CTC) count, markers of bone turnover, lactate
dehydrogenase, and hemoglobin are traditional biomarkers that are correlated with overall survival in
castration-resistant prostate cancer (CRPC) but are
not predictive of benefit from specific therapies in
CRPC.
Predictive biomarkers of sensitivity to systemic therapies in CRPC are needed and are likely to be based
on the mechanism of action of a given agent (i.e.,
androgen receptor activity and MDV3100, microtubule mutations and docetaxel sensitivity, androgen
precursor levels and abiraterone sensitivity, phosphatase and tensin homolog loss or phosphoinositide
3-kinase (PI3K) pathway activation and PI3K pathway inhibitors) and may be determined from tumor
tissue or blood-/plasma-/urine-based assays such as
CTCs or cell-free tumor DNA.
Improvements in PSA, CTC count, pain, radiographic
tumor burden, and/or quality of life and delays in
progression are associated with improved prognosis
post-treatment, and work is ongoing to determine the
association of these improvements with overall survival in patients with CRPC in a range of contexts.
Although there are no current validated surrogates
for overall survival in CRPC, it is imperative that
we incorporate intermediate endpoints and biomarkers into clinical trial design to develop potential
surrogates of activity in order to accelerate drug
development.
Table 1. Prognostic Factors in CRPC a

Baseline Prognostic Factors
Performance status
Visceral metastatic disease
Anemia
Alkaline phosphatase and
other bone turnover
markers (e.g., TRAP)
PSA
PSA kinetics
Type of progression (bone,
measurable disease, PSA
only)
CTC count
LDH
Albumin
Pain
Number of disease sites
Post-Treatment Prognostic Factors
PSA decline
Pain improvement
Quality-of-life improvement
Change in CTC count to 5
Predictive Factors
None validated
See Table 3
PSA PFS
Radiographic response PFS
Induction of immunity to tumor
antigens (sipuleucel-T)
Skeletal-related event
development
Development of anemia
LDH changes
Type of progression
Alkaline phosphatase
improvements
Age
VEGF levels
Interleukin-6 levels
Chromogranin-A
C-reactive protein
Bone turnover markers
Gleason sum in primary
Urinary N-telopeptide
Abbreviations: CRPC, castration-resistant prostate cancer; CTC, circulating
tumor cell; LDH, lactate dehydrogenase; PFS, progression-free survival; PSA,
prostate-specific antigen; TRAP, tartrate-resistant acid phosphatase; VEGF,
vascular endothelial growth factor.
a
Adapted from Armstrong and colleagues. 35
fall plots provides a visual clue to the effect of a systemic

agent on PSA. However, many agents in prostate cancer
have demonstrated benefit without notable early PSA
changes, particularly immunotherapies.5
PSA doubling time (PSADT) is prognostic for OS in
CRPC.4 In nonmetastatic CRPC, both the absolute PSA and
the PSADT can identify men at high risk for early disease
progression. In metastatic but asymptomatic CRPC, rapid
PSA kinetics may suggest the need for more aggressive
therapy, whereas reduction in PSADT with chemotherapy is
suggestive of a better prognosis.4,6 However, PSADT may
also change over time without intervention, and changes in
PSA kinetics have not been demonstrated to be surrogates of
OS broadly, so these findings must be interpreted cautiously.12
In neuroendocrine or small cell prostate cancer, very little
or no PSA is produced, and therefore PSA changes do not
correlate with disease status. Chromogranin A levels are
prognostic but not predictive in these cases, and thus other
biomarkers such as CTCs will be needed in this emerging
and virulent subset of CRPC.13
CTCs
To establish distant metastases, invasive cancer cells

likely circulate in the bloodstream and home in on their
target of choice, which in CRPC is often bone. CellSearch
(Veridex LLC of Raritan, NJ) is the only FDA-cleared
technology for the detection of CTCs in patients with cancer.
Using this technology, CTCs are defined as nucleated cells at
least 4 microns in diameter, immunomagnetically captured
from the bloodstream using antibodies against epithelial cell
293
BITTING AND ARMSTRONG
adhesion molecule and further characterized by cytokeratin

expression and lack of the leukocyte marker CD45.14 The
CTC count is a prognostic biomarker in CRPC and in other
solid tumors. Before the start of chemotherapy for CRPC,
the detection of five or more CTCs is associated with an
inferior OS compared to patients with less than five. Furthermore, a drop in CTCs to below five during treatment is
associated with an improvement in survival.15 Whether
CTCs have a greater association with survival than PSA or
radiographic changes over time is an important and unresolved question.16
Flares in the CTC count have not been reported, and
changes in the CTC count often precede changes in PSA.17
Therefore, the CTC trend may be valuable for therapeutic
decision making in cases where other clinical assessments of
response are equivocal, though its use in this setting is
purely speculative. Prospective evaluation of whether the
CTC count may act as a surrogate for OS is ongoing, but the
initial study assessing its use as a surrogate in CRPC was
promising.18
The potential use of CTCs as a biomarker is not restricted
to the current CTC detection method and definition. One
limitation of the CellSearch epithelial-based method is the
lack of detection of CTCs in many men with progressive,
metastatic CRPC.17 Recent findings suggest that there is
CTC phenotypic heterogeneity, with some CTCs expressing
not only epithelial proteins, but also mesenchymal and
stemness proteins.19 Therefore, epithelial-mesenchymal
transitions may explain the relative underdetection of CTCs
in patients with advanced malignancy using the standard
epithelial antigen-based technology. A number of technologies that employ nonepithelial targets for CTC capture and
characterization are under development, though the prognostic and predictive implications of these CTC phenotypes
must be independently and prospectively validated.
Because CTCs may be a direct measurement of the underlying tumor biology, enhanced capture of CTCs may aid in
the development of CTCs as a predictive biomarker. For
example, the identification of phosphatase and tensin homolog (PTEN) loss and androgen receptor amplification in
CTCs suggests that personalization of therapy using biomarker guidance is achievable.21,22 As additional CTC phenotypes are discovered, potential therapeutic targets may
emerge. Thus, CTCs may provide prognostic information
through enumeration, but more importantly, they may provide a noninvasive window into tumor biology and provide
predictive information for therapeutic decision making.
Lactate Dehydrogenase
The enzyme lactate dehydrogenase (LDH) converts pyruvate to lactate and vice versa as part of the normal glycolysis
and gluconeogenesis pathways of the cell, and those pathways are preferentially upregulated in cancer cells as a
result of oncogenic signaling.22 LDH is an independent
prognostic biomarker in CRPC and other tumor types, and
elevations are thought to reflect the underlying tumor burden.23 Assessments of LDH in conjunction with other biomarkers such as PSA or CTCs may improve on the current
clinical utility of LDH for risk stratification and prognostication.18 Also, although baseline LDH is strongly prognostic
in multivariate models in CRPC, increases in LDH following
therapy carry an unfavorable prognosis and may be useful in
interpreting treatment response.24 Given the strong associ-
294
ation of LDH with OS in CRPC over time, however, serial

measurement and reporting of this factor during treatment
and in the context of clinical trials are useful and recommended.
Markers of Bone Turnover
Prostate cancer commonly metastasizes to the bone, and

this may be mediated by adhesion molecules that target the
bone microenvironment and promote osteomimicry.25 As
such, agents that impede this tumor-bone stromal interaction such as zoledronic acid and denosumab delay the
development of skeletal-related events in CRPC.26,27 Bony
metastatic effects can be indirectly measured using bone
turnover markers such as N-telopeptide, tartrate-resistant
acid phosphatase 5b, C-telopeptide, and osteopontin, and
these are being evaluated for use as prognostic and predictive biomarkers.28 Bone-derived alkaline phosphatase,
which is a measure of osteoblastic activity, is an established
prognostic biomarker in CRPC.29 The reduction in total
alkaline phosphatase with docetaxel is independently prognostic in CRPC and may provide evidence of a survival
benefit even in the absence of decline in PSA or improvement in imaging.30 Overall, the measurement of total or
bone-derived alkaline phosphatase at baseline and over time
provides prognostic information but does not predict response to bone-targeted therapy.
Urinary N-telopeptide, which is a breakdown product of
type 1 collagen, is elevated in men with CRPC and bony
metastases, and high N-telopeptide levels are associated
with an increased risk of skeletal-related events, disease
progression, and death.31 Treatment with the receptor activator of nuclear factor kappa B (RANK)-ligand antagonist
denosumab reduces the levels of N-telopeptide in the urine
beyond the reduction seen with zoledronic acid, though
whether this reduction correlates with denosumabs superiority in preventing skeletal-related events is unclear.27 The
utility of N-telopeptide and other bone turnover markers as
predictive biomarkers for the use of bone-targeted agents is
an area of active investigation.
Hemoglobin
Anemia in the setting of CRPC is often multifactorial,

related to androgen deprivation therapy, renal disease,
chemotherapy toxicity, bone marrow infiltration or failure,
or other comorbid conditions. Anemia has long been recognized as a prognostic factor in CRPC and, in multivariate
analysis, is among the strongest predictors for docetaxelrelated PSA declines, tumor response, and OS for men with
CRPC.32 Thus, the measurement of hemoglobin over time
has both practical management and prognostic implications
in the treatment of men with CRPC.
Post-treatment Prognostic and Potential Surrogate
Endpoints in CRPC
Although imaging measurements are performed routinely

to assess disease status in oncology, the absolute size of a
metastatic lesion does not always correlate with clinical
outcome. The use of the Response Evaluation Criteria in
Solid Tumors (RECIST) in CRPC is especially problematic,
as RECIST does not account for skeletal lesions, the most
common site of metastasis and source of morbidity in
CRPC.33 Bone scintigraphy is used to assess skeletal lesions;

Table 2. Post-treatment Prognostic and Potential Surrogate Endpoints in CRPC a
Biomarker
PSA decline
CTCs
Bone turnover markers
Quality-of-life or pain improvement
Radiographic changes (RECIST 1.1, PCWG2)
PFS (PCWG2)
References
Pros
Cons
6, 7
Easily measurable
Widely available
Time 3 mo
Evidence to support use with cytotoxic therapy
Not validated for use with novel agents

PSA flares may occur
Threshold of response unclear
Not useful with immunotherapy
Subgroups of CRPC do not make PSA
Not present in all CRPC using CellSearch
Expensive
Requires specialized lab
Requires processing within 72 h
New CTC definitions and technologies require validation
Normal in patients without bony metastases
May be normal if bony metastases
Clinical relevance of partial changes unclear
Qualitative, requires validated scales
Subjective and variable
May be multifactorial
Inherent biases
Not useful if nonmetastatic disease
Not always measurable (i.e., bone lesions)
Only modestly prognostic
Bone scan flare may occur
Some agents (i.e., immunotherapy) may have benefit
without affecting imaging
Exact definition is critical
Not validated as a surrogate for OS
Censorship prevents current surrogate analyses
15, 18, 19
30
6, 36
8, 16, 34
9, 10
Change before PSA

Tumor-specific
Strongly prognostic
Surrogacy use under evaluation
May reflect current disease phenotype
Reflects tumor-stromal interaction
Widely available
Prognostic
Direct patient measure
Well-defined criteria for response
May capture clinical benefit

Can measure effect of cytostatic agents
Abbreviations: CRPC, castration-resistant prostate cancer; CTC, circulating tumor cell; h, hours; mo, months; OS, overall survival; PFS, progression-free survival;
PSA, prostate-specific antigen.
a
however, it is an indirect measurement that detects not only

the cancer but also unassociated inflammation and degeneration. In fact, effective therapies sometimes cause what
appear to be enlarging bony lesions but actually represent
healing bone surrounding cancerous bone, known as the
flare phenomenon or osteoblastic reactions.34 To account for
this, bone scan progression per PCWG2 criteria requires the
confirmation of bony lesions on a subsequent scan.8 Overall,
however, there is only a modest correlation between bone
scan changes and survival in CRPC to date.9
The time from the initiation of therapy or study entry to
the documentation of disease progression is known as
progression-free survival. Because PFS is generally measurable sooner than OS, has more independent events than OS,
and is not confounded by subsequent therapeutic interventions, PFS is a convenient endpoint for use in clinical
trials.17 The major limitation of the PFS endpoint is that
disease progression can be suggested by PSA alone, radiographic changes based on lesion size or the presence of new
lesions, and/or symptomatology, and the clinical significance
of each varies. The PCWG2 criteria for disease progression
provides a consensus on the definition, but the relevance of
this outcome depends on the mechanism of action of the
therapy.8 Immunotherapy, for example, prolongs OS without affecting PFS, therefore, the use of PFS as a surrogate
for survival must be considered in the context of the therapy
in question and drug mechanism.5 Given the poor correlation of PFS and OS across phase III trials in CRPC, we are
faced with a conundrum of how to approve drugs based only
on PFS benefit without knowing the full effect of a novel
agent on OS.4 Table 2 provides a list of potential surrogate
biomarkers in CRPC as well as a brief summary of advantages and disadvantages for each.
Predictive Biomarkers in CRPC
As described above, there are a number of prognostic

biomarkers in CRPC. However, there is a paucity of biomarkers under consideration for predictive use, and only
CTCs are being evaluated in clinical trials for surrogacy use.
In an era where costs of medical therapies are skyrocketing
Table 3. Predictive Biomarkers under Consideration a
Biomarker
CTC count
High urinary N-telopeptide

or bone turnover markers
Androgen receptor splice
variants
Levels of androgen
precursors
c-Met/HGF activity
PTEN loss in CTCs or
metastases
Ras/raf mutations
Tubulin mutations
DNA repair defects
Myc amplification
Potential CRPC Application
Potential surrogate for OS with abiraterone,

prognostic predocetaxel, may be mechanismindependent
Benefit from denosumab or zoledronic acid
Predict sensitivity or resistance to anti-androgens
(e.g., MDV3100) or to abiraterone
Predict benefit from androgen synthesis inhibitors
(e.g., abiraterone acetate or TAK700)
Enrich for benefit from c-met inhibitors (e.g.,
cabozantinib)
Enrich for benefit from PI3 kinase pathway
inhibitors
Potential benefit from ras pathway inhibitors (e.g.,
sorafenib, vemurafenib)
May predict resistance to microtubule-based
therapies (e.g., docetaxel, cabazitaxel)
Enrich for benefit from PARP inhibitors
Predict sensitivity to cell-cycle inhibitors
(antiproliferation agents)
Abbreviations: CRPC, castration-resistant prostate cancer; CTC, circulating

tumor cell; HGF, hepatocyte growth factor; OS, overall survival; PARP, poly(adenosine diphosphate [ADP] ribose) polymerase; PTEN, phosphatase and tensin
homolog.
a
295
BITTING AND ARMSTRONG
and treatment options are expanding, the use of predictive

biomarkers may allow providers to select or enrich for
patients with CRPC who are most likely to benefit from a
specific therapy. Similar to the benefit seen with HER2targeted therapy in HER2-amplified breast cancer, one
would hypothesize that patients with CRPC with androgen
receptor amplification are most likely to benefit from an
androgen receptor inhibitor like MDV3100. Likewise, those
with PTEN deletions in their tumor would be expected to
benefit from treatment with a PI3 kinase inhibitor and those
with c-Met overexpression would be expected to benefit from
c-Met inhibition. As prostate cancer subtypes are increasingly identified, this biomarker-driven paradigm will allow
for the customizing of therapeutic interventions. With the
advancing technology for CTC detection and characterization, this molecular information may eventually be obtainable from a blood test. Because a biomarker must be
evaluated in a series of clinical trials to generate evidence
for its use, it is imperative that investigators employ trial

designs that allow for appropriate assessment of biomarkers. Table 3 lists predictive biomarkers under consideration
and the evidence supporting their use in CRPC.
Conclusion
All biomarkers currently used clinically in CRPC have

prognostic implications when measured before starting therapy but have not yet been established as predictive or
surrogate markers in the on- or post-treatment settings.
Ongoing randomized studies of active systemic therapies
with prospectively embedded biomarker-based validation
studies are needed before these can be used for definitive
clinical decision making. It is imperative that biomarkers be
studied rigorously in parallel with drug development, given
the potential to maximize benefit and minimize harms and
costs to individual men with CRPC and to society.
Author
Rhonda L. Bitting
Andrew J. Armstrong
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership

Novartis; Sanofi
Honoraria
Amgen;
Dendreon;
Johnson &
Johnson; Pfizer;
Sanofi
Research
Funding
Veridex, LLC
Active Biotech;
Bristol-Myers
Squibb;
Dendreon;
ImClone
Systems;
Johnson &
Johnson;
Medivation;
Novartis; Pfizer;
Sanofi
Expert
Testimony
Other
Remuneration
REFERENCES
1. Biomarkers Definitions Working Group. Biomarkers and surrogate
endpoints: preferred definitions and conceptual framework. Clin Pharmacol
Ther. 2001;69:89-95.
2. Dancey JE, Dobbin KK, Groshen S, et al. Guidelines for the development
and incorporation of biomarker studies in early clinical trials of novel agents.
Clin Cancer Res. 2010;16:1745-1755.
3. Prentice RL. Surrogate and mediating endpoints: current status and
future directions. J Natl Cancer Inst. 2009;101:216-217.
4. Armstrong AJ, Garrett-Mayer ES, Yang YC, et al. A contemporary
prognostic nomogram for men with hormone-refractory metastatic prostate
cancer: a TAX327 study analysis. Clin Cancer Res. 2007;13:6396-6403.
6. Armstrong AJ, Garrett-Mayer E, Ou Yang YC, et al. Prostate-specific
antigen and pain surrogacy analysis in metastatic hormone-refractory prostate cancer. J Clin Oncol. 2007;25:3965-3970.
7. Petrylak DP, Ankerst DP, Jiang CS, et al. Evaluation of prostate-specific
antigen declines for surrogacy in patients treated on SWOG 99-16. J Natl
Cancer Inst. 2006;98:516-521.
8. Scher HI, Halabi S, Tannock I, et al. Design and end points of clinical
trials for patients with progressive prostate cancer and castrate levels of
testosterone: recommendations of the Prostate Cancer Clinical Trials Working Group. J Clin Oncol. 2008;26:1148-1159.
9. Halabi S, Vogelzang, NJ, Ou SS, et al. Progression-free survival as a
predictor of overall survival in men with castrate-resistant prostate cancer.
J Clin Oncol. 2009;27:2766-2771.
10. Hussain M, Goldman B, Tangen C, et al. Prostate-specific antigen
progression predicts overall survival in patients with metastatic prostate
cancer: data from Southwest Oncology Group Trials 9346 (Intergroup Study
0162) and 9916. J Clin Oncol. 2009;27:2450-2456.
11. Thuret R, Massard C, Gross-Goupil M, et al. The postchemotherapy
PSA surge syndrome. Ann Oncol. 2008;19:1308-1311.
12. Smith MR, Manola J, Kaufman DS, et al. Rosiglitazone versus placebo
296
for men with prostate carcinoma and a rising serum prostate-specific antigen
level after radical prostatectomy and/or radiation therapy. Cancer. 2004;101:
1569-1574.
13. Taplin ME, George DJ, Halabi S, et al. Prognostic significance of
plasma chromogranin a levels in patients with hormone-refractory prostate
cancer treated in Cancer and Leukemia Group B 9480 study. Urology.
2005;66:386-391.
14. Allard WJ, Matera J, Miller MC, et al. Tumor cells circulate in the
peripheral blood of all major carcinomas but not in healthy subjects or
patients with nonmalignant diseases. Clin Cancer Res. 2004;10:6897-6904.
15. de Bono JS, Scher HI, Montgomery RB, et al. Circulating tumor cells
predict survival benefit from treatment in metastatic castration-resistant
prostate cancer. Clin Cancer Res. 2008;14:6302-6309.
16. Sonpavde G, Pond GR, Berry WR, et al. The association between
radiographic response and overall survival in men with metastatic castrationresistant prostate cancer receiving chemotherapy. Cancer. 2011;117:39633971.
17. Scher HI, Morris MJ, Basch E, et al. End points and outcomes in
castration-resistant prostate cancer: from clinical trials to clinical practice.
J Clin Oncol. 2011;29:3695-3704.
18. Scher HI, Heller G, Molina A, et al. Evaluation of circulating tumor cell
(CTC) enumeration as an efficacy response biomarker of overall survival (OS)
in metastatic castration-resistant prostate cancer (mCRPC): planned final
analysis (FA) of COU-AA-301, a randomized double-blind, placebo-controlled
phase III study of abiraterone acetate (AA) plus low-dose prednisone (P) post
docetaxel. J Clin Oncol. 2011;29:293s (suppl; abstr LBA4517).
19. Armstrong AJ, Marengo MS, Oltean S, et al. Circulating tumor cells
from patients with advanced prostate and breast cancer display both epithelial and mesenchymal markers. Mol Cancer Res. 2011;9:997-1007.
20. Attard G, Swennenhuis JF, Olmos D, et al. Characterization of ERG,
AR and PTEN gene status in circulating tumor cells from patients with
castration-resistant prostate cancer. Cancer Res. 2009;69:2912-2918.
21. Shaffer DR, Leversha MA, Danila DC, et al. Circulating tumor cell

analysis in patients with progressive castration-resistant prostate cancer.
Clin Cancer Res. 2007;13:2023-2029.
22. Vander Heiden MG, Cantley LC, and Thompson CB. Understanding
the Warburg effect: the metabolic requirements of cell proliferation. Science.
2009;324:1029-1033.
23. Halabi S, Small EJ, Kantoff PW, et al. Prognostic model for predicting
survival in men with hormone-refractory metastatic prostate cancer. J Clin
Oncol. 2003;21:1232-1237.
24. Halabi S, Ou S, Vogelzang NJ, et al. The prognostic value of change in
hemoglobin (HGB), LDH and PSA levels at 3 months from baseline in men
with castrate recurrent prostate cancer. In ASCO Prostate Cancer Symposium Proceedings. Alexandria, VA: ASCO, 2007 (abstr 217).
25. Chung LW, Huang WC, Sung SY, et al. Stromal-epithelial interaction
in prostate cancer progression. Clin Genitourin Cancer. 2006;5:162-170.
26. Saad F, Gleason DM, Murray R, et al. Long-term efficacy of zoledronic
acid for the prevention of skeletal complications in patients with metastatic
hormone-refractory prostate cancer. J Natl Cancer Inst. 2004;96:879-882.
27. Fizazi K, Carducci M, Smith M, et al. Denosumab versus zoledronic
acid for treatment of bone metastases in men with castration-resistant
prostate cancer: a randomised, double-blind study. Lancet. 2011;377:813-822.
28. Coleman RE, Major P, Lipton A, et al. Predictive value of bone
resorption and formation markers in cancer patients with bone metastases
receiving the bisphosphonate zoledronic acid. J Clin Oncol. 2005;23:49254935.
29. Cook RJ, Coleman R, Brown J, et al. Markers of bone metabolism and
survival in men with hormone-refractory metastatic prostate cancer. Clin

Cancer Res. 2006;12:3361-3367.
30. Sonpavde G, Pond GR, Berry WR, et al. Serum alkaline phosphatase
changes predict survival independent of PSA changes in men with castrationresistant prostate cancer and bone metastasis receiving chemotherapy. Urol
Oncol. Epub 2010 Sep 29.
31. Brown JE, Cook RJ, Major P, et al. Bone turnover markers as
predictors of skeletal complications in prostate cancer, lung cancer, and other
solid tumors. J Natl Cancer Inst. 2005;97:59-69.
32. Armstrong AJ, Tannock IF, de Wit R, et al. The development of risk
groups in men with metastatic castration-resistant prostate cancer based on
risk factors for PSA decline and survival. Eur J Cancer. 2010;46:517-525.
33. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation
criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J
Cancer. 2009;45:228-247.
34. Ryan CJ, Shah S, Efstathiou E, et al. Phase II study of abiraterone
acetate in chemotherapy-naive metastatic castration-resistant prostate cancer displaying bone flare discordant with serologic response. Clin Cancer Res.
2011;17:4854-4861.
35. Armstrong AJ, Eisenberger MA, Halabi S, et al. Biomarkers in the
management and treatment of men with metastatic castration-resistant
prostate cancer. Eur Urol. 2012;61:549-559.
36. Berthold DR, Pond GR, Roessner M, et al. Treatment of hormonerefractory prostate cancer with docetaxel or mitoxantrone: relationships
between prostate-specific antigen, pain, and quality of life response and
survival in the TAX-327 study. Clin Cancer Res. 2008;14:2763-2767.
297
KIDNEY CANCER BIOLOGY AND THERAPEUTICS:

VEGFR, MTOR, AND BEYOND
CHAIR
Toni K. Choueiri, MD, MSc
Boston, MA
SPEAKERS
William G. Kaelin, MD
Boston, MA
Daniel Y. C. Heng, MD, MPH
Tom Baker Cancer Centre, University of Calgary
Calgary, AB, Canada
The Evolving Landscape of Metastatic Renal

Cell Carcinoma
By Daniel Y. C. Heng, MD, MPH, and Toni K. Choueiri, MD, MSc
Overview: The treatment paradigm in metastatic renal cell

carcinoma (mRCC) has evolved over the last 5 years. There are
now seven approved targeted therapies against the vascular
endothelial growth factor (VEGF) and mammalian target of
rapamycin (mTOR) pathways. The use of targeted therapy,
HE ELUCIDATION of the von Hippel Lindau (VHL)

tumor suppression gene associated with the hereditary
and sporadic forms of clear cell renal cell carcinomas (RCCs)
has sparked a revolution of targeted therapy for this disease.
The loss of VHL leads to downstream accumulation of
hypoxia inducible factor (HIF) and subsequent activation of
tumor promoting pathways including VEGF.1 In fact, malignancies of the kidney show the greatest range and maximum expression of VEGF,2 suggesting a rational target in
this disease. mTOR is a serine threonine kinase that is
another important central target in RCC, as it enhances
translation of proteins involved in cellular growth, survival,
and angiogenesis.
Inhibitors of VEGF and mTOR have dominated the scene
in the treatment of mRCC. These drugs have become commonplace in the treatment algorithm (Table 1) based on the
registration phase III clinical trials.
Choices of Targeted Therapy
The list of choices for first-line targeted therapy is ever

increasing. Oral VEGF tyrosine kinase inhibitors such as
sunitinib3 and pazopanib4 and intravenous antibodies
against VEGF such as bevacizumab5,6 in combination with
interferon-alpha
have
demonstrated
a
convincing
progression-free survival (PFS) benefit in the first-line setting in patients with favorable or intermediate prognosis.
Patients with poor prognosis benefit from temsirolimus
(mTOR inhibitor),7 as it is the only therapy to improve
overall survival (OS) in the setting of a phase III trial in
patients with poor prognosis.
After failure of initial therapy, there are also choices to be
made. On progression after immunotherapy, sorafenib has
demonstrated a PFS benefit compared with placebo.8 After
failure of initial VEGF therapy, both everolimus (mTOR
inhibitor) and axitinib (VEGF inhibitor) have also demonstrated a PFS benefit. Everolimus was examined in patients
who progressed taking initial sunitinib, sorafenib, or both
drugs and were randomly selected to take everolimus compared with placebo. The PFS for the everolimus versus
placebo group was 4.9 versus 1.9 months (p 0.0001),
respectively.9 Axitinib was recently approved by the FDA
based on the registration phase III trial of axitinib compared
with sorafenib in patients previously treated with a broad
range of frontline therapies, including mostly prior sunitinib
and cytokines. The response rates and PFS for the axitinib
versus sorafenib group was 19% versus 9% and 6.7 versus
4.7 months (p 0.0001).10
Although there is no level I evidence for the drug of choice
for patients in whom initial mTOR inhibitors failed, the use
of a sequential targeted therapy not previously used remains a standard practice for these patients. This is also
sequences, combinations, and investigational compounds will

be discussed. Prognostic and predictive tools are detailed,
although much work must be done to find predictive biomarkers in an effort to individualize therapy for patients.
true for the choice of third-line targeted therapy, as there is

no level I evidence to guide us in this setting.
We are now faced with several targeted therapies to
choose from in each treatment setting. As we do not have
any robust, externally validated predictors of response to
targeted therapy, the choice of targeted therapy is usually
based on physician preference, intravenous versus oral therapy, reimbursement issues, and toxicity profile. For example, patients with significant lung disease receiving oxygen
or poorly controlled diabetes mellitus may not be optimal
candidates for an mTOR inhibitor, as there is a risk of
noninfectious pneumonitis and hyperglycemia with this
class of agents. Similarly, patients with refractory hypertension treated with several antihypertensive agents or severe
cardiovascular (CV) morbidities may not initially choose a
VEGF inhibitor, as these are known to increase the risk of
hypertension and CV events. These examples may be encountered in daily practice. However, no routine markers to
predict response or toxicity to allow for a more informed
decision about which targeted therapy to choose are available.
Sequences of VEGF and mTOR inhibitors are being investigated in several randomized trials including the RECORD
3 (NCT00903175), 404 study (NCT00474786), and START
(NCT01217931) clinical trials (Table 2). These may help
shed light onto the best sequence of drugs to use although
they do not add to the personalized medicine approach.
Different combinations of targeted therapy have also been
studied, including the TORAVA trial of temsirolimus and
bevacizumab in the frontline setting, which demonstrated
only increased toxicity and no convincing evidence of added
benefit.11 Other combinations such as sunitinib and bevacizumab have proven to be toxic and have adverse effects such
as thrombotic thrombocytopenic purpura, which is rarely
seen when the agents are used alone.11 A phase III CALGB
trial investigating second-line use of everolimus compared
with everolimus plus bevacizumab is underway with OS as
the primary endpoint (NCT01198158). Unless substantial
differences in durable complete responses or OS are documented, combination therapy remains investigational.
From the Tom Baker Cancer Center, University of Calgary, Calgary, Alberta, Canada;
Dana Farber Cancer Institute and Harvard Medical School, Boston, MA.
Address reprint requests to Daniel Y. C. Heng, MD, MPH, Tom Baker Cancer Center,
University of Calgary, Calgary, Alberta, Canada, 1331 29th St. NW, Calgary, Alberta,
Canada, T2N 4N2; email: daniel.heng@albertahealthservices.ca.
1092-9118/10/1-10
299
HENG AND CHOUEIRI

Prognostic Factors in Advanced RCC
Prognostic factors have been developed to assist in patient

counseling, risk-directed treatment, and clinical trial design. These clinical factors that predict prognosis are a
combination of patient factors, indicators of tumor burden,
pro-inflammatory markers, and treatment-related factors
including prior cytoreductive nephrectomy.12
Many multivariable models such as the Memorial SloanKettering Cancer Center (MSKCC) model13 have been created in an effort to stratify groups of patients with different
biology. More recently, the International mRCC Database
Consortium developed criteria from a large populationbased database of patients treated with VEGF targeted
therapy. The independent predictors of poor OS include
anemia, hypercalcemia, thrombocytosis, neutrophilia, a
Karnofsky Performance Status of less than 80%, and a
diagnosis-to-treatment interval of less than 1 year. Patients
are then segregated into three risk categories: favorable risk
(no risk factors, median OS not reached), intermediate risk
(one to two risk factors, median OS of 27 months), and poor
risk (three or more risk factors, median OS of 8.8 months).14
This multivariable model has been externally validated in
an even more modern set of patients treated with VEGF
targeted therapy with median OSs at 44 months, 21 months,
and 8 months for the favorable, intermediate, and poor risk
groups, respectively (p 0.0001). This same model has also
been validated in a set of patients who were previously
treated with VEGF inhibitors and received a second line of
systemic therapy.15 Data from these models suggest continued improvements in OS across different risk groups, which
is a testament to the effectiveness of modern-day targeted
therapy.
Prediction of Response to Targeted Therapy: Steps
Toward Personalized Medicine
Currently there are no clinical factors or biomarkers that

can conclusively predict which targeted therapies patients
will respond to. There are some serum levels of proteins in
KEY POINTS
300
There are now seven targeted therapies that are

approved by the U.S. Food and Drug Administration
for the treatment of metastatic renal cell carcinoma
and are included in the treatment algorithm. The
choice of drug is dependent on the context of the
corresponding clinical trial, physician experience,
drug availability, and patient preference.
Currently there are no standard, routinely used baseline biomarkers to predict response and toxicities.
Recent studies have found that the development of
hypertension as well as other therapy-related toxicities and single nucleotide polymorphisms may indicate improved patient outcomes, but using these
potential biomarkers still requires prospective validation.
There are new agents on the horizon, including
more targeted therapies and next-generation
immunotherapy.
Table 1. A Proposed Treatment Algorithm for Patients

with mRCC
Therapies with
Level 1 Evidence
Setting
Patients
First-Line Therapy
Good or intermediate
risk
Sunitinib
Pazopanib
Bevacizumab
IFN
Poor risk
Temsirolimus
Prior cytokines
Sorafenib
Axitinib
Pazopanib
Axitinib
Everolimus
Second-Line
Therapy
Prior VEGF
Third-Line
Therapy
Prior mTOR
No data
available
Any
No data
available
Other Options
High dose IL-2 in

highly select patients
Sorafenib
Clinical trial
Observation in select
patients
Other VEGF inhibitors
Clinical trial
Sunitinib
Clinical trial
Targeted therapy not
previously used
Clinical trial
previously used
Clinical trial
previously used
Clinical trial
Abbreviations: IFN, interferon; IL, interleukin; mRCC, metastatic renal cell

carcinoma; mTOR, mammalian target of rapamycin; VEGF, vascular endothelial
growth factor.
the angiogenesis pathway that may be prognostic of OS, but

predictive markers of response remain elusive.12 Single
nucleotide polymorphisms (SNPs) are single-base pair
changes within a gene that may or may not affect gene
function, and many have been explored for their prognostic
or predictive value.
For patients treated with pazopanib, SNPs in two
interleukin-8 and HIF1A loci were associated with a significant difference in PFS whereas SNPs in HIF1A, NR1/2, and
three VEGFA loci were associated with overall response
rates.16 For patients treated with sunitinib, 136 patients
with clear cell mRCC were examined to determine a favorable genetic profile, which was found to include an A allele in
the CYP3A5 6986A/G loci, an absent CAT copy in the NR1/3
haplotype, and a TCG copy in the ABCB1 haplotype. Patients with this favorable profile had an improved PFS and
OS compared to those without.17 A VEGF SNP in a different
loci was found to be associated with sunitinib-induced hypertension and another VEGF SNP and VEGF Receptor-2
SNP were found to be together associated with OS.18 Another study found two VEGF Receptor-3 SNPs to be associated with PFS when treated with sunitinib.19 These results
are interesting but are currently restricted to the caucasian
population, as there are substantial racial differences in
SNPs. These SNPs require further prospective evaluation
while ensuring correction for multiple testing to see whether
incorporating them in the decision-making process of choosing targeted therapy for an individual patient improves
outcome.
Recently, studies of toxicities due to targeted therapy have
demonstrated better treatment outcomes when toxicity is
encountered. For example, the development of hypertension
during the first cycle of sunitinib treatment was associated
with a better overall response rate, PFS, and OS.20 Similar
findings, including fatigue and hand-foot syndrome being
METASTATIC RENAL CELL CARCINOMA

Table 2. Selected Ongoing mRCC Clinical Trials Investigating Sequencing, Combinations, and New Drugs
Trials
Sequencing
Combinations
New Drugs
RECORD 3 Randomized Phase II

NCT00903175
START Trial Phase III NCT01217931
Population
Randomization
Treatment naive mRCC

Treatment naive mRCC with
nephrectomy
404 Study Phase III NCT00474786
mRCC refractory to sunitinib
BEST Trial Phase III NCT 00378703
Treatment naive clear cell

predominant mRCC
CALGB 90,802 Phase III

NCT01198158
mRCC with one prior VEGF

inhibitor
ADAPT Trial AGS-003 (Autologous

Dendritic Cell Immunotherapy)
Phase III
BMS936558 (PD-1 inhibitor) Phase
II (phase III trial upcoming)
NCT01354431
Treatment naive mRCC
Dovitinib (TKI258) (FGF and VEGF

inhibitor) Phase III NCT01030783
Tivozanib (AV951) (potent and
specific VEGF inhibitor) Phase III
NCT01030783
mRCC with at least one prior

therapy
mRCC with clear cell component

with one prior VEGF inhibitor
and one prior mTOR inhibitor
mRCC with clear cell component
and previous nephrectomy
Sunitinib 3 everolimus vs.

everolimus 3 sunitinib
Pazopanib 3 bevacizumab vs.
pazopanib 3 everolimus vs.
everolimus 3 bevacizumab
vs. everolimus 3 pazopanib
vs. bevacizumab 3
pazopanib vs. bevacizumab
3 everolimus
Temsirolimus vs. sorafenib
Bevacizumab vs. temsirolimus vs.
bevacizumab sorafenib vs.
temsirolimus sorafenib
Everolimus bevaciuzmab vs.
everolimus
Primary Endpoint
PFS of first-line treatment

(finished accrual)
Time to treatment failure of a
sequence
PFS of second-line treatment

(finished accrual)
PFS of first-line treatment
OS
AGS-003 with sunitinib/standard

therapy vs. sunitinib/standard
therapy
BMS936558 0.3 mg/kg q 3
weekly vs. BMS936558 2
mg/kg q 3 weekly vs.
BMS936558 10 mg/kg
q 3 weekly
Dovitinib vs. sorafenib
PFS of third-line treatment
Tivozanib vs. sorafenib
PFS (data to be presented)
OS
PFS
Abbreviations: FGF, fibroblast growth factor; mRCC, metastatic renal cell carcinoma; mTOR, mammalian target of rapamycin; OS, overall survival; PD-1, programmed
death 1; PFS, progression-free survival; VEGF, vascular endothelial growth factor.
associated with better outcomes, have been shown. These

are important associations; however, clinicians should
not discontinue targeted therapy if toxicities do not develop
because the discriminatory value and accuracy in determining hypertension or other toxicities is unknown.
Additionally, these biomarkers are only helpful once the
administration of the drug has already started and
thus do not help the clinician choose which drug to use
initially.
Exploring New Agents and Mechanisms of Action
Newer agents are on the horizon to expand the treatment

armamentarium in advanced RCC. They are not yet FDA
approved at this time. A more potent and specific VEGF
tyrosine kinase inhibitor Tivozanib (AV-951) is expected to
report PFS endpoints against sorafenib in the treatment
naive setting (Table 2). Dovitinib (TKI258) is a fibroblast
growth factor (FGF) inhibitor as well as a VEGF inhibitor
and is currently being studied as part of the GOLD trial
against sorafenib in third-line therapy after failure of one
VEGF and one mTOR inhibitor. The FGF pathway is hypothesized to be an angiogenic escape mechanism that is
upregulated when the tumor develops resistance against our
current VEGF and mTOR inhibitors. Dovitinib will test the
hypothesis that targeting the FGF angiogenic escape mechanism will lead to further responses and prolongation of
survival.
Next-generation immunotherapy is being investigated in
mRCC. Programmed death-1 (PD-1) is a member of the

immunoglobulin gene family and is expressed after T-cell
activation to inhibit T-cell receptor signaling as a way to
regulate the T-cell response. Higher preoperative soluble
PD-1 ligand levels in patients with RCC were associated
with larger tumors (p 0.001), tumors of advanced stage
(p 0.017), grade (p 0.044), and tumors with necrosis (p
0.003). A doubling of these levels was associated with a 41%
increased risk of death (p 0.010).21 Thus, not only is the
upregulation of the PD-1 mechanism a potential prognostic
factor, but it has become a potential target for nextgeneration immune-based targeted therapy.
BMS 936558 is a PD-1 inhibitor and was studied in a
larger phase I multicenter trial where 126 patients (18 with
RCC) were treated with several dose levels.22 Across all
doses, the most common adverse events grades 3 to 4 were
fatigue (6.3%) and diarrhea (0.8%). One patient died with
sepsis while being treated for drug-related grade 4 pneumonitis. There were 18 patients with mRCC, 16 of whom were
treated with the 10 mg/kg dose. The objective investigator
assessed overall response rate was 31.2% (5 out of 16). The
median duration of treatment was 7.6 months and the
median duration of response was 4.0 months. Thus, a phase
II dose-varying trial for mRCC is underway (Table 2), with a
phase III trial anticipated shortly. This drug is promising
but remains investigational at this time.
To further immunotherapy and personalized medicine,
AGS-003 is being investigated. It is an autologous dendritic
301
HENG AND CHOUEIRI
cell immunotherapy in which a small tumor specimen isolated from the patient during nephrectomy or metastatectomy is taken along with a leukopheresis sample of
monocytes that differentiate into dendritic cells. They are
coelectroporated with the RCC and CD40L RNA and then
eventually injected back into the patient. Recently, AGS-003
was studied along with sunitinib in a phase II trial of 21
patients with no grade 3 or 4 adverse events. The overall
response rate was 38% and the median PFS was 11.2
months.23 Because almost half of the patients had poor
prognostic profiles,14 this PFS is encouraging and warrants
further study. Thus, a phase III randomized trial of
AGS-003 with sunitinib/standard therapy compared with

sunitinib/standard therapy is open and enrolling patients.
Conclusion
The treatment of mRCC has certainly evolved over the

last 5 years, with more and more treatments available.
There is an urgent need for biomarkers to help clinicians
select which drug is most suitable for each specific patient.
New drugs with different mechanisms of action are currently being investigated, making this an exciting time in
the realm of mRCC research.
Author
Daniel Y. C. Heng
Toni K. Choueiri
Employment or
Leadership
Positions
Consultant or
Advisory Role
Bayer; Novartis;
Pfizer
Bayer;
GlaxoSmithKline;
Novartis; Pfizer
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Pfizer
REFERENCES
1. Rini BI, Small EJ. Biology and clinical development of vascular endothelial growth factor-targeted therapy in renal cell carcinoma. J Clin Oncol.
2005;23:1028-1043.
2. Jubb AM, Pham TQ, Hanby AM, et al. Expression of vascular endothelial
growth factor, hypoxia inducible factor 1alpha, and carbonic anhydrase IX in
human tumours. J Clin Pathol. 2004;57:504-512.
3. Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon alfa
in metastatic renal-cell carcinoma. N Engl J Med. 2007;356:115-124.
4. Sternberg CN, Davis ID, Mardiak J, et al. Pazopanib in locally advanced
or metastatic renal cell carcinoma: results of a randomized phase III trial.
J Clin Oncol. 2010;28:1061-1068.
5. Rini BI, Halabi S, Rosenberg JE, et al. Phase III trial of bevacizumab
plus interferon alfa versus interferon alfa monotherapy in patients with
metastatic renal cell carcinoma: final results of CALGB 90206. J Clin Oncol.
2010;28:2137-2143.
6. Escudier BJ, Bellmunt J, Negrier S, et al. Final results of the phase III,
randomized, double-blind AVOREN trial of first-line bevacizumab (BEV)
interferon-2a (IFN) in metastatic renal cell carcinoma (mRCC). J Clin Oncol.
2009;27:239s (suppl; abstr 5020).
7. Hudes G, Carducci M, Tomczak P, et al. Temsirolimus, interferon alfa, or
both for advanced renal-cell carcinoma. N Engl J Med. 2007;356:2271-2281.
8. Escudier B, Eisen T, Stadler WM, et al. Sorafenib in advanced clear-cell
renal-cell carcinoma. N Engl J Med. 2007;356:125-134.
9. Motzer RJ, Escudier B, Oudard S, et al. Efficacy of everolimus in
advanced renal cell carcinoma: a double-blind, randomised, placebocontrolled phase III trial. Lancet. 2008;372:449-456.
10. Rini BI, Escudier B, Tomczak P, et al. Comparative effectiveness of
axitinib versus sorafenib in advanced renal cell carcinoma (AXIS): a randomised phase 3 trial. Lancet. 2011;378:1931-1939.
11. Negrier S, Gravis G, Perol D, et al. Temsirolimus and bevacizumab, or
sunitinib, or interferon alfa and bevacizumab for patients with advanced
renal cell carcinoma (TORAVA): a randomised phase 2 trial. Lancet Oncol.
2011;12:673-680.
12. Tang PA, Vickers MM, Heng DY. Clinical and molecular prognostic
factors in renal cell carcinoma: what we know so far. Hematol Oncol Clin
North Am. 2011;25:871-891.
13. Motzer RJ, Bacik J, Murphy BA, et al. Interferon-alfa as a comparative
302
treatment for clinical trials of new therapies against advanced renal cell
carcinoma. J Clin Oncol. 2002;20:289-296.
14. Heng DY, Xie W, Regan MM, et al. Prognostic factors for overall
survival in patients with metastatic renal cell carcinoma treated with
vascular endothelial growth factor-targeted agents: results from a large,
multicenter study. J Clin Oncol. 2009;27:5794-5799.
15. Heng DY, Xie W, Bjarnason GA, et al. A unified prognostic model for
first- and second-line targeted therapy in metastatic renal cell carcinoma
(mRCC): results from a large international study. J Clin Oncol. 2010;28:347s
16. Xu CF, Bing NX, Ball HA, et al. Pazopanib efficacy in renal cell
carcinoma: evidence for predictive genetic markers in angiogenesis-related
and exposure-related genes. J Clin Oncol. 2011;29:2557-2564.
17. van der Veldt AA, Eechoute K, Gelderblom H, et al. Genetic polymorphisms associated with a prolonged progression-free survival in patients with
metastatic renal cell cancer treated with sunitinib. Clin Cancer Res. 2011;17:
620-629.
18. Kim JJ, Vaziri SA, Rini BI, et al. Association of VEGF and VEGFR2
single nucleotide polymorphisms with hypertension and clinical outcome in
metastatic clear cell renal cell carcinoma patients treated with sunitinib.
Cancer. Epub 2011 Aug 31.
19. Garcia-Donas J, Esteban E, Leandro-Garcia LJ, et al. Single nucleotide
polymorphism associations with response and toxic effects in patients with
advanced renal-cell carcinoma treated with first-line sunitinib: a multicentre,
observational, prospective study. Lancet Oncol. 2011;12:1143-1150.
20. Rini BI, Cohen DP, Lu DR, et al. Hypertension as a biomarker of
efficacy in patients with metastatic renal cell carcinoma treated with
sunitinib. J Natl Cancer Inst. 2011;103:763-773.
21. Frigola X, Inman BA, Lohse CM, et al. Identification of a soluble form
of B7-H1 that retains immunosuppressive activity and is associated with
aggressive renal cell carcinoma. Clin Cancer Res. 2011;17:1915-1923.
22. McDermott DF, Drake CG, Sznol M, et al. A phase I study to evaluate
safety and antitumor activity of biweekly BMS-936558 (anti-PD-1, MDX1106/ONO-4538) in patients with RCC and other advanced refractory malignancies. J Clin Oncol. 2011;29 (suppl; abstr 331).
23. Figlin RA, Amin A, Dudek A, et al. Phase II study combining personalized dendritic cell (DC)-based therapy, AGS-003, with sunitinib in metastatic renal cell carcinoma (mRCC). J Clin Oncol. 2012;30 (suppl; abstr 348).
RECENT INNOVATIONS IN THE MANAGEMENT OF

GENITOURINARY CANCERS (eQ&A)
CHAIR
Matt D. Galsky, MD
The Tisch Cancer Institute, Mount Sinai School of Medicine
New York, NY
SPEAKERS
Elisabeth I. Heath, MD
Detroit, MI
Robert John Motzer, MD
New York, NY
New Developments in Urothelial Cancer

By Matthew D. Galsky, MD
Overview: Urothelial cancer is among the most chemotherapysensitive neoplasms of all the solid tumors. However, for the
majority of patients with advanced disease, response durations with conventional treatment are relatively short. Secondline systemic treatment regimens are associated with modest
response rates and poor outcomes. Trials in both the first- and
second-line settings have demonstrated that a ceiling in
efficacy has likely been reached with cytotoxic drugs, particularly in unselected patient populations. Promising areas of
ACH YEAR, approximately 56,000 patients in the

United States will be diagnosed with urothelial cancer
and more than 10,000 patients will succumb to the disease.1
Although the majority of patients present with clinically
localized disease, a substantial proportion will ultimately
develop metastatic recurrence whereas a smaller percentage
will have metastatic disease at the time of initial presentation. Once metastasis occurs, the median survival for patients with urothelial cancer is approximately 1 to 2 years.
In an attempt to improve outcomes in advanced disease,
research efforts have traditionally focused on the development of combination cytotoxic regimens, and more recently
have begun to unravel the molecular pathogenesis of urothelial cancer in an effort to target tumor-specific genetic
aberrations.
Chemotherapy in Advanced Urothelial Cancer
The Development of MVAC
During the 1980s, multiagent chemotherapeutic regimens

were developed in an attempt to exploit non cross-resistant
cytotoxics with somewhat nonoverlapping toxicity profiles.
In 1985, a landmark trial reported the initial experience
with the MVAC (methotrexate, vinblastine, doxorubicin,
and cisplatin) regimen.2 Remarkably, of the 24 patients
enrolled, responses were observed in 71% (95% CI, 53% to
89%), with complete clinical responses in 50% (95% CI, 30%
to 70%). Subsequent studies confirmed the activity of MVAC
in larger cohorts, albeit with slightly lower response proportions, and established MVAC as a standard of care. Attempts to improve on the results with MVAC have included
dose-dense administration with growth factor support. In a
randomized phase III trial of dose-dense compared with
conventional administration of MVAC, the former regimen
failed to detect a meaningful improvement in median survival, but did result in an improvement in 5-year survival
rate (21.8% [95% CI, 14.5% to 29.2%] compared with 13.5%
[95% CI, 7.4% to 19.6%; p 0.042]).3
Newer-Generation Cytotoxic Agents
During the 1990s, several newer cytotoxics, including the

taxanes and gemcitabine, were explored in urothelial cancer. Only a few of these regimens demonstrated sufficient
activity to advance to randomized phase III trials (Table 1).
These trials demonstrated superior outcomes with MVAC
compared with docetaxel plus cisplatin and uninterpretable
results from a comparison of MVAC with paclitaxel plus
carboplatin, as a result of early study closure.4-5 The results
of a randomized trial of gemcitabine plus cisplatin (GC)
304
investigation include integrating predictive biomarkers to

optimize patient selection for specific therapies, disrupting
driving oncogenomic mutations, and associated signaling
pathways and cotargeting both tumor and the immune system
or tumor stroma. In addition, expanded sources of evidence
generation are of interest in an effort to refine treatment for
the general population of patients with advanced urothelial
cancer, not only those who meet the narrow eligibility criteria
used in most clinical trials.
compared with MVAC, however, did alter the standard of

care.6 Although not designed as a noninferiority trial, this
study demonstrated similar antitumor outcomes, yet a favorable toxicity profile, with GC compared with MVAC.
Since the establishment of GC as a treatment standard in
advanced urothelial carcinoma, there have only been two
completed phase III trials attempting to improve on this
therapy. The addition of paclitaxel to the GC regimen did
not improve median survival in patients with advanced
urothelial carcinoma.7 Bamias and colleagues performed a
randomized phase III trial of dose-dense MVAC (n 118)
compared with dose-dense GC (gemcitabine 2,500 mg/m2
plus cisplatin 70 mg/m2 administered every 2 weeks, n
57).8 As a result of poor accrual, the MVAC arm was
supplemented with additional patients who received MVAC
off-study, leading to difficulty interpreting the overall study
results. The dose-dense GC arm demonstrated similar outcomes compared with the dose-dense MVAC arm (overall
response rate, 47.4% compared with 47.4%, p 0.9; median
survival, 18.4 months compared with 20.7 months, p 0.7).
However, the methodologic issues with this study have at
least, in part, limited the penetration of dose-dense GC into
routine practice.
Second-Line Chemotherapy
There have been only two completed phase III trials

evaluating chemotherapeutic regimens for patients experiencing disease progression despite first-line treatment. Bellmunt and colleagues randomly assigned 370 patients with
progressive urothelial cancer after first-line platinum-based
chemotherapy to vinflunine compared with best supportive
care.9 In the intent-to-treat population, treatment with
vinflunine did not result in a substantial improvement in
survival. However, when only eligible patients were analyzed, there was a substantially longer median survival with
vinflunine, resulting in marketing authorization by the
Committee for Medicinal Products for Human Use of the
European Medicines Agency.
The German Association of Urological Oncology performed a randomized phase III trial of gemcitabine plus
paclitaxel as second-line therapy in patients with advanced
From the Tisch Cancer Institute, Division of Hematology and Oncology, Department of
Medicine, Mount Sinai School of Medicine, New York, NY.
Address reprint requests to Matthew D. Galsky, MD, Mount Sinai School of Medicine,
1 Gustave L Levy Place, New York, NY 10029; email: matthew.galsky@mssm.edu.
1092-9118/10/1-10
NOVEL DEVELOPMENTS IN UROTHELIAL CANCER

Table 1. Randomized Phase III Trials of Cisplatin-Based
Chemotherapy in Advanced Urothelial Carcinoma
Regimen
MVAC
Cisplatin
MVAC
CISCA
MVAC
FAP
MVAC
DD-MVAC
MVAC
Gemcitabine cisplatin
MVAC
Docetaxel cisplatin
MVAC
Paclitaxel carboplatin
paclitaxel
DD-MVAC
DD-gemcitabine cisplatin
Response (%)
No. of
Patients
Overall
Complete
Survival
(months)
120
126
55
55
86
83
129
134
205
203
109
111
44
41
314
312
36
11
65
46
59
42
58
72
46
50
54
37
40
28
44
56
13
3
35
25
24
10
9
21
12
12
23
13
13
3
11
14
12.5
8.2
12.6
10
12.5
12.5
14.1
15.5
14.8
13.8
14.2
9.3
14.2
13.8
12.7
15.8
118
57
47
47
15
10
18.4
20.7
p
0.0002
0.05
0.17
Cisplatin-Ineligible Patients
0.122
There have not been adequately powered randomized

phase III trials evaluating cisplatin-compared with
carboplatin-based therapy in advanced urothelial cancer to
address a potential survival benefit with the former regimens. However, a meta-analysis of randomized trials revealed that cisplatin-based chemotherapy was associated
with a significant increase in the likelihood of achieving an
objective response (response rate [RR] 1.33, [95% CI, 1.04
to 1.71], p 0.025) or a complete response (RR 3.54 [95%
CI, 1.48 to 8.49], p 0.004) compared with carboplatinbased chemotherapy.14 Although these data support the
preference for cisplatin-based regimens, urothelial cancer is
largely a disease of the elderly and as a result of multiple
reasons including impairment in renal function and performance status, approximately 30% to 50% of patients are
deemed ineligible for cisplatin-based chemotherapy.15 Investigators have long appreciated the disconnect between
the efficacy of cisplatin-based therapy and the effectiveness
of this treatment when applied to the general population of
patient with bladder cancer and have designed trials specifically for cisplatin-ineligible patients.
Two randomized phase III trials have been initiated in the
cisplatin-ineligible population. EORTC 30986 was a randomized phase II/III trial of gemcitabine plus carboplatin
(GCa) compared with methotrexate, vinblastine, and carboplatin (M-CAVI) in unfit patients (WHO performance status of 2 and/or creatinine clearance 30 60 mL/min) with
metastatic urothelial cancer.16 There was no significant
difference in the progression-free survival (GCa, 5.8 months;
M-CAVI, 4.2 months; hazard ratio [HR], 1.04; 95% CI, 0.8 to
1.35; p 0.78) or the overall survival (GCa, 9.3 months;
M-CAVI, 8.1 months; HR, 0.94; 95% CI, 0.72 to 1.22; p
0.64) between the treatment arms. However, GCa was better
tolerated, solidifying the role of this regimen as a standard
in the unfit population. A phase III trial comparing vinflunine plus gemcitabine compared with placebo plus gemcitabine closed early as a result of poor accrual.
Trials in the unfit patient population have been hampered by a lack of a standard definition of what constitutes
cisplatin ineligibility. In an effort to develop a consensus
definition of patients with metastatic urothelial cancer unfit for cisplatin-based chemotherapy, a working group assembled and surveyed genitourinary oncologists with the
goal of establishing uniform eligibility criteria for future
clinical trials.17 According to this consensus definition, patients meeting at least one of the following criteria were
considered unfit for cisplatin: Eastern Cooperative Oncology
Group (ECOG) performance status of 2, creatinine clearance
below 60 mL/min, grade 2 or greater hearing loss, grade 2 or
greater neuropathy, and/or New York Heart Association
Class III heart failure.
0.746
0.025
0.41
0.075
0.7
Abbreviations: MVAC, methotrexate, vinblastine, doxorubicin, cisplatin;

CISCA, cyclophosphamide, cisplatin, doxorubicin; CMV, cisplatin, methotrexate,
vinblastine; MV, methotrexate, vinblastine; FAP, fluorouracil, interferon 2b,
cisplatin; DD, dose dense.
urothelial cancer, with the primary goal of evaluating the

impact of duration of therapy.10 In this trial, 102 patients
were randomly assigned to short-term (six cycles) compared with prolonged therapy (treatment until disease progression) with gemcitabine plus paclitaxel. Notably, the
response rate with both treatment arms was approximately
40%. Neither overall survival nor progression-free survival
differed between the treatment arms, although the pro-
KEY POINTS
longed schedule was difficult to evaluate because the majority of patients experienced rapid disease progression and/or
toxicity.
In the absence of Level 1 evidence, the taxanes or pemetrexed are commonly used in the second-line setting on the
basis of modest activity demonstrated in single-arm phase II
trials.11-13
Cisplatin-based combination chemotherapy regimens, particularly methotrexate, vinblastine, doxorubicin and cisplatin (MVAC) and gemcitabine plus
cisplatin, have long been the first-line treatment
standards for advanced urothelial cancer.
There have been no improvements in the efficacy of
first-line chemotherapy for advanced urothelial cancer in the last 30 years.
A large proportion urothelial carcinomas are considered unfit for cisplatin, and the combination of gemcitabine plus carboplatin is a reasonable standard on
the basis of phase III testing.
Ongoing efforts to improve outcomes focus on integrating antiangiogenic agents, targeting oncogenes
and associated signaling pathways, stimulating an
antitumor immune response, and refining patient
selection for specific treatments.
Expansion of the evidence base beyond randomized
trials is likely needed to help address many unanswered questions encountered in the daily care of
patients with advanced urothelial cancer.
305
MATTHEW D. GALSKY
Novel Approaches in Advanced Urothelial Cancer
Identifying Predictive Biomarkers and Therapeutic Targets
Because urothelial cancer is relatively chemotherapy sensitive, yet only a fraction of patients respond to any given
regimen, there has been much interest in developing tools
to allow more rational use of existing and future drugs.
DNA-repair genes, or their protein products, have been of
particular interest as predictive biomarkers on the basis of
the mechanism of action of the conventional cytotoxics used
in the management of urothelial cancer. In a retrospective
study, levels of the DNA-repair genes ERCC1, RRM1,
BRCA1, and caveolin-1, in tumor tissue from 57 patients
with bladder cancer treated with cisplatin-based combination chemotherapy, were analyzed.18 The median survival
of patients in this cohort was higher in patients with low
ERCC1 levels (25.4 vs. 15.4 months; p 0.03). However,
development of ERCC1 as a potential prognostic and/or
predictive biomarker has been hampered by difficulties with
analytic validation.
The genomic complexity of most solid tumors suggests
that relying on a single gene or protein as a predictive
biomarker is unlikely to yield major improvements in patient selection. Alternatively, gene signatures of response
to chemotherapy may be more attractive. The conventional
approach to predictive gene expression model development
is limited, however, because the signature is applicable only
to the particular drug or combination and patient population
for which the signature was developed. To overcome these
limitations, Theodorescu and colleagues have developed a
novel bioinformatics approach known as Coexpression Extrapolation (COXEN).19 COXEN uses the publicly available
gene-expression profiling data and drug sensitivity data
from the NCI-60 cell line panel as a Rosetta Stone to
predict chemotherapy sensitivity of gene-expressionprofiled tumor samples. A neoadjuvant study to demonstrate
the clinical utility of COXEN in treatment selection for
patients with urothelial cancer is currently being planned.
Profiling tumors for potentially actionable genomic mutations also offers the potential for personalized application
of targeted therapeutics. Sjodahl and colleagues performed
mutation analyses of 16 genes (FGFR3, PIK3CA, PIK3R1,
PTEN, AKT1, KRAS, HRAS, NRAS, BRAF, ARAF, RAF1,
TSC1, TSC2, APC, CTNNB1, and TP53) in 145 urothelial
cancer samples.20 Notably, this study confirmed that FGFR3
and PIK3CA mutations were most commonly found in noninvasive low-grade tumors, although a smaller proportion of
muscle-invasive specimens also harbored these mutations.
The potential importance of adenomatous polyposis coli
signaling and the mammalian target of rapamycin (mTOR)
regulatory tuberous sclerosis complex genes (TSC2 and
TSC2) were also identified in this study. Actionable mutations/aberrations found in other solid tumors, such EGFR
mutations and EML4-ALK fusions, have been identified
very rarely in urothelial cancer specimens.
Targeting Oncogenes
Although activating mutations in EGFR are not detected

in urothelial cancer, overexpression or amplification of
EFGR pathway family members are present in a subset of
tumors, and have been implicated in the pathogenesis of the
disease. Multiple phase II trials have explored single-agent
inhibitors of the EGFR pathway using different agents,
306
methods of patient selection, and trial designs. A study of

lapatinib as second-line therapy in patients with metastatic
urothelial cancer reported a response rate of only 1.7%;
however, stable disease was correlated with EGFR overexpression, and, to some extent, HER2 overexpression.21 A
randomized discontinuation trial of lapatinib in patients
with tumors harboring HER2 gene amplification, which
included patients with urothelial cancer, demonstrated no
objective responses in the urothelial cancer cohort.22 Trials
combining chemotherapy with inhibitors of the EGFR pathway have yielded more intriguing results, but the single-arm
nature of these trials has precluded an assessment of the
contribution of the EGFR pathway inhibitor, and few randomized trials have been initiated.
Although activating mutations in FGFR3 are most commonly found in nonmuscle-invasive urothelial cancer, recent studies have demonstrated the presence of FGFR3
mutations in 10% to 20% of muscle-invasive urothelial
cancer specimens.23 Dovitinib is a small-molecule inhibitor
of several tyrosine kinase receptors, including the vascular
endothelial growth factor receptor (VEGFR) and fibroblast
growth factor receptor (FGFR), and has demonstrated inhibition of tumor growth and proliferation in urothelial carcinoma models. A multicenter, two-stage, open-label phase II
trial is currently evaluating the safety and efficacy of
dovitinib in patients with advanced urothelial carcinoma
who have experienced disease progression despite prior
systemic therapy, both in cohorts with and without FGFR3
mutations (NCT00790426).
Targeting Angiogenesis
Multiple lines of evidence support a therapeutic role for

targeting the tumor vasculature in urothelial cancer. The
triplet of gemcitabine, cisplatin, and the anti-VEGF antibody bevacizumab in a phase II of 43 chemotherapy-nave
patients with metastatic urothelial cancer, demonstrated an
overall response rate of 73% and an encouraging median
overall survival of 19.1 months; it is currently being explored in a randomized trial by the Cancer and Leukemia
Group B.24
On the basis of preclinical studies demonstrating synergistic antitumor activity by combining VEGFR and plateletderived growth factor receptor (PDGFR) kinase inhibitors,
phase II trials of sorafenib, sunitinib, and pazopanib, have
been performed in urothelial cancer and demonstrated modest antitumor activity with the latter drugs.25-26 Although
supported by preclinical studies showing additive to synergistic activity, the combination of GC plus sunitinib was
poorly tolerated in a phase II trial as a result of hematologic
toxicity.27
Two randomized phase II trials evaluating the impact of
adding antiangiogenic therapy to cytotoxic regimens in
urothelial carcinoma have been reported. In a placebocontrolled phase II trial of docetaxel with or without vandetanib, a small-molecule inhibitor of the VEGFR and
epithelial growth factor receptor (EGFR) tyrosine kinases, in
the second-line setting, the combination failed to improve
outcomes.28 Similarly, a trial of GC with or without
sorafenib failed to improve outcomes, but closed early as a
result of poor accrual limiting the power to detect differences
between the arms.29
NOVEL DEVELOPMENTS IN UROTHELIAL CANCER

Targeting the Immune System
Preclinical and clinical studies suggest that urothelial

cancer is immunogenic. For example, higher numbers of
CD8-positive tumor-infiltrating lymphocytes in bladder cancer specimens have been correlated with substantial improved disease-free and overall survival.30 Despite the
immunogenicity of urothelial cancer, patients with urothelial cancer also exhibit tumor-associated immunologic tolerance. Blocking immune regulatory checkpoints may
overcome tumor-induced immune tolerance in urothelial
cancer. Ipilimumab, a human immunoglobulin G (IgG1)
anti-CTLA-4 monoclonal antibody, was administered before
cystectomy in a proof-of-concept study of 12 patients with
clinically localized bladder cancer.31 Treatment resulted in
perivascular infiltration of cells positive for CD3, CD8, CD4,
and granzyme. In addition, patients demonstrated an increase in CD4-positive T cells with high expression of
inducible costimulator (ICOShi) in tumor tissue and systemic circulation, and eight of 12 patients experienced
downstaging of their primary tumor. A phase II trial of
sequential chemotherapy followed by chemotherapy plus
ipilimumab, in an attempt to autovaccinate patients before
introduction of immune checkpoint blockade, has recently
been initiated.
Given the high risk of recurrence in patients with muscleinvasive bladder cancer after cystectomy, immune-based
approaches have also been of interest in decreasing the
likelihood of relapse. Lapuleucel-T is an activated cellular
therapy consisting of autologous peripheral blood mononuclear cells cultured ex vivo with BA7072, a recombinant
fusion antigen consisting of portions of the intracellular and
extracellular regions of HER2/neu linked to granulocytemacrophage colony-stimulating factor. A placebo-controlled
randomized phase II trial is ongoing exploring the impact of
Lapuleucel-T on overall survival in patients with Her-2

expressing muscle-invasive bladder cancer after cystectomy.
The Rapid Learning Health System and
Urothelial Cancer
Although randomized controlled clinical trials are the

standard for evidence development in medicine, given issues
of feasibility and limited patient and economic resources,
there are questions that arise in the daily care of patients
with metastatic urothelial cancer that will never be answered by randomized trials. Furthermore, the pace of
evidence development through traditional clinical trials is
painstakingly slow. In the meantime, only a small percentage of patients treated for cancer have their clinical data
captured for research purposes. Complementary approaches
to evidence generation must be embraced to support realtime clinical practice decisions and to enhance the care of all
patients with urothelial cancer, not just those who fit the
narrow eligibility criteria used in randomized trials. The
rapid learning health care model envisions the use of routinely collected real-time clinical data, from electronic
medical records or other sources, potentially linked to
omic data, to continuously drive both scientific discovery
and patient care through an iterative process.32 However,
linking data across institutions requires investment,
information-technology infrastructure, and attention to
privacy regulations. Pilot efforts such as the Retrospective
International Study of Invasive/Advanced Cancer of the
Urothelium (RISC; www.mssm.edu/research/programs/
genitourinary-cancer-research/risc) are currently underway
to link clinical data derived from patients with bladder
cancer from a variety of institutions to demonstrate feasibility, standardize data collection and reporting, establish
infrastructure, and ultimately aid in the design of future
projects that will bring rapid learning to the care of patients
with bladder cancer.
Author
Matthew D. Galsky
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Amgen; AVEO;
Bristol-Myers
Squibb;
GlaxoSmithKline;
Pfizer
Research
Funding
Expert
Testimony
Other
Remuneration
Celgene;
Novartis; Pfizer;
Viatar
REFERENCES
J Clin. 2011;61:69-90.
2. Sternberg CN, Yagoda A, Scher HI, et al. Preliminary results of M-VAC
(methotrexate, vinblastine, doxorubicin and cisplatin) for transitional cell
carcinoma of the urothelium. J Urol. 1985;133:403-407.
3. Sternberg CN, de Mulder P, Schornagel JH, et al. Seven year update of
an EORTC phase III trial of high-dose intensity M-VAC chemotherapy and
G-CSF versus classic M-VAC in advanced urothelial tract tumours. Eur J
Cancer. 2006;42:50-54.
4. Bamias A, Aravantinos G, Deliveliotis C, et al. Docetaxel and cisplatin
with granulocyte colony-stimulating factor (G-CSF) versus MVAC with
G-CSF in advanced urothelial carcinoma: a multicenter, randomized, phase
III study from the Hellenic Cooperative Oncology Group. J Clin Oncol.
2004;22:220-228.
5. Dreicer R, Manola J, Roth BJ, et al. Phase III trial of methotrexate,
vinblastine, doxorubicin, and cisplatin versus carboplatin and paclitaxel in
patients with advanced carcinoma of the urothelium. Cancer. 2004;100:16391645.
6. von der Maase H, Hansen SW, Roberts JT, et al. Gemcitabine and
cisplatin versus methotrexate, vinblastine, doxorubicin, and cisplatin in
advanced or metastatic bladder cancer: results of a large, randomized,
multinational, multicenter, phase III study. J Clin Oncol. 2000;18:3068-3077.
7. Bellmunt J, Von der Maase H, Mead GM, et al. Randomized phase III
study comparing paclitaxel/cisplatin/gemcitabine and gemcitabine/cisplatin
in patients with locally advanced or metastatic urothelial cancer without
prior systemic therapy; EORTC30987/Intergroup Study. J Clin Oncol.
2007;25 (suppl; abstr 5030).
8. Bamias A, Karadimou A, Lampaki S, et al. Prospective, randomized
phase III study comparing two intensified regimens (methotrexate/vinblastine/doxorubicin hydrochloride/cisplatin [MVAC] versus gemcitabine/cisplatin) in patients with inoperable or recurrent urothelial cancer. J Clin Oncol.
2011;29 (suppl; abstr 4510).
9. Bellmunt J, Theodore C, Demkov T, et al. Phase III trial of vinflunine
plus best supportive care compared with best supportive care alone after a
platinum-containing regimen in patients with advanced transitional cell
carcinoma of the urothelial tract. J Clin Oncol. 2009;27:4454-4461.
307
MATTHEW D. GALSKY
10. Albers P, Park SI, Niegisch G, et al. Randomized phase III trial of 2nd
line gemcitabine and paclitaxel chemotherapy in patients with advanced
bladder cancer: short-term versus prolonged treatment [German Association
of Urological Oncology (AUO) trial AB 20/99]. Ann Oncol. 2011;22:288-294.
11. Galsky MD, Mironov S, Iasonos A, et al. Phase II trial of pemetrexed as
second-line therapy in patients with metastatic urothelial carcinoma. Invest
New Drugs. 2007;25:265-270.
12. McCaffrey JA, Hilton S, Mazumdar M, et al. Phase II trial of docetaxel
in patients with advanced or metastatic transitional-cell carcinoma. J Clin
Oncol. 1997;15:1853-1857.
13. Sweeney CJ, Roth BJ, Kabbinavar FF, et al. Phase II study of
pemetrexed for second-line treatment of transitional cell cancer of the
urothelium. J Clin Oncol. 2006;24:3451-3457.
14. Galsky MD, Chen GJ, Oh WK, et al. Comparative effectiveness of
cisplatin-based and carboplatin-based chemotherapy for treatment of advanced urothelial carcinoma. Ann Oncol. 2012;23:406-410.
15. Dash A, Galsky MD, Vickers AJ, et al. Impact of renal impairment on
eligibility for adjuvant cisplatin-based chemotherapy in patients with urothelial carcinoma of the bladder. Cancer. 2006;107:506-513.
16. De Santis M, Bellmunt J, Mead G, et al. Randomized phase II/III trial
comparing gemcitabine/carboplatin (GC) and methotrexate/carboplatin/vinblastine (M-CAVI) in patients (pts) with advanced urothelial cancer (UC)
unfit for cisplatin-based chemotherapy (CHT): phase III results of EORTC
study 30986. J Clin Oncol. 2010;28 (suppl; abstr LBA4519).
17. Galsky MD, Hahn NM, Rosenberg J, et al. Treatment of patients with
metastatic urothelial cancer unfit for cisplatin-based chemotherapy. J Clin
Oncol. 2011;29:2432-2438.
18. Bellmunt J, Paz-Ares L, Cuello M, et al. Gene expression of ERCC1 as
a novel prognostic marker in advanced bladder cancer patients receiving
cisplatin-based chemotherapy. Ann Oncol. 2007;18:522-528.
19. Williams PD, Cheon S, Havaleshko DM, et al. Concordant gene expression signatures predict clinical outcomes of cancer patients undergoing
systemic therapy. Cancer Res. 2009;69:8302-8309.
20. Sjodahl G, Lauss M, Gudjonsson S, et al. A systematic study of gene
mutations in urothelial carcinoma; inactivating mutations in TSC2 and
PIK3R1. PLoS One. 2011;6:e18583.
21. Wulfing C, Machiels JP, Richel DJ, et al. A single-arm, multicenter,
open-label phase 2 study of lapatinib as the second-line treatment of patients
308
with locally advanced or metastatic transitional cell carcinoma. Cancer.

2009;115:2881-2890.
22. Galsky MD, Von Hoff DD, Neubauer M, et al. Target-specific, histologyindependent, randomized discontinuation study of lapatinib in patients with
HER2-amplified solid tumors. Invest New Drugs. Epub 2010 Sep 22.
23. Kompier LC, Lurkin I, van der Aa MN, et al. FGFR3, HRAS, KRAS,
NRAS and PIK3CA mutations in bladder cancer and their potential as
biomarkers for surveillance and therapy. PLoS One. 2010;5:e13821.
24. Hahn NM, Stadler WM, Zon RT, et al. Phase II trial of cisplatin,
gemcitabine, and bevacizumab as first-line therapy for metastatic urothelial
carcinoma: Hoosier Oncology Group GU 04-75. J Clin Oncol. 2011;29:15251530.
25. Dreicer R, Li H, Stein M, et al. Phase 2 trial of sorafenib in patients
with advanced urothelial cancer: a trial of the Eastern Cooperative Oncology
Group. Cancer. 2009;115:4090-4095.
26. Gallagher DJ, Milowsky MI, Gerst SR, et al. Phase II study of sunitinib
in patients with metastatic urothelial cancer. J Clin Oncol. 2010;28:13731379.
27. Galsky MD, Sonpavde G, Hellerstedt BA, et al. Phase II study of
gemcitabine, cisplatin, and sunitinib in patients with advanced urothelial
carcinoma. Presented at the ASCO Genitourinary Oncology Symposium,
March 5-7, 2010, San Francisco, CA (abstr 276).
28. Choueiri T, Vaishampayan U, Yu EY, et al. A double-blind randomized
trial of docetaxel plus vandetanib versus docetaxel plus placebo in platinumpretreated advanced urothelial cancer. Presented at the ASCO Genitourinary
Oncology Symposium, March 5-7, 2010, San Francisco, CA (abstr LBA239).
29. Krege S, Rexer H, vom Dorp F, et al. Gemcitabine and cisplatin with or
without sorafenib in urothelial carcinoma (AUO-AB 31/05). J Clin Oncol.
2010;28 (suppl; abstr 4574).
30. Sharma P, Shen Y, Wen S, et al. CD8 tumor-infiltrating lymphocytes
are predictive of survival in muscle-invasive urothelial carcinoma. Proc Natl
Acad Sci U S A. 2007;104:3967-3972.
31. Carthon BC, Wolchok JD, Yuan J, et al. Preoperative CTLA-4 blockade:
tolerability and immune monitoring in the setting of a presurgical clinical
trial. Clin Cancer Res. 2010;16:2861-2871.
for cancer care. J Clin Oncol. 2010;28:4268-4274.
New Developments in Prostate Cancer Therapy

By Madhuri Bajaj, MD, and Elisabeth I. Heath, MD
Overview: Prostate cancer is the most common nonskin

malignant neoplasm in men worldwide. In the United States,
241,740 new diagnoses of prostate cancer and 28,170 prostate
cancer deaths have been estimated for 2012, representing
28% of new cancer cases and 10% of male cancer deaths. 1
Although metastatic prostate cancer remains an incurable
disease, substantial advances have been made in therapeutic
options available for men in the past several years. Development of novel agents that modulate the androgen receptor
pathway, growth factor signaling pathways, and immune function and bone targeting pathways has been the focus of
therapeutic strategies because of its significance in the biology of prostate cancer progression. Several of the agents
have gained U.S. Food and Drug Administration (FDA) approval, whereas many are in late-stage clinical trials. With the
growth of available treatment options, a major challenge as
we move forward will be to determine the best sequence
and/or combination of therapy that will result in maximum
clinical efficacy with minimum toxicity. Highlighted in this
publication are several of the exciting advances in prostate
cancer therapy for patients with metastatic, castrate-resistant
prostate cancer.
NDROGENS ARE the key regulators of cell growth

and proliferation in prostate cancer. Androgen deprivation therapy is initially a highly effective therapy because
of the induction of apoptosis. Persistent androgen receptor
(AR) activation is an important mediator of disease progression in castrate-resistant prostate cancer (CRPC).2 There
are multiple mechanisms by which this activation happens,
including AR overexpression, AR mutations that increase
androgen sensitivity to or activation by other steroids,
increased local androgen production by prostate cells via
expression of steroidogenic enzymes, AR activation via
crosstalk of signal transduction pathways (epidermal
growth factor, insulinlike growth factor, interleukin 6),
modulated expression of coactivators or corepressors of AR,
and proteolytic processing of AR to an androgenindependent isoform.3 Preclinical research has validated
these concepts and thus has served as the basis for the
translation of novel, potent AR-targeted therapies for patients with prostate cancer who experience relapse after
initial androgen inhibition. Several clinical studies have
demonstrated that CRPC cells continue to be under the
influence of androgen signaling as evidenced by the high
number of AR expression.4 As a result, these newer agents
tend to be more specific targets of enzymes downstream in
the hormonal cascade.
PSA response rate (29% vs. 6%, p 0.001), favored abiraterone.7 Common adverse effects with this agent include
hypokalemia, hypertension, and pedal edema. The effects
are explained by a syndrome of mineralocorticoid excess. On
the basis of these results, the FDA granted approval in April
2011 of abiraterone for the treatment of patients with
metastatic CRPC whose disease had progressed regardless
of their docetaxel-based chemotherapy. The second of the
phase III trials is investigating abiraterone in asymptomatic
or mildly symptomatic men with metastatic CRPC who had
not received prior chemotherapy. This trial has completed
accrual, with final results pending (NCT00887198).
The clinical trial of abiraterone compared with placebo in
the postdocetaxel population is also important because considerable progress has been made in the area of circulating
tumor cells. Evaluation of circulating tumor cells was embedded in the phase III trial as a potential surrogate
endpoint for overall survival. At the 2011 Annual Meeting of
the American Society of Clinical Oncology, results from this
phase III trial confirmed that pretreatment circulating tumor cells and lactate dehydrogenase, alone and in combination, served as prognostic biomarkers.8 Interestingly, PSA
did not. This important trial will set the foundation for
future trials that incorporate biomarkers as surrogate endpoints.
Abiraterone
TAK-700
Abiraterone acetate is more potent than ketoconazole and

a selective inhibitor of the 17-hydroxylase and the C17,
20-lyase function of CYP17A.5 In a phase II trial, 47 men
with metastatic CRPC deemed refractory to docetaxel-based
chemotherapy were given oral abiraterone at 1,000 mg/d.
The results were prostate-specific antigen (PSA) decreases
of 30% or higher, 50% or higher, and 90% or higher seen in
32 of 47 patients (68%), 24 of 47 patients (51%), and 7 of 47
patients (15%), respectively.6 These results, coupled with a
favorable safety profile, have laid the foundation for the
development of two randomized, double-blind, placebocontrolled, phase III clinical trials. Recently, in the first of
the phase III trials, 1,195 patients with metastatic CRPC
previously treated with docetaxel who were given abiraterone plus prednisone showed improved overall survival
compared with those given placebo plus prednisone (median
overall survival time, 14.8 vs. 10.9 months; hazard ratio
[HR], 0.65; p 0.0001). All secondary end points, including
time to PSA progression (10.2 vs. 6.6 months; p 0.001),
progression-free survival (5.6 vs. 3.6 months; p 0.001), and
TAK-700 is a novel, selective CYP450c17 inhibitor similar

to abiraterone in its mechanism of reducing testosterone and
dehydroepiandrosterone levels. In a phase I/II study of this
compound in asymptomatic patients with metastatic CRPC,
the drug was tolerated well at various doses, and there was
a 50% decrease in 12 of 15 patients who were treated with
300 mg or more twice daily for 3 months or longer.9 No
dose-limiting toxicity was seen, and the most common adverse events were fatigue (62%), nausea (38%), constipation
(35%), and vomiting (30%). The preliminary phase I/II study
results have led to a phase II, ongoing evaluation of TAK-
From the Karmanos Cancer Institute/Wayne State University School of Medicine, Detroit,
MI.
Address reprint requests to Elisabeth I. Heath, MD, Karmanos Cancer Institute/Wayne
State University School of Medicine, 4100 John R, Detroit, MI; email: heathe@
karmanos.org.
1092-9118/10/1-10
309
BAJAJ AND HEATH
700 at a dose of 400 mg twice daily with prednisone in

patients with metastatic CRPC.8 There are now two phase
III, multicenter, randomized, double-blind trials that are
evaluating TAK-700 plus prednisone compared with placebo
plus prednisone in patients with metastatic CRPC. One trial
is evaluating chemotherapy-naive patients (primary endpoints: overall survival and radiographic progression-free
survival)10 (NCT01193244), whereas the other focuses on
the postdocetaxel progression patient population (primary
endpoint: overall survival)11 (NCT01193257). Both trials are
actively recruiting, with a target accrual of 1,000 to 1,400
patients, and results are expected to be available by 2013
2014. Accrual to the postdocetaxel chemotherapy protocol
may be more challenging in the future because of the FDA
approval of abiraterone and the positive phase III results of
MDV3100.
Studies evaluating combination therapy of abiraterone or
TAK-700 with docetaxel chemotherapy are under way
(NCT01400555 and NCT01084655, respectively). These results are eagerly awaited because each of these agents
results in a reasonable PSA response, leading to the unanswered question of whether the synergistic combination will
produce an even stronger PSA response.
MDV3100
Another hormonally driven strategy is to target the AR

directly. MDV3100 is an oral AR antagonist that directly
inhibits AR by irreversibly binding to the receptor. This
interaction impairs AR nuclear translocation, DNA binding,
and recruitment of coactivators.12 Preclinical studies have
demonstrated that MDV3100 is a more potent binder to the
AR receptor than bicalutamide, thus leading to complete
suppression of the AR pathway.13 In a phase I/II study,
MDV3100 showed antitumor activity in patients with metastatic CRPC.14 In this trial, 56% of 140 patients demon-
KEY POINTS
310
Currently, multiple new treatment options approved

by the US Food and Drug Administration (FDA) are
available for the treatment of metastatic castrateresistant prostate cancer (CRPC), including cabazitaxel, sipuleucel-T, and abiraterone.
MDV3100, an androgen receptor antagonist, and
radium-223, a novel radiopharmaceutical, have also
shown improvement in overall survival when compared with placebo in men and will likely have a role
in treating patients with CRPC after chemotherapy.
Denosumab, a monoclonal antibody to RANK ligand,
is now an FDA-approved therapy for the prevention
of skeletal-related events from bone metastases secondary to prostate cancer.
Additional novel agents that inhibit androgen, angiogenesis, cell survival and signaling, bone interface,
and immunologic pathways are actively being investigated in ongoing phase III clinical trials.
Optimal sequencing and/or combination therapy of
active agents has yet to be determined in men with
metastatic CRPC.
strated decreases in serum PSA of 50% or more, and 61 of

the 109 patients had stabilized bone disease after treatment.
The Atrial Fibrillation Follow-up Investigation of Rhythm
Management (AFFIRM) trial is a randomized, double blind,
placebo-controlled, multinational trial of 1,199 men assigned to 160 mg/d of MDV3100 (800 patients) or placebo
(399 patients). MDV3100 was associated with a median
overall survival of 18.4 months compared with 13.6 months
for patients assigned to placebo (HR, 0.631). Median
progression-free survival also favored MDV3100 (8.3 vs. 2.9
months).15 Approximately 30% of patients assigned to
MDV3100 had complete or partial response compared with
1.3% in the placebo group. The drug also was associated with
a PSA reduction of at least 50% from baseline in 54% of the
MDV3100 group compared with 1.5% of the placebo group
and at least a 90% reduction from baseline in 25% of the
MDV3100 group compared with 1% of the placebo group.
Median time to PSA progression was 8.3 months in the
experimental group compared with 3 months in the placebo
group. MDV3100 was well tolerated. The most common
adverse events that occurred more frequently in the
MDV3100 group ( 2%) than in the placebo group included
fatigue, diarrhea, and hot flushes. Five of 800 patients
treated with MDV3100 in the study reported seizures compared with no seizures among the placebo arm. The 0.6%
seizure rate attributed to MDV3100 is below the approximate 1.5% rate seen in an earlier study of the drug using
higher doses.
The Prevention of VTE after Acute Ischemic Stroke with
LMWH Enoxaparin (PREVAIL) trial is a phase III trial
evaluating patients with chemotherapy-naive CRPC treated
with 160 mg/d of MDV3100 with standard of care compared
with placebo with standard of care (NCT01212991). With a
target accrual goal of 1,700 patients nearly complete, the
study has the primary endpoints of overall survival and
progression-free survival and secondary endpoints of time
to initiation of cytotoxic chemotherapy and time to first
skeletal-related event (SRE).16
ARN-509 is a second-generation antiandrogen that is
currently undergoing clinical evaluation. ARN-509 inhibits
both AR nuclear translocation and AR binding to androgen
response elements in DNA. The compound also does not
exhibit agonist activity in prostate cancer cells that overexpress AR. In a recent study, ARN-509 was optimized for
inhibition of AR transcriptional activity and prostate cancer
cell proliferation, pharmacokinetics, and in vivo efficacy. In
a clinically valid murine xenograft model of human CRPC,
ARN-509 showed greater efficacy than MDV3100. Maximal
therapeutic response in this model was achieved at 30 mg/kg
daily of ARN-509, whereas the same response required 100
mg/kg daily of MDV3100 and higher steady-state plasma
concentrations.17 Thus, ARN-509 appears to have a higher
therapeutic index than current AR antagonists. ARN-509
seems to be a promising therapy in both castration-sensitive
and castration-resistant forms of prostate cancer. Currently,
it is in phase II clinical trials (NCT01171898).
Agents such as abiraterone, TAK-700, and MDV3100 are
appealing options even during clinical trials because patients are often inclined to consider treatment with pills
rather than systemic chemotherapy. Furthermore, there is a
familiarity of androgen modulators as therapy because patients undergoing treatment with luteinizing hormonereleasing hormone or gonadotropin-releasing hormone
DEVELOPMENTS IN PROSTATE CANCER THERAPY
agents have most likely been treated with antiandrogens,

such as flutamide and nilutamide, in the past. Finally,
agents modulating the androgen signaling axis frequently
result in PSA reduction, another familiar signal of treatment effect. However, a major challenge now is how to best
position these oral agents to maximize efficacy. MDV3100
has shown success in patients in whom docetaxel-based
chemotherapy has failed, which is in the same patient group
as those who benefited from abiraterone. Therefore, with
two highly active oral agents, there will be concerns regarding which agent should be initially administered in patients
in whom chemotherapy has failed. The treatment landscape
will also be affected when the results of the clinical trials
conducted in patients before chemotherapy are available.
The sequencing and/or role of combination therapy is important and practical and must be further explored in future
clinical trials.
Cell Signaling Pathways
Angiogenesis, the process of new blood vessel formation, is

a crucial step in the propagation of malignant tumor growth
and metastasis. Among the multiple proangiogenic factors
that promote the process of vessel formation, vascular endothelial growth factor (VEGF) is one of the most important.
Bevacizumab is a humanized monoclonal antibody directed
against VEGF-A and causes potent inhibition of VEGF
receptor (VEGFR) signaling and angiogenesis. Bevacizumab
is approved for use in combination with chemotherapy for
patients with metastatic colorectal, breast, and lung cancers. In prostate cancer, bevacizumab has been evaluated
in several clinical trials, including the Cancer and Leukemia
Group B (CALGB) phase II trial of bevacizumab in combination with docetaxel and estramustine in 79 patients with
metastatic CRPC.18 A PSA decrease of more than 50%
from baseline occurred in 81% of patients, the median time
to progression was 9 months, and overall survival was
21 months. These favorable trials led to a recent phase
III randomized placebo-controlled trial of docetaxel, prednisone, and bevacizumab compared with docetaxel and
prednisone in 1,050 patients with chemotherapy-naive metastatic CRPC with the primary endpoint of overall survival.
Final results published in 2011 reported that although there
was median progression-free survival in the bevacizumab
arm of 9.9 months compared with the 7.5 months of the
control arm (p 0.0001), the overall survival time was not
statistically significant.19 Bevacizumab has notable toxicities, including hypertension, thromboembolism, hemorrhage,
gastrointestinal perforation, and proteinuria. Unfortunately,
a negative study with bevacuzimab poses a challenge for
further development of agents modulating the VEGF signaling axis. Multitargeted tyrosine kinase inhibitors against
VEGFR, such as cediranib and sunitinib, have been evaluated in phase II and III clinical trials, respectively, but the
response rates have not been overwhelmingly encouraging.
In fact, the phase III sunitinib trial was recently terminated
for futility.
However, a novel VEGFR tyrosine kinase inhibitor that is
showing tremendous response rates in imaging studies,
including bone scans, is cabozantinib (XL184). Cabozantinib
is an inhibitor of multiple kinase signaling pathways, including MET, RET, VEGFR2/KDR, and KIT. MET is a
receptor tyrosine kinase that has roles in oncogenic signaling, angiogenesis, and metastasis. Androgen deprivation
activates MET signaling in prostate cancer cells. Activated

MET is particularly highly expressed in bone. Preclinical
studies have suggested that MET signaling may promote
survival of prostate cancer cells.20 In a recent phase II study,
cabozantinib showed promising activity in men with bone
metastases, with substantial improvement in bone scans in
most patients. Patients with metastatic CRPC with progressive measurable disease received cabozantinib at 100 mg/d
orally during 12 weeks, with a primary endpoint of objective
response rates. Accrual was halted at 168 patients based on
an observed high rate of clinical activity. Of the 100 evaluable patients with a median age of 68, 86% of patients had
complete or partial resolution of lesions on bone scan as
early as week 6. A total of 64% had improved pain, and the
most common related grade 3/4 adverse events were fatigue
(11%), hypertension (7%), and hand-foot syndrome (5%); no
related grade 5 adverse events were reported. The PSA
changes were independent of clinical activity, and the overall week 12 disease control rate was 71%.21 Current studies
are proceeding with a lower dose of cabozantinib to improve
drug tolerability (NCT01347788, NCT01428219). The relationship between c-met inhibition and metastatic bone disease has yet to be properly elucidated, but multiple research
efforts are under way to improve our understanding of this
novel finding.
Another antiangiogenesis inhibitor in a phase III clinical
trial is tasquinimod. Tasquinimod is an orally active
quinolone-3-carboxamide. Nearly 70% of men who took tasquinimod in a phase II trial did not progress at 6 months
compared with 30% of men who took placebo. The phase III
study is now open and expects to enroll 1200 patients
(NCT01234311).
Bone Targeting
Until recently, the only standard of care was to administer

vitamin D, calcium, and a bisphosphonate, such as zoledronic acid, to help minimize bone resorption, which leads to
reduction of SREs. Denosumab, a humanized monoclonal
antibody with specificity for the RANK ligand, was shown to
be superior to zoledronic acid in a study of 1,901 men with
CRPC. Denosumab delayed or prevented SREs more effectively than zoledronic acid.22 The median time to first
on-study SRE, the primary endpoint, was 20.7 months for
denosumab compared with 17.1 months for zoledronic acid
(HR, 0.82; 95% CI, 0.71 to 0.95; p 0.0002). No differences
were found in PSA time, overall disease progression, or
overall survival. The two treatment groups had a similar
frequency of serious toxicities; the cumulative incidence of
osteonecrosis of the jaw was similar in the two groups (2.3%
for denosumab vs. 1.3% for zoledronic acid). These results,
combined with two other pivotal phase III trials of the same,
led to the FDA approval of denosumab for prevention of
skeletal complications in patients with bone metastases
from solid tumors, except multiple myeloma.
Denosumab was also found to improve bone metastasis
free survival (29.5 months) compared with placebo in men
with M0 CRPC disease (25.2 months). However, no survival
benefits were seen with denosumab in the phase III trial
that was reported at the European Multidisciplinary Cancer
Congress and the European Society for Medical Oncology
Meeting 2011. In September 2011, denosumab received FDA
approval for its indication to increase bone mass in patients
at high risk for fracture receiving androgen deprivation
311
BAJAJ AND HEATH
therapy for nonmetastatic prostate cancer. Fortunately, for

men with CRPC, there are now two FDA-approved agents to
help improve and strengthen metastatic bones. Zoledronic
acid is administered via intravenous infusion and requires
monitoring of renal function. Denosumab is administered
via subcutaneous injection and requires monitoring of calcium and other electrolytes. Both agents have a role in the
treatment of prostate cancer.
Another agent that targets the osteoclasts, osteoblasts,
and bone microenvironment is dasatinib. Dasatinib is an
oral tyrosine kinase inhibitor with activity against Src
kinases. A phase I/II trial evaluating dasatinib alone and in
combination with docetaxel chemotherapy in CRPC reported
activity with bone turnover markers with mild PSA response.23 Currently, a phase III trial with an estimated
enrollment of 1,500 patients comparing docetaxel and dasatinib with docetaxel and placebo is under way
(NCT00744497).
A new drug that targets bone signaling pathways leading
to actual overall survival is radium-223 chloride. Radium223 chloride is an intravenously administered radiopharmaceutical that targets bone metastasis with high-energy,
short-range -particles. At the European Multidisciplinary
Cancer Congress and the European Society for Medical
Oncology Meeting 2011, results from a trial of nearly 1,000
patients with CRPC with 2:1 randomization to either
radium-223 chloride or placebo revealed a superior median
overall survival of 14 months in the radium-223 chloride
arm compared with the 11.2 months in the placebo arm.24
This 30% reduction in the risk of death (HR, 0.695, p
0.00185) in overall survival is important because the typical
treatment of bone targeting agents results in mediating
symptom relief and not necessarily affecting survival. The
study results were promising enough that the data safety
and monitoring committee for this study halted the accrual.
The adverse effects include hematologic effects, such as
anemia, and mild gastrointestinal toxicities. Next steps in
the development of radium-223 chloride are eagerly
awaited.
Cytotoxic Therapy
In 2004, our current stand-of-care chemotherapy, docetaxel, in combination with prednisone had shown efficacy
in patients with CRPC. An overall survival benefit of almost
t3 months and a PSA response of more than 50% were seen
in almost one-half of patients compared with mitoxantrone
and prednisone.9,25 In March 2010, the efficacy results of a
new taxane, cabazitaxel, for use in docetaxel-treated pa-
tients with CRPC were presented. This large, multicenter

study showed an overall survival benefit of 3 months when
compared with mitoxantrone.26,27 On the basis of these
results, the FDA-approved cabazitaxel as a second-line
therapy for prostate cancer. As expected, clinical trials
conducted in 2011 are primarily combination therapy evaluating cabazitaxel with chemotherapy, such as docetaxel
(NCT01308567), abiraterone (NCT01522536), and tasquinimod (NCT01513733). Clinical trials evaluating optimal sequencing of cabazitaxel and abiraterone and other future
agents must be efficiently designed to quickly meet its
objectives.
Conclusion
In the last several years, notable advances have been

made in the field of prostate cancer. Treatments emerging
from our knowledge of the cell biology, androgen regulation,
immunology, and chemoresistance of prostate cancer have
led to the development of mechanism-based drug discovery.
This in turn has led to various clinical trials based on a
sound biologic rationale. Several phase III trials testing
rational drug combinations in prostate cancer are ongoing.
However, clinical trials to determine optimal sequence of
therapy are yet to be conducted.
The role of immunotherapy, including the 2010 FDAapproved sipuleucel-T, is also an important part of the
treatment paradigm and one that would benefit from biomarker identification. Sipuleucel-T is the first therapeutic
cancer vaccine to gain FDA approval for patients with
metastatic CRPC. The landmark phase III study of
sipuleucel-T showed an overall survival benefit of 4.1
months; however, tumor response rates were minimal.28
The study validated the efficacy of immunotherapy in prostate cancer and has led to an investigation of additional
clinical trials of sipuleucel-T in prostate cancer. Additional
promising agents in phase III clinical trials include antiCTLA-4 (NCT01057810) and ProstVAC-V/F (NCT01322490).
Multiple new drugs have recently been approved for the
treatment of prostate cancer, but they all have not been able
to demonstrate cure of the disease. In addition to targeting
the different mechanisms of action discussed in this publication, multiple other molecular targets also promise to
provide the next generation of advances. We must continue
to maintain a close collaboration between basic and clinical
science so that our knowledge of the molecular physiology
can lead to strategic development of further viable drug
targets.
Author
Madhuri Bajaj*
Elisabeth I. Heath
312
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Amgen;
AstraZeneca;
BiPar Sciences;
Bristol-Myers
Squibb;
GlaxoSmithKline;
Pfizer; Seattle
Genetics;
Zymogenetics
Expert
Testimony
Other
Remuneration
DEVELOPMENTS IN PROSTATE CANCER THERAPY
REFERENCES
J Clin. 2012;62:10-29.
2. Chen Y, Sawyers CL, Scher HI. Targeting the androgen receptor
pathway in prostate cancer. Curr Opin Pharmacol. 2008;8:440-448.
3. Devlin HL, Mudryj M. Progression of prostate cancer: multiple pathways
to androgen independence. Cancer Lett. 2009;274:177-186.
4. Taplin ME, Regan MM, Ko YJ, et al. Phase II study of androgen
synthesis inhibition with ketoconazole, hydrocortisone, and dutasteride in
asymptomatic castration-resistant prostate cancer. Clin Cancer Res. 2009;15:
7099-7105.
5. Yap TA, Carden CP, Attard G, et al. Targeting CYP17: established and
novel approaches in prostate cancer. Curr Opin Pharmacol. 2008;8:449-457.
6. Danila DC, Morris MJ, de Bono JS, et al. Phase II multicenter study of
abiraterone acetate plus prednisone therapy in patients with docetaxeltreated castration-resistant prostate cancer. J Clin Oncol. 2010;28:14961501.
7. de Bono JS, Logothetis CJ, Molina A, et al. Abiraterone and increased
survival in metastatic prostate cancer. N Engl J Med. 2011;364:1995-2005.
8. Scher HI, Heller G, Molina A, et al. Evaluation of circulating tumor cell
(CTC) enumeration as an efficacy response biomarker of overall survival (OS)
in metastatic castration-resistant prostate cancer (mCRPC): Planned final
analysis (FA) of COU-AA-301, a randomized double-blind, placebo-controlled
phase III study of abiraterone acetate (AA) plus low-dose prednisone (P) post
docetaxel. J Clin Oncol. 2011;29 (suppl; abstr LBA4517)
9. Petrylak DP, Tangen CM, Hussain MH, et al. Docetaxel and estramustine compared with mitoxantrone and prednisone for advanced refractory
prostate cancer. N Engl J Med. 2004;351:1513-1520.
10. U.S. National Institutes of Health. Study Comparing Orteronel Plus
Prednisone in Patients With Chemotherapy-Naive Metastatic CastrationResistant Prostate Cancer [Clinical Trials.gov identifier NCT01193244].
http://clinicaltrials.gov/ct2/show/NCT01193244. Accessed March 20, 2012.
11. U.S. National Institutes of Health. Study Comparing Orteronel Plus
Prednisone in Patients With Metastatic Castration-Resistant Prostate Cancer [Clinical Trials.gov identifier NCT01193257]. http://clinicaltrials.gov/ct2/
show/NCT01193257. Accessed March 20, 2012.
12. Wu Y, Rosenberg JE, Taplin ME. Novel agents and new therapeutics in
castration-resistant prostate cancer. Curr Opin Oncol. 2011;23:290-296.
13. Tran C, Ouk S, Clegg NJ, et al. Development of a second-generation
antiandrogen for treatment of advanced prostate cancer. Science. 2009;324:
787-790.
14. Scher HI, Beer TM, Higano CS, et al. Antitumour activity of MDV3100
in castration-resistant prostate cancer: a phase 1-2 study. Lancet. 2010;375:
1437-1446.
15. Scher HI, Fizazi K, Saad F, et al. Effect of MDV3100, an androgen
receptor signaling inhibitor (ARSI), on overall survival in patients with
prostate cancer postdocetaxel: Results from the phase III AFFIRM study.
J Clin Oncol. 2012;30 (suppl 5; abstr LBA1).
16. U.S. National Institutes of Health. PREVAIL: A Multinational Phase 3
R, Double-Blind, Placebo-Controlled Efficacy and Safety Study of Oral

MDV3100 in Chemotherapy-Naive Patients with Progressive Metastatic
Prostate Cancer Who Have Failed Androgen Deprivation Therapy [Clinical
Trials.gov
identifier
NCT01212991].
clinicaltrials.gov/ct2/show/
NCT01212991. Accessed March 20, 2012.
17. Clegg NJ, Wongvipat J, Tran C, et al. ARN-509: a novel anti-androgen
for prostate cancer treatment. Cancer Res. Epub 2012 Jan 29.
18. Picus J, Halabi S, Kelly WK, et al. A phase 2 study of estramustine,
docetaxel, and bevacizumab in men with castrate-resistant prostate cancer:
results from Cancer and Leukemia Group B Study 90006. Cancer. 2011;117:
526-533.
19. Kelly WK, Halabi S, Carducci MA, et al. A randomized, double-blind,
placebo-controlled phase III trial comparing docetaxel, prednisone, and placebo with docetaxel, prednisone, and bevacizumab in men with metastatic
castration-resistant prostate cancer (mCRPC): Survival results of CALGB
90401. J Clin Oncol. 2010;28:18s (suppl; abstr LBA4511).
20. Zhang S, Zhau HE, Osunkoya AO, et al. Vascular endothelial growth
factor regulates myeloid cell leukemia-1 expression through neuropilin-1dependent activation of c-MET signaling in human prostate cancer cells. Mol
Cancer. 2010;9:9.
21. Hussain M, Smith MR, Sweeney C, et al. Cabozantinib (XL184) in
metastatic castration-resistant prostate cancer (mCRPC): results from a
phase II randomized discontinuation trial. J Clin. Oncol. 2011;29 (suppl;
abstr 4516).
22. Fizazi K, Carducci M, Smith M, et al. Denosumab versus zoledronic
acid for treatment of bone metastases in men with castration-resistant
prostate cancer: a randomised, double-blind study. Lancet. 2011;377:813-822.
23. Araujo JC, Mathew P, Armstrong AJ, et al. Dasatinib combined with
docetaxel for castration-resistant prostate cancer: results from a phase 1-2
study. Cancer. 2012;118:63-71.
24. Parker C, Heinrich, D, OSullivan JM., et al. Overall urvival benefit of
adium-223 chloride (Alpharadin) in the treatment of atients with Ssymptomatic bone metastases in castration-resistant prostate cancer (CRPC): a
Pphase III randomized trial (ALSYMPCA). In: European Multidisciplinary
Cancer Congress; 2011 September 24, 2011; Stockholm, Sweden; 2011. p.
1LBA.
25. Tannock IF, de Wit R, Berry WR, et al. Docetaxel plus prednisone or
mitoxantrone plus prednisone for advanced prostate cancer. N Engl J Med.
2004;351:1502-1512.
26. de Bono JS, Oudard S, Ozguroglu M, et al. Prednisone plus cabazitaxel
or mitoxantrone for metastatic castration-resistant prostate cancer progressing after docetaxel treatment: a randomised open-label trial. Lancet. 2010;
376:1147-1154.
27. Galsky MD, Dritselis A, Kirkpatrick P, et al. Cabazitaxel. Nat Rev Drug
Discov. 2010;9:677-678.
313
ADJUVANT THERAPY FOR OLDER PATIENTS

CHAIR
Arti Hurria, MD
City of Hope
Duarte, CA
SPEAKERS
Cynthia Owusu, MD, MS
Case Western Reserve University
Cleveland, OH
Nadine Jackson McCleary, MD
Boston, MA
Jeffrey Crawford, MD
Durham, NC
Adjuvant Therapy for Older Women with

Early-Stage Breast Cancer: Treatment
Selection in a Complex Population
By Cynthia Owusu, MD, MS, Arti Hurria, MD, and Hyman Muss, MD
Overview: Breast cancer is a disease of aging. However,

older women with breast cancer are less likely to participate in
clinical trials or to receive recommended treatment. This
undertreatment has contributed to a lag in breast cancer
survival outcomes for older women compared with that for
their younger counterparts. The principles that govern recommendations for adjuvant treatment of breast cancer are the
same for younger and older women. Systemic adjuvant treatment recommendations should be offered on the basis of
tumor characteristics that divide patients into three distinct
subgroups. These include (1) older women with hormone
receptor (HR)-positive and human epidermal growth factor 2
(HER2)-negative breast cancer who should be offered endocrine therapy; (2) older women with HR-negative and HER2negative breast cancer who should be offered adjuvant
REAST CANCER is the most common cancer in American women and the second leading cause of cancerrelated deaths among women. In 2011, approximately
230,480 new cases were diagnosed in the United States,
with an expected 39,520 deaths.1,2 The most important risk
factor for breast cancer is age. The estimated lifetime risk of
a new breast cancer is 1 in 15, 1 in 29, 1 in 27 and 1 in 207
for women 70 years or older, 60 to 69, 40 to 59, and 39 or
younger, respectively.1 The median age at the time of breast
cancer diagnosis is currently 61 years and an estimated 45%
of women are 65 or older at the time of initial diagnosis.2,3
Recent gains in life expectancy, coupled with aging as a risk
factor for breast cancer, makes breast cancer primarily a
disease of older women, with increasing public health importance. In 1980, persons 65 and older represented 11.3%
of the total population, but by 2030 this proportion is
expected to increase to 20%.3 In addition, by 2030, persons
older than 75 will be expected to account for just under 50%
of the total cohort older than 65.4 Given the nonlinear
age-risk relationship and increasing life expectancy, a substantial proportion of older women are expected to be affected by breast cancer.
Age-Related Cancer Health Disparities
Evaluation of the biology of breast cancer by patient age

has shown that hormone receptor-positive, low S-phase, low
tumor grade and HER2-negative tumors are more common
among older than younger women,5 although these differences are relatively modest. Despite the favorable tumor
profile of breast cancer in older women, this has not translated into any major survival advantage for older women
with breast cancer in comparison with their younger counterparts. In a study that drew data from National Vital
Statistics Reports and the Surveillance Epidemiology and
End Results database of the National Cancer Institute,
Smith and colleagues6 found that although the rate of breast
cancer death in the general population and the adjusted risk
of death among women with newly diagnosed disease are
declining among all age groups, the least decline has been
among older women. Relative to 1990, the rate of breast
chemotherapy; and (3) older women with HER2-positive disease who should be offered chemotherapy with trastuzumab.
Exceptions to these guidelines may be made for older women
with small node-negative tumors or frail older women with
limited life expectancy, where close surveillance may be a
reasonable alternative. Addressing the current age-related
disparities in breast cancer survival will require that older
women are offered the same state-of-the-art-treatment as
their younger counterparts, with a careful weighing of the
risks and benefits of each treatment in the context of the
individuals preferences. In addition, older women should be
encouraged to participate in breast cancer clinical trials to
generate additional chemotherapy efficacy, toxicity, and quality of life data.
cancer death in the general population decreased 2.5% per

year for women age 20 to 49, 2.1% per year for those age 50
to 64, 2% per year for those age 65 to 74, but 1.1% per year
for those age 75 years and older. Moreover, among women
with newly diagnosed breast cancer between 1980 and 1997,
the adjusted risk of death decreased by 3.6% per year among
women younger than age 75 compared with 1.3% per year
among those age 75 and older (p 0.01). These differences
were even greater for older black women. The age-related
disparity in survival outcomes was hypothesized to be related to the undertreatment of older women with breast
cancer. In an analysis of 9,766 patients enrolled in the
TEAM (Tamoxifen Exemestane Adjuvant Multinational)
randomized controlled trial conducted with postmenopausal
women with hormone receptor-positive breast cancer, increasing age was associated with a higher disease-specific
mortality.7 Treatments received and tumor characteristics
did not completely explain the age-related differences in
survival outcomes but older patients in this trial were much
less likely to receive chemotherapy. Together, these data
clearly underscore the fact that breast cancer is an important disease of older women who bear a disproportionate
burden of the morbidity and mortality associated with the
disease and who should be offered proven treatments that
improve survival outcomes. Given that treatment differences do not completely explain the age-related disparities
in survival outcomes, additional population and translational studies are needed to provide further insight into the
reasons for these disparities.
From the Case Western Reserve University School of Medicine, Cleveland, OH; City of
Hope Medical Center and Beckman Research Institute, Duarte, CA; and, Lineberger
Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC.
Address reprint requests to Cynthia Owusu, MD, MS, Case Western Reserve University
School of Medicine, Division of Hematology/Oncology and Case Comprehensive Cancer
Center-BHC 5055, 11100 Euclid Avenue, Cleveland, OH 44106; e-mail: cynthia.owusu@
case.edu.
1092-9118/10/1-10

Adjuvant Systemic Therapy
Adjuvant systemic therapy refers to the administration of

anticancer therapy after primary breast surgery for early
stage breast cancer with the goal of eradicating occult
micrometastatic disease thought to be responsible for distant recurrence. Systemic treatment modalities include
endocrine therapy and chemotherapy with or without trastuzumab. Treatment decision making regarding adjuvant
systemic therapy for older women should involve consideration of such factors as the risk of morbidity and mortality
from breast cancer, life expectancy and treatment tolerance,
and patient preference. A geriatric assessment that includes
evaluation of functional, cognitive, nutritional, and psychosocial status, and review of comorbidities and concomitant
medications is particularly helpful in estimating the health
status and life expectancy of older adults. Each of the
components of a geriatric assessment can identify older
adults at increased risk for morbidity and mortality. This
assessment can also identify areas of vulnerability that may
affect the patients ability to participate in a treatment plan
(for example, ability to take medications on ones own).
Items in a geriatric assessment are able to identify older
adults at risk of chemotherapy toxicity.8,9 In a multisite
study of 500 older adults with cancer, conducted by the
Cancer and Aging Research Group, five geriatric assessment
questions were independent predictors of the risk of chemotherapy toxicity, in addition to age, tumor, treatment, and
laboratory variables. The geriatric assessment questions
that were predictive of chemotherapy toxicity included hearing impairment (rated at fair or worse), difficulty in walking
one block, need for assistance with taking medications, one
or more falls in the past 6 months, and decrease in social
activities because of either physical or emotional health.8
This predictive model for chemotherapy toxicity is being
validated specifically in older adults receiving adjuvant
chemotherapy for breast cancer (clinicaltrials.gov
NCT01472094). In another study of 518 evaluable older
adults with cancer (Chemotherapy Risk Assessment Scale
for High Age Patients), predictors of chemotherapy toxicity
included measures of functional status (ability to complete
KEY POINTS
The undertreatment of older women with breast

cancer has contributed to poorer survival outcomes
for older women than for younger women.
The principles that guide breast cancer treatment
recommendations for younger and older women are
fundamentally the same.
Breast cancer treatment decision making should consider the risk of breast cancer relapse, patient health
status, life expectancy, and patient preference.
The use of systemic therapy in the adjuvant setting
should be based on tumor characteristics and include
endocrine therapy, and/or chemotherapy with or
without trastuzumab.
Omission of postoperative radiation therapy is a
reasonable strategy for older women with small
hormone-receptor positive tumors.
instrumental activities of daily living), cognition (MiniMental Status score), and nutrition (Mini-Nutritional Assessment score).9 This information can be used in the
treatment decision-making process in order to estimate life
expectancy and risk of chemotherapy side effects and to
identify areas of intervention to assist the patient during her
treatment course.
Endocrine Therapy
The majority of older patients present with hormone

receptorpositive breast cancer. Endocrine therapy is the
mainstay of adjuvant therapy for these patients and results
in proportional reductions in the risk of relapse and dying
of breast cancer that exceed any available chemotherapy
regimen. Current consensus guidelines recommend adjuvant systemic endocrine therapy for hormone receptorpositive breast cancer. The National Comprehensive Center
Network (NCCN) guidelines10 recommend the use of adjuvant endocrine therapy for women with hormone receptorpositive breast cancer regardless of age, menopausal status,
or HER2 status, with the possible exception of women
with lymph node-negative cancers 0.5 cm or less, or 0.6 to
1.0 cm in diameter with favorable prognostic features;
benefit from endocrine therapy is likely to be small for
tumors of this size. In contrast, the St. Gallen International
Consensus Panel recommends adjuvant systemic endocrine
therapy for all women with endocrine-responsive disease
with no exceptions and has defined endocrine-responsive
tumors as those with as few as 1% of cells staining positive
for hormone receptor proteins.11 The Society of International Geriatric-Oncology (SIOG) breast cancer task force
recommended that older women with endocrine-responsive
breast cancer be offered systemic endocrine therapy. However, for women with minimal risk disease, the decision to
offer endocrine therapy should be based on a risk-benefit
analysis.12
Tamoxifen is the most firmly established adjuvant endocrine therapy for both premenopausal and postmenopausal
women. Supporting these recommendations are results
from the Early Breast Cancer Trialists Collaborative
Group (EBCTCG) overview analysis, which demonstrated
that over a 15-year period, use of adjuvant tamoxifen
therapy for women with known estrogen receptor-positive
disease, compared with no tamoxifen, decreased the 15-year
risk of recurrence and death by 39% and 31%, respectively,
regardless of age.13 It is clear from these data that
tamoxifen is of benefit for older women. In addition to
decreasing the risk of disease relapse and death, there are
also potential nonbreast cancer benefits of tamoxifen therapy in postmenopausal women. Tamoxifen may prevent
osteoporosis14 and reduce the risk of cardiovascular disease.15 Adjuvant tamoxifen therapy is, however, underutilized in older women. Women 80 years or older are half as
likely to report having had a discussion about tamoxifen
with their doctor compared with women 65 to 79 years, and
women age 85 to 92 years are 25% less likely to receive a
tamoxifen prescription than those 80 to 84 years.16 Additionally, older women are more likely to self-discontinue and
to be nonadherent to tamoxifen before the recommended
treatment period of 5 years,17,18 undercutting the treatment
benefit from tamoxifen. Oncologists should always ask their
patients if they are taking their prescribed medications and
should reinforce the importance of adherence to maximize

treatment benefit.
The adjuvant use of aromatase inhibitors (anastrozole,
letrozole, exemestane) for postmenopausal women with
early breast cancer has been evaluated in several studies.
These studies have involved the use of aromatase inhibitors
either as initial adjuvant therapy,19 as sequential therapy
after 2 to 3 years of tamoxifen,20 or as extended therapy
after 4.5 to 6 years of tamoxifen.21 The findings of these
studies are consistent in demonstrating that the use of a
third-generation aromatase inhibitor for postmenopausal
women with hormone receptorpositive breast cancer, regardless of patient age, is superior in decreasing the risk of
disease recurrence, including ipsilateral and contralateral
breast cancer, and distant metastatic disease compared with
tamoxifen. Additionally, sequential use of aromatase inhibitors20 or extended therapy21 has been shown to provide an
overall survival advantage compared with tamoxifen use for
5 years. No survival advantage has been demonstrated with
the upfront use of aromatase inhibitors for 5 years compared
with tamoxifen.
There are differences in the toxicity profiles of aromatase
inhibitors and tamoxifen. The incidence of venous thromboembolic disease, cerebrovascular events, endometrial cancer, vaginal bleeding, and hot flashes are less likely to be
associated with aromatase inhibitors than tamoxifen,
whereas the incidence of musculoskeletal pain, osteoporosis,
and bone fractures have been found to be higher with
aromatase inhibitors.19 Emerging data also suggest that
aromatase inhibitors may be associated with a small but
higher risk of cardiovascular events compared with tamoxifen,22,23 but not compared with placebo.24 In a metaanalysis of seven trials in which aromatase inhibitors were
compared with tamoxifen, longer duration of aromatase
inhibitor use or aromatase inhibitor use alone for 5 years
was associated with a higher likelihood of cardiovascular
events compared with tamoxifen alone (odds ratio [OR]
1.26, 95% CI 1.10 1.43).25 Because studies of aromatase
inhibitors have not included extensive follow-up, the full
effect of aromatase inhibitors on cardiovascular disease and
coronary heart risk remains to be determined.
Although there is little evidence of age-related differences
in the benefits of aromatase inhibitors for postmenopausal
women, results of studies designed to examine age-related
differences in the pattern of toxicity have been mixed. In
general, the incidence of grade 35 nonfracture-related adverse events is higher among women 75 and older than
among women less than 75 years.26 However, a comparison
of the quality of life and the side effect profile for women who
participated in MA-17, a study in which 5 years of letrozole
was compared with placebo, showed no age-related differences in side effects.24 The long-term consequences and
implications of these side effects and any-age-related differences remain to be well characterized.
Based on the results from recent studies that favor aromatase inhibitors over tamoxifen, current guidelines recommend that aromatase inhibitors should be offered to all
postmenopausal women with hormone receptor-positive
early stage breast cancer, either alone, as sequential therapy after 23 years of tamoxifen. Given the lack of overall
survival advantage associated with aromatase inhibitor use
for 5 years, for women with pre-existing heart disease or
bone loss, use of tamoxifen for 5 years or a switching
strategy is a reasonable approach. For women with low

grade, node-negative tumors 1 cm or smaller, endocrine
therapy may be optional and observation acceptable, although the risks and benefits should be discussed with the
patient.
Adjuvant Chemotherapy
Cytotoxic chemotherapy can be considered for older

women with either node-positive or high-risk node-negative
disease, particularly, triple-negative breast cancer. An
abundance of literature has demonstrated the benefit of
adjuvant chemotherapy for younger women, with benefit
decreasing as age increases. The EBCTCG overview analysis,13 which has 15 years of follow-up data, demonstrated
that adjuvant chemotherapy reduced the annual risk of
recurrence by 37% and 19%, for women younger than 50 and
50 to 69, respectively. The annual risk of death was reduced
by 30% and 12%, for women younger than 50 and 50 to 69,
respectively. The benefit of adjuvant chemotherapy for
women with early stage breast cancer over age 70 was
difficult to assess in the EBCTCG overview analysis
because of the paucity of randomized trials that incorporated this age group. Of 29,000 women included in 60
adjuvant polychemotherapy trials, 4% were 70 and older.
This paucity of data has prevented definitive estimates
regarding the magnitude of benefit of chemotherapy for
women age 70 and older. In addition, with advancing age,
organ function and performance status decline, and comorbidities increase, making the risks associated with chemotherapy even greater. Moreover, the risk reductions for
chemotherapy are lower for postmenopausal women than for
premenopausal women, although the reasons are unclear.
Coupled with the apparent decline in the efficacy of chemotherapy with age is the increased risk of death from competing illnesses (comorbidity), leading to additional decline in
the benefit from chemotherapy. As a result of all these
factors, older women with early stage breast cancer receive
adjuvant chemotherapy considerably less frequently than do
younger women.
However, a growing body of evidence suggests that adjuvant chemotherapy leads to improved survival outcomes
for older women with breast cancer, particularly older
women with hormone receptor-negative and node-positive
breast cancer. A retrospective analysis of four Cancer and
Leukemia Group B (CALGB) randomized clinical trials
showed superior disease-free and overall survival benefits
with the use of more aggressive chemotherapy (compared
with less aggressive chemotherapy), among 6,487 women
with node-positive breast cancer in all age groups, including
70 and older.27 This benefit, however, came at the cost of
increased risk of toxicity in older women, with older women
more likely to discontinue treatment and to have an increased risk of treatment-related mortality. Additionally,
data from large population studies have demonstrated a
survival benefit from adjuvant chemotherapy for older
women with hormone-negative or node-positive early stage
breast cancer.28,29
Results from randomized controlled clinical trials that
have specifically focused on older women with breast cancer,
though scant, have helped to fill the gap on chemotherapy
benefit for older women with early stage breast cancer.
In the largest study in this population to date, a CALGB and
Breast Cancer Intergroup study, 33 patients 65 and older
with early-stage breast cancer were randomly assigned to

either standard treatment (doxorubicin and cyclophosphamide [AC] for four cycles or cyclophosphamide, methotrexate, and 5-fluorouracil [CMF] for six cycles) or to an
experimental arm of single-agent capecitabine for six cycles.30 At a median follow-up of 2.4 years, the relapse-free
survival for patients receiving single-agent capecitabine was
inferior to that for women receiving standard therapy (hazard ratio [HR] 2.09 (1.38 3.17) p 0 0.0001), as was
overall survival (HR 1.85 (1.113.08) p 0.02). An
unplanned subset analysis demonstrated that standard
combination chemotherapy was particularly effective in
patients with hormone receptor-negative disease. Overall,
these results demonstrate that standard combination chemotherapy in the adjuvant setting, provides an overall
survival benefit for older women with breast cancer, particularly those with hormone-receptor negative disease.
The optimum chemotherapy regimen for treating breast
cancer in the adjuvant setting remains debatable. Regardless, anthracycline-based regimens have become the norm,
particularly for high-risk disease. However, the long-term
complications associated with anthracycline use include,
but are not limited to, dose-dependent cardiomyopathy. Age
is a risk factor for cardiac disease, including anthracyclinerelated cardiomyopathy. Often, this precludes the use of
anthracycline-based regimens in older women with breast
cancer. Jones and colleagues,31 in a US Oncology trial,
compared an anthracycline-based regimen (AC for four cycles) with a nonanthracycline taxane-based regimen docetaxel and cyclophosphamide (TC) for four cycles in 1,016
women with node-negative and node-positive disease. At a
median follow-up of 7 years, TC use was associated with
superior disease-free survival (HR 0.74, 95% CI (0.56
0.98) p 0 0.03) and overall survival (HR 0.69, 95%
(0.50 0.97) p 0 0.03) compared with AC. Sixteen percent
of the study population were 65 and older. Unplanned
subgroup analysis showed that TC was associated with
superior disease-free and overall survival in all age groups,
including older ages. TC therefore is a reasonable treatment
regimen in the adjuvant setting for older women, particularly those with pre-existing heart disease and other contraindications to anthracycline-based regimens. A recent study
has shown that this regimen is well tolerated in older
women.32 In another EBCTCG overview analysis in which
different polychemotherapy regimens for early stage breast
cancer were compared, adding a taxane to an anthracyclinebased chemotherapy regimen or a higher cumulative-dosage
anthracycline-based regimen reduced breast cancer mortality, on average, by one-third. This benefit was irrespective
of node status, tumor size, tumor grade, or age (70).33
No definitive conclusion could be drawn regarding women
70 and older because few women in that age group were
included in the meta-analyses.
For older women with node-negative, hormone-positive
breast cancer, gene-expression profiling analysis can be used
to identify women with high-risk disease who are likely to
benefit from chemotherapy. The most widely used assay for
this purpose is the 21-gene assay, which quantifies the
likelihood of breast cancer recurrence in women with nodenegative, estrogen receptor-positive breast cancer and predicts the magnitude of endocrine therapy and chemotherapy
benefit.34 To the extent that the 21-gene assay allows for
individualization of cancer treatment, it is indeed a useful

tool for the management of older patients with breast
cancer, who were well represented in the validation cohorts
(NSABP-14 and NSABP-20) for the assay. Moreover, the
predictive ability of the 21-gene assay has been found to be
independent of age.34 Older patients with low scores are not
likely to derive substantial benefit from chemotherapy,
whereas those with high scores may derive great benefit.
The benefit of adjuvant chemotherapy among patients with
intermediate scores on the assay is being evaluated in the
Tailor Rx study.
Adjuvant! Online is another tool that can assist in decision
making regarding the benefits of adjuvant endocrine therapy and chemotherapy for an individual patient. This online
tool (available at www.adjuvantonline.com) summarizes the
absolute benefit of chemotherapy and endocrine therapy,
taking into account the patients age, brief assessment of
comorbid medical illnesses, and tumor characteristics.35 To
further inform this discussion, the risks associated with
chemotherapy can be calculated with the predictive models
for chemotherapy toxicity8,9 as described earlier (See Adjuvant Systemic Therapy).
Adjuvant Trastuzumab for HER2-Positive
Breast Cancer
Amplification or overexpression of HER2 is seen in approximately 10% to 15% of invasive breast cancers in older
women,5 and it is associated with an unfavorable prognosis.
A substantial body of literature from phase III trials in
the adjuvant setting has demonstrated considerable benefit
in disease-free and overall survival when trastuzumab is
used either concurrently or sequential to chemotherapy
compared with chemotherapy alone.36-38 The main adverse
effect associated with trastuzumab use is cardiotoxicity. In
five phase III trials of adjuvant trastuzumab, the incidence
of severe heart failure (New York Heart Association class III
or IV), ranged from 0 to 3.9% among patients receiving
trastuzumab, compared with 0 to 1.3% among patients who
did not receive trastuzumab.39 In the Breast Cancer International Research Group (BCIRG) 006 study,38 two
trastuzumab-containing regimens (AC plus docetaxel and
trastuzumab and a nonanthracycline regimen of docetaxel,
carboplatin, and trastuzumab [TCH]) were compared with
standard chemotherapy alone. This study demonstrated
disease-free and overall survival benefits with the use of
trastuzumab plus chemotherapy compared with chemotherapy alone. There was no substantial difference between the
two trastuzumab-containing arms. Moreover, the incidence
of cardiotoxicity associated with the nonanthracycline-based
trastuzumab regimen was lower than that associated with
the anthracycline-based trastuzumab regimen.
Consistent with the under-representation of older women
in breast cancer clinical trials of chemotherapy, older women
have also been under-represented in clinical trials of trastuzumab therapy. With the notable exception of cardiac
dysfunction, which was found to be associated with increasing age (older than 50), limitations in data collection precluded a determination of whether the toxicity profile of
younger patients. The reported clinical experience was also
not adequate to determine whether the efficacy improvements (overall and disease-free survival) associated with

Table 1. Summary of Recommendations for Adjuvant Systemic Therapy in Early Stage Breast Cancer in Older Women
Node-negative,
Tumor size 1 cm
Node-negative,
Tumor size 1 cm
Node-positive
Endocrine-positive,
HER2-negative
Consider hormonal therapy if tumor size 0.6 cm,
grade 2, or other high risk features
Hormonal therapy
Consider chemotherapy based on geneexpression profiling results
Hormonal therapy
Chemotherapy
Endocrine-negative,
HER2-negative
Consider chemotherapy if tumor size 0.6 cm plus
other high risk features
Chemotherapy alone
Chemotherapy alone
HER2-positive
Consider chemotherapy with trastuzumab if tumor
size 0.6 cm plus high risk features

Add hormonal therapy if hormone
positive

Add hormonal therapy
if hormone positive
patients younger than 65. Regardless, in the absence of

contraindications, trastuzumab is currently recommended
for the adjuvant treatment of HER2-positive breast cancer,
even for older women. In older women, a nonanthracycline,
trastuzumab-containing regimen of TCH is often used because of the increased risk of cardiotoxicity associated with
the anthracycline and trastuzumab regimen.
In summary, the principles that govern recommendations
for systemic adjuvant treatment of breast cancer are the
same for younger and older women. Older women should
be offered guideline-recommended therapies (Table 1).
Broadly, these recommendations should be offered along
three clinically distinct subgroups based on tumor characteristics. (1) Older women with hormone receptor-positive
and HER2-negative breast cancer who should be offered
endocrine therapy regardless of node status. Gene expression profiling assay may be used to determine the added
benefit of chemotherapy for those with node-negative
disease; (2) older women with hormone receptornegative
and HER2-negative breast cancer (triple-negative breast
cancer) who should be offered adjuvant chemotherapy; and
(3) older women with HER2-positive disease who should
be offered chemotherapy with trastuzumab. In the last
group, women with hormone receptor-positive tumors
should also be offered endocrine therapy. Exceptions to
these guidelines may be made for older women in any of the
three subgroups who have node-negative disease and a
tumor less than 1 cm or for frail older women with limited
life expectancy, where close surveillance may be a reasonable alternative.
Adjuvant Radiation Therapy
Until recently, the guideline recommendation, regardless

of age, was for all women to receive radiation therapy
after breast-conserving surgery and for postmastectomy
radiation to be offered to women with a high probability of
local recurrence. A recent meta-analysis of the EBCTCG
supports these recommendations, showing that radiation
therapy decreased the 10-year risk of any first recurrence
from 35% to 19% and the 15-year risk of breast cancer death
from 25% to 21% among women treated with breastconserving surgery.40 Although the proportional reductions
in relapse were similar among all women, the absolute
benefits varied substantially by age, grade, estrogen receptor status, tamoxifen use, and extent of surgery. The authors
concluded that radiation therapy after breast-conserving
surgery halves the rate at which the disease recurs and
reduces the breast cancer death rate by about a sixth.
However, older women with small tumors can be spared
radiation therapy. In a landmark randomized controlled

study by Hughes and colleagues41,42 636 women 70 or older
who had undergone lumpectomy for stage I hormone
receptor-positive breast cancer were randomly assigned to
receive either radiation therapy and adjuvant tamoxifen
for 5 years or to adjuvant tamoxifen for 5 years alone. The
results demonstrated no substantial differences between
the two groups with regard to mastectomy rates for local
recurrence, distant metastases, or overall survival (median
follow-up of 12 years). Of the 49% of patients who died
during follow-up, 3% died of breast cancer. The only statistically significant difference was found in the rate of local
or regional recurrence at 5 years, (2% among women who
had radiation therapy compared with 9% who did not).
Based on results from this study, one may reasonably
consider lumpectomy (with surgically clear margins)
without radiation therapy for women 70 and older with
clinically negative lymph nodes, a tumor 2 cm or smaller,
and hormone receptor-positive breast cancer who agree
to take endocrine therapy. This strategy is limited by the
high rate of nonadherence and early discontinuation of
adjuvant systemic endocrine therapy among older
women.43,44 Omission of postoperative radiation therapy,
coupled with nonadherence to adjuvant systemic endocrine
therapy, may result in earlier recurrences and, ultimately,
poorer survival outcomes for older women. A favorable
outcome from this approach can be achieved only when
older women are adherent to prescribed oral endocrine
therapies. Adherence to prescribed endocrine therapy
should therefore be discussed and encouraged at follow-up
visits.
Conclusion
Breast cancer is a disease of aging. With minor differences, existing data support similar recommendations for
both younger and older women. However, there are agerelated differences in treatment patterns, with older women
less likely than younger women to receive standard therapies. Furthermore, survival outcomes lag behind that of
younger women. Closing the current gap in age-related
disparities in breast cancer survival will require that
older women are offered the same state-of-the-art treatment
as younger women, with a careful weighing of the risks
and benefits of each treatment in the context of the individuals preferences. Newer tools that estimate life
expectancy and toxicity as well as the potential benefits of
therapy should make it easier for oncologists to make better
treatment decisions with older patients. In addition, older
women should be encouraged to participate in breast cancer
clinical trials to generate additional efficacy and toxicity

data. Such information will provide further knowledge so
that oncologists can offer older women the best treatment

options.
Author
Cynthia Owusu*
Arti Hurria
Hyman Muss
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Amgen;
Genentech; GTx
Research
Funding
Expert
Testimony
Other
Remuneration
Abraxis
BioScience;
Celgene;
GlaxoSmithKline
Boehringer
Ingelheim; Eisai;
Pfizer; Sandoz
REFERENCES
CA Cancer J Clin. 2011;61:212-236.
2. DeSantis C, Siegel R, Bandi P, et al. Breast cancer statistics, 2011. CA
Cancer J Clin. 2011;61:409-418.
3. SEER stat fact sheets. http://seer.cancer.gov/csr/1975_2008/results_
single/sect_01_table.11_2pgs.pdf. Accessed March 8, 2012.
4. Chu KC, Tarone RE, Kessler LG, et al. Recent trends in U.S. breast
cancer incidence, survival, and mortality rates. J Natl Cancer Inst. 1996;88:
1571-1579.
5. Diab SG, Elledge RM, Clark GM. Tumor characteristics and clinical
outcome of elderly women with breast cancer. J Natl Cancer Inst. 2000;92:
550-556.
6. Smith BD, Jiang J, McLaughlin SS, et al. Improvement in breast cancer
outcomes over time: are older women missing out? J Clin Oncol. 2011;29:
4647-4653.
7. van de Water W, Markopoulos C, van de Velde CJ, et al. Association
between age at diagnosis and disease-specific mortality among postmenopausal women with hormone receptor-positive breast cancer. JAMA. 2012;
307:590-597.
2011;29:3457-3465.
9. Extermann M, Boler I, Reich RR, et al. Predicting the risk of chemotherapy toxicity in older patients: The Chemotherapy Risk Assessment Scale for
High-Age Patients (CRASH) score. Cancer. Epub 2011 Nov 9. doi: 10.1002/
cncr.26646
in Breast Cancer. Version 1. 2012. www.nccn.org. Accessed February 28,
2012.
11. Goldhirsch A, Wood WC, Gelber RD, et al. Meeting highlights: updated
international expert consensus on the primary therapy of early breast cancer.
J Clin Oncol. 2003;21:3357-3365.
12. Wildiers H, Kunkler I, Biganzoli L, et al. Management of breast cancer
in elderly individuals: recommendations of the International Society of
Geriatric Oncology. Lancet Oncol. 2007;8:1101-1115.
13. Effects of chemotherapy and hormonal therapy for early breast cancer
on recurrence and 15-year survival: an overview of the randomised trials.
Lancet. 2005;365:1687-1717.
14. Love RR, Mazess RB, Barden HS, et al. Effects of tamoxifen on bone
mineral density in postmenopausal women with breast cancer. N Engl J Med.
1992;326:852-856.
15. Love RR, Wiebe DA, Feyzi JM, et al. Effects of tamoxifen on cardiovascular risk factors in postmenopausal women after 5 years of treatment. J Natl
Cancer Inst. 1994;86:1534-1539.
16. Blackman SB, Lash TL, Fink AK, et al. Advanced age and adjuvant
tamoxifen prescription in early-stage breast carcinoma patients. Cancer.
2002;95:2465-2472.
17. Partridge AH, Wang PS, Winer EP, et al. Nonadherence to adjuvant
tamoxifen therapy in women with primary breast cancer. J Clin Oncol.
2003;21:602 606.
18. Owusu C, Buist DS, Field T, et al. Tamoxifen discontinuance among
older women with estrogen-receptor-positive breast cancer. J Clin Oncol.
2006;24: (suppl; abstr 648).
19. Howell A, Cuzick J, Baum M, et al. Results of the ATAC (Arimidex,
Tamoxifen, Alone or in Combination) trial after completion of 5 years

adjuvant treatment for breast cancer. Lancet. 2005;365:60-62.
20. Coombes RC, Hall E, Gibson LJ, et al. A randomized trial of exemestane after two to three years of tamoxifen therapy in postmenopausal women
22. Mouridsen H, Keshaviah A, Coates AS, et al. Cardiovascular adverse
events during adjuvant endocrine therapy for early breast cancer using
letrozole or tamoxifen: safety analysis of BIG 1-98 trial. J Clin Oncol.
2007;25:5715-5722.
23. Buzdar A, Howell A, et al. Comprehensive side-effect profile of anastrozole and tamoxifen as adjuvant treatment for early-stage breast cancer:
long-term safety analysis of the ATAC trial. Lancet Oncol. 2006;7:633-643.
24. Muss HB, Tu D, Ingle JN, et al. Efficacy, toxicity, and quality of life in
older women with early-stage breast cancer treated with letrozole or placebo
after 5 years of tamoxifen: NCIC CTG intergroup trial MA.17. J Clin Oncol.
2008;26:1956-1964.
25. Amir E, Seruga B, Niraula S, et al. Toxicity of adjuvant endocrine
therapy in postmenopausal breast cancer patients: a systematic review and
meta-analysis. J Natl Cancer Inst. 2011;103:1299-1309.
26. Crivellari D, Sun Z, Coates AS, et al. Letrozole compared with tamoxifen for elderly patients with endocrine-responsive early breast cancer: the
BIG 1-98 trial. J Clin Oncol. 2008;26:1972-1979.
27. Muss HB, Woolf S, Berry D, et al. Adjuvant chemotherapy in older and
younger women with lymph node-positive breast cancer. JAMA. 2005;293:
1073-1081.
28. Du XL, Jones DV, Zhang D. Effectiveness of adjuvant chemotherapy for
node-positive operable breast cancer in older women. J Gerontol A Biol Sci
Med Sci. 2005;60:1137-1144.
29. Elkin EB, Hurria A, Mitra N, et al. Adjuvant chemotherapy and
survival in older women with hormone receptor-negative breast cancer:
assessing outcome in a population-based, observational cohort. J Clin Oncol.
2006;24:2757-2764.
30. Muss HB, Berry DA, Cirrincione CT, et al. Adjuvant chemotherapy in
older women with early-stage breast cancer. N Engl J Med. 2009;360:20552065.
31. Jones S, Holmes FA, OShaughnessy J, et al. Docetaxel with cyclophosphamide is associated with an overall survival benefit compared with doxorubicin and cyclophosphamide: 7-year follow-up of US Oncology Research
32. Freyer G, Campone M, Peron J, et al. Adjuvant docetaxel/cyclophosphamide in breast cancer patients over the age of 70: results of an observational
study. Crit Rev Oncol Hematol. 2011;80:466-473. Epub 2011 May 11.
33. Peto R, Davies C, et al. Comparisons between different polychemotherapy regimens for early breast cancer: meta-analyses of long-term
outcome among 100,000 women in 123 randomised trials. Lancet. 2012;379:
432-444.
34. Paik S, Shak S, Tang G, et al. A multigene assay to predict recurrence
of tamoxifen-treated, node-negative breast cancer. N Engl J Med. 2004;351:
2817-2826.
35. Ravdin PM, Siminoff LA, Davis GJ, et al. Computer program to assist
in making decisions about adjuvant therapy for women with early breast

2005;353:1659-1672.
2005;353:1673-1684.
38. Slamon D, Eiermann W, Robert N, et al. Adjuvant trastuzumab in
HER2-positive breast cancer. N Engl J Med. 2011;365:1273-1283.
39. Ewer SM, Ewer MS. Cardiotoxicity profile of trastuzumab. Drug Saf.
2008;31:459-467.
40. Darby S, McGale P, et al. Effect of radiotherapy after breast-conserving
surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis
of individual patient data for 10,801 women in 17 randomised trials. Lancet.
2011;378:1707-1716.
42. Hughes KS, Schnaper LA, Cirrincione C, et al. Lumpectomy plus
tamoxifen with or without irradiation in women age 70 or older with early
breast cancer. J Clin Oncol. 2010;28:15s (suppl; abstr 507).
43. Owusu C, Buist DS, Field TS, et al. Predictors of tamoxifen discontinuation among older women with estrogen receptor-positive breast cancer.
J Clin Oncol. 2008;26:549-555.
44. Partridge AH, LaFountain A, Mayer E, et al. Adherence to initial
adjuvant anastrozole therapy among women with early-stage breast cancer.
J Clin Oncol. 2008;26:556-562.
Adjuvant Treatment of Older Patients with

Lung Cancer
By Jeffrey Crawford, MD
Overview: Although advances in the molecular biology of

lung cancer have rapidly impacted management of patients
with advanced stage non-small cell lung cancer (NSCLC), the
principal treatment in the adjuvant setting of early stage
NSCLC remains platinum-based chemotherapy regimens. The
evidence available from clinical trials demonstrates a similar
benefit of adjuvant chemotherapy in fit, older patients as well
as younger patients. Observational studies suggest that adjuvant chemotherapy for older patients provides comparable
survival benefit, along with increased toxicity. The lower use
UNG CANCER is the leading cause of cancer mortality

worldwide. Largely because of the cumulative effects of
smoking and other exposures, the risk of lung cancer increases with age. In the United States, Surveillance, Epidemiology, and End Results (SEER) data demonstrate that
68% of patients diagnosed with NSCLC are age 65 or older.1
The median age at diagnosis is 71. The incidence of lung
cancer peaks in the age 65 to 74 group with an incidence of
31.9%. However, the rate of lung cancer in the age 75 to 84
group is nearly as high at 28.9%. At age 85 and older, the
incidence drops to 7.3%. Whereas this decline in the very old
may partially represent under diagnosis, it also speaks to
the complex interaction between smoking, cumulative life
exposure, and competing risks for mortality, which reduces
the likelihood of long-term survival in smokers and former
smokers.
Increasing age is also associated with increasing likelihood of comorbidities, which may limit treatment options for
patients. The intersection of age-related functional impairment and comorbidities affect the patients overall performance status, which has been a powerful predictor of
outcomes for patients with lung cancer ever since Karnofskys initial description.2 For these reasons, older patients
may be less likely to receive surgery and adjuvant treatment than their younger counterparts. When this decision is
made on the basis of significant functional impairment that
substantially alters the risk/benefit ratio for therapeutic
intervention, this may be quite appropriate. However, withholding treatment based on age alone is not appropriate
given the substantial body of data suggesting that the fit
elderly population can have similar benefits compared to
younger patients.
Further complicating this field is the limited evidence
specific to the elderly population, since the vast majority of
patients enrolled on clinical trials in lung cancer are
younger than age 70.3 Therefore, the clinician is often left
with uncertainty when evaluating the true risk/benefit of
treatment in a 75- or 80-year-old individual with or without
substantial comorbidities.
of chemotherapy in the older population also suggests that the

selection of appropriate patients remains an important part of
the decision process. Carboplatin therapy may be substituted
for cisplatin in selected older patients, and different options
exist for the second cytotoxic chemotherapy agent. As in all
patients, and particularly in this vulnerable population receiving cytotoxic chemotherapy, supportive care is vital. Increasing enrollment of the older population in clinical trials will be
important to improve the evidence for our decision-making in
the future.
function decreases with age, including a decrease in elasticity of the arterial system, loss of myocytes, loss of atrial
pacemaker cells, and increased fibrosis of the cardiac fibroskeleton.4 Thus, even the patient with no prior cardiac
history, an increase in arrhythmias and other cardiac events
may occur. Secondly, the age-related decline in renal function occurs with a decline in renal flow, decreased glomerular filtration rate, and creatinine clearance.5
Both renal and cardiac function issues have a substantial
impact on subsequent decisions regarding drug administration, such as cisplatin-based chemotherapy and related
hydration, and also significant considerations for anesthesia
and other perioperative medications. The decline in hematopoietic reserve may be associated with increased toxicity of
chemotherapy but also may substantially affect the need for
hematopoietic support during and following surgical interventions.6 Particularly important when considering lung
cancer surgery is the decline in lung function that occurs as
a function of age.7 For subjects who have never smoked,
there is a gradual reduction in lung function, with a decline
in FEV1 of 25% between ages 50 and 75, with further decline
over age 75. However, for those who have smoked the
majority of their adult lives, there may be a more substantial
reduction in lung function by as much as 75% or greater.
With emphysema and other smoking-related lung diseases,
operative and postoperative risks and long-term quality of
life may be significantly compromised and/or the patient
may not be a candidate at all, despite having early-stage
disease. Fortunately, for former smokers who quit smoking
before significant lung injury occurs, the subsequent loss of
lung function is more consistent with the gradual agerelated decline seen in the nonsmoking population.
Because of improvements in surgical techniques, including less invasive surgery such as video-assisted thoroscopic
approaches, better preoperative evaluation, and improved
perioperative management of patients, defining which patients with early-stage lung cancer are operable has been
evolving. Studies over the last decade have suggested improvement in outcomes for older patients.8
Surgery in the Older Patient with Lung Cancer
Traditionally, lung cancer surgery has been reserved for

the healthy elderly population. Even in this population, the
risk of operative and postoperative complications is increased, related, at least in part, to the decline in organ
function with age. It is well known that cardiovascular
From the Duke University Medical Center, Durham, NC.

Address reprint requests to Jeffrey Crawford, MD, Duke University Medical Center,
2301 Erwin Rd., Durham, NC 27710; email: Crawf006@mc.duke.edu.
1092-9118/10/1-10
315
JEFFREY CRAWFORD
Table 1. Surgical Management and Outcomes in Elderly
Patients with Early Stage NSCLC
Lobectomy
Pneumonectomy
30 d mortality
90 d mortality
age 70
age 70
n 1,969
77%
11%
2.2%
2.5%
n 1,969
79%
8%
3.6%
4.7%
NS
NS
p 0.01
p 0.0002
More recent studies provide a current idea of the risks and

outcomes for older patients with early stage NSCLC. A
nested case-control study in France between January 2004
and December 2008 matched nearly 2,000 patients age 70 or
older with stage I/II NSCLC with matched controls younger
than age 70 in order to compare surgical treatment and
postoperative outcomes.9 These results are outlined in Table
1. Although the frequency of radical lymph-node dissection
was more common in the younger patient, there was otherwise no significant difference in the type of resection between younger and older patients. The frequency of
lobectomy was 77% in the younger patient and 79% in the
older patients, with pneumonectomy in 11% of the younger
patients and 8% in the older patients. Sublobar resections
occurred in only 7% of the patients younger than age 70 and
6% over the age of 70. Postoperative mortality was higher in
the older patient with a 30-day mortality of 3.6% compared
with 2.2% in the younger patients (p 0.01). For the older
population, perhaps an even more important endpoint is the
90-day mortality. This endpoint was 4.7% in those patients
older than age 70 compared with 2.5% in the population
younger than age 70 (p 0.0002). This result in particular
shows the vulnerability of this population after thoracic
surgery. However, the relatively low mortality in this age
group across a wide number of institutions in France reassures us that patients appropriately selected for surgery can
have similar outcomes.
Improved outcomes with higher surgery rates for older
patients with early-stage NSCLC have been recently reported from a national cohort study of over 17,000 Medicare
KEY POINTS
316
The majority of lung cancer occurs in the older

population.
Advances in staging, surgery, and perioperative management have increased the number of older patients
who are candidates for surgical resection for earlystage lung cancer.
Although the elderly population is clinically underrepresented in clinical trials, the data suggest that
older, fit patients may have similar benefit with
adjuvant chemotherapy than younger patients with
a modest increase in toxicities.
Observational databases suggest that these benefits
extend to the patients outside of clinical trials. In the
United States, carboplatin is commonly substituted
for cisplatin in the older patient population and
benefits may be comparable within the limits of this
data.
patients with stage I or II NSCLC, identified through the

SEER database between 2001 and 2005.10 In this study,
areas of high and low rates of curative surgery for earlystage lung cancer were compared in an attempt to determine
the effectiveness of surgery in older and sicker patients. In
the low surgery areas, less than 63% of patients underwent
surgery compared with greater than 79% in high surgery
areas. Those high surgery areas also operated on more
patients with advanced age and patients with chronic obstructive pulmonary disease. Despite this, the overall 1-year
mortality was 18% in the high surgery area compared with
22.8% in the low surgery area (adjusted odds ratio [OR] 0.89;
95% confidence interval [CI] 0.86 0.93). In addition, the
1-year lung cancer specific mortality was lower in the high
surgery area at 12% compared with 16.9% for the low
surgery area (OR 0.86; 95% CI 0.82 0.91). The training
effect and expertise of the areas with higher rates of surgery
was associated with older and sicker patients undergoing
resection with improved survival. This speaks strongly to
the need for identifying broader expertise and centers of
excellence for surgical intervention in older patients, both in
the academic and community settings.
Adjuvant Chemotherapy Clinical Trials
The 5-year survival rate for surgical resection alone in

NSCLC is approximately 55% to 65% for patients with stage
I disease, 40% to 55% for patients with stage II, and 20% to
25% for stage III. Based on the Mountain staging system,
which has served as the basis for the majority of adjuvant
chemotherapy trial results currently available,11 the Lung
Adjuvant Cisplatin Evaluation (LACE) pooled analysis has
reported survival improvement following surgical resection
with the use of adjuvant chemotherapy between 4% and
15%.12 Among these trials, JBR.10 was a study conducted by
the National Cancer Institute of Canada, comparing cisplatin and vinorelbine with observation. A subset analysis has
been reported for the patients older than age 65 in this
trial.13 Three hundred twenty-seven patients age 65 or
younger were compared to 155 patients older than age 65.
The baseline demographics were similar for the two groups,
with the exception of histology with 58% of younger patients
having adenocarcinoma compared with 43% of the older
patients. Squamous cell cancer was seen in 32% of the
younger patients and 49% of the older patients (p 0.001).
Performance status 0 was also more frequent in the younger
population53% compared with 41% (p 0.01). Chemotherapy significantly prolonged overall survival in the older
population with a hazard ratio of 0.61 (95% CI, 0.38 0.98).
This benefit was quite similar to the effect seen for the
overall study population. Differences were seen in the mean
dose intensity of vinorelbine and cisplatin, with the elderly
patients receiving fewer doses of both agents. Fewer elderly
patients completed treatment and more patients refused
treatment. There were no significant differences in toxicities, hospitalizations, or treatment-related deaths by age
group. There was a higher mortality rate from nonmalignant causes in the older population21.1% compared with
11.9% in the younger population, although this did not reach
statistical significance. The authors concluded that, despite
the fact that the older population received less chemotherapy overall, the improvement in survival of adjuvant vinorelbine and cisplatin was comparable to the result seen in
the younger population and also had acceptable toxicity.
OLDER PATIENTS WITH LUNG CANCER
Fig. 1. Analysis by age of survival comparing surgery with chemotherapy to surgery alone (NSCLC) meta-analysis.
A meta-analysis has also been performed of individual

patient data comparing the effects of adding adjuvant chemotherapy to surgery through the NSCLC Meta-analyses
Collaborative Group.14 The first meta-analysis of surgery
plus chemotherapy compared with surgery alone includes 34
trials in 8,447 patients. The overall benefit of adding chemotherapy after surgery was an absolute increase in survival of 4% at 5 years with a hazard ratio of 0.86 (95% CI
0.81 0.92; p 0.0001). Figure 1 demonstrates the hazard
ratio from that study. As shown, in all age subgroups, the
addition of chemotherapy to surgery improved survival, with
a trend for the greatest effect in the group older than age 70.
Although this is clearly a subset analysis and must be taken
with caution, it at least suggests that adjuvant chemotherapy in older patients, who are appropriately selected, as they
would have been in these clinical trials, can provide survival
benefit at least comparable to that of younger patients.
Observational Studies of Adjuvant Chemotherapy in
the Older Population
How do these results from randomized clinical trials and

meta-analyses of these studies apply to clinical practice? An
observational cohort study utilizing the SEER registry and
Medicare database has been reported to address this question.15 In this study, 3,324 patients older than age 65 were
identified as having surgery for stage II and IIIA NSCLC.
This included cases of lung cancer diagnosed up to 2005 with
follow-up data through December of 2007. The primary
endpoint was to look at overall survival. In this group of
patients, 21% received platinum-based chemotherapy.
There was improvement in overall survival for patients who
received chemotherapy, with a hazard ratio of 0.78. Beneficial results were seen in both stage II and stage IIIA
patients. Within age strata, improved survival was seen in
the population younger than age 70 (HR 0.74, CI 0.62 0.88).
There was also improvement in overall survival for the
population age 70 to 79 (HR 0.82, CI 0.71 0.94). However,
no survival benefit was observed in the population older
than age 80 (HR 1.33, CI 0.86 2.06). The use of adjuvant
chemotherapy was associated with an increased odds ratio
of serious adverse events as determined by hospitalization
(OR 2.0, CI 1.52.6). Given the nature of this study, there
is no comparison population younger than age 65 and details
of exact toxicities are limited. This study extends the observations from the clinical trial setting to clinical practice
regarding the potential benefit of adjuvant chemotherapy in
the elderly. Appropriate caveats also include that the benefit
of adjuvant chemotherapy is clearly not established in the
population older than age 80 and the effect of adverse events
must be also considered. Overall, there was 3.1% mortality
within 12 weeks of treatment for this population.
At the American Society of Clinical Oncology Annual
Meeting in 2011 (ASCO 2011), Cuffe and colleagues reported
on the patterns of use of adjuvant chemotherapy among
surgically resected patients with NSCLC in Ontario, with
the focus on the population of patients age 70 and older.16
Although this represents more than 50% of the patients with
lung cancer, in JBR.10, only 15% of patients were older than
age 70, and the overall LACE analysis included only 9% of
patients in this age group. This study evaluated the use of
adjuvant chemotherapy and associated outcomes from 2001
to 2003, before results of JBR.10 and other clinical trials
demonstrating the benefit of adjuvant chemotherapy were
known. This preadoption time period was compared to the
postadoption time period of 2004 to 2006. The primary
study outcome was overall survival, with a secondary endpoint of rate of hospitalization within six months of surgery
as a surrogate for toxicity. In this study, 6,570 patients were
identified who underwent surgical resection within 24 weeks
of diagnosis. Patients who received neoadjuvant radiation
and/or chemotherapy were excluded, leaving a population of
6,304 patients. In this group, 3,541 patients were younger
than age 70 with 1,217 patients age 70 to 74, 980 patients
age 75 to 79, and 466 patients older than age 80. Other
variables associated with survival differences included age
and Charlson comorbidity scores. The majority of chemotherapy was delivered in the group younger than age 70.
Very few patients older than age 80 were treated with
chemotherapy. Overall survival by age group clearly favored
the population younger than age 70, with the worse survival
rate in the population older than age 80. However, when the
age groups were compared between the pre- and postadoption time periods, improvement in survival was seen in both
the population younger than age 70 and older than age 70,
with hazard ratios of 0.85 and 0.87 respectively. By comparison, no difference was seen in the population older than age
80, with a hazard ratio of 1.0. Toxicities, defined by hospitalization within 6 weeks of surgery, were higher in the
population older than age 75 (p 0.001) and ranged between 11% and 18%, reflecting postoperative complications.
By contrast, hospitalization rates between 6 and 24 weeks of
surgery, when a patient would have received chemotherapy,
were similar across all age groups and varied between 27%
and 32%. The authors suggest that there is an association
between the adoption of adjuvant chemotherapy in the older
population and the survival improvement, although the
majority of patients did not receive chemotherapy. The
benefit of adjuvant chemotherapy in patients older than age
80 was not clarified by this study, since so few patients
received treatment.
An additional study further explored the SEER Medicare
database to form a comparison between carboplatin- and
cisplatin-based regimens.17 Although cisplatin-based chemotherapy has been the standard recommended therapy in
the adjuvant setting, the older patient population may have
poor tolerance for this medication or be unable to receive it
317
JEFFREY CRAWFORD
because of organ dysfunction, neuropathy, or other comorbidities. One prospective trial, CALGB 9633, was performed
utilizing carboplatin and paclitaxel in patients with stage IB
disease.18 However, this trial was underpowered and only
showed a trend toward overall survival. Because of this, the
effectiveness of carboplatin-based adjuvant chemotherapy
has been questioned. In this particular study, the authors
looked at 3,324 patients age 65 and older from the SEER
Medicare database with resected stage II and IIIA NSCLC
diagnosed between 1992 and 2005.15 In this study, 636
patients, or 19%, received platinum-based chemotherapy
within 3 months of surgery. The overall population had a
significant improvement in survival with a hazard ratio of
0.79 (95% CI, 0.71 0.89). Among these patients, 105, or
16.5%, received cisplatin-based therapy, and 489 patients,
or 76.9%, received carboplatin-based therapy. The hazard
ratio for cisplatin-based chemotherapy was 0.76 (95% CI,
0.60 0.96). The hazard ratio for carboplatin-based chemotherapy compared with no adjuvant chemotherapy was 0.76
(95%, CI 0.68 0.86). When cisplatin- and carboplatin-based
regimens were compared directly, there appeared to be no
difference in survival with a hazard ratio for carboplatin of
0.91 (95% CI, 0.70 1.18). Chemotherapy-related toxicities
were similar between the two platinum-based regimens,
except for a lower rate of infection and emesis for the
carboplatin-treated patient and borderline reduction in dehydration, also favoring carboplatin. This study documented
the common practice of substituting carboplatin for cisplatin
in the older patient receiving adjuvant chemotherapy. The
same caveat exists that the healthier patients with better
outcomes may have potentially received chemotherapy.
With that said, the survival appeared similar in those
patients treated with cisplatin- or carboplatin-based therapy. In this regard, the use of a comprehensive geratric
assessment can be helpful in determining both life expectancy and morbidity from treatment of the older patient
with cancer.22 These tools are being incorporated into prospective clinical trials of older patients with lung cancer and
need to be studied in the surgical population as well so we
can better inform our patients of the risks and benefits of our
treatments.
Current Status
Although the majority of clinical trial data involves

cisplatin-based chemotherapy with vinorelbine, as well as
older regimens such as cisplatin/etoposide, the comparability of cisplatin-based regimens in advanced stage disease
has led to the conclusion that multiple options for platinumbased chemotherapy can be considered in the adjuvant
setting. The current intergroup study, E1505, allows the
physician a choice of cisplatin-based regimens, including
vinorelbine, docetaxel, gemcitabine, or pemetrexed.19 In
particular, the cisplatin/pemetrexed regimen was added
most recently to this study, based on its efficacy and potential lower toxicity in advanced stage patients with nonsquamous NSCLC.20 A phase II comparison between adjuvant
cisplatin/vinorelbine and cisplatin/pemetrexed was reported
at ASCO 2011.21 The aim of this phase II trial was to
compare dose delivery and clinical feasibility of cisplatin/
pemetrexed compared with cisplatin/vinorelbine in the adjuvant setting. This study was done predominantly in a
younger population with a mean age of 59, with 132 patients
randomly selected. However, delivery of the intended dose of
318
cisplatin/pemetrexed was higher at 74.6%, compared with

20% for the cisplatin/vinorelbine. The median number of
cycles of treatment was four of cisplatin/pemetrexed compared with three for the cisplatin/vinorelbine arm. The time
to withdrawal from therapy differed significantly, favoring
the cisplatin/pemetrexed arm (p 0.001). In early follow-up
at 4 months, there had been two deaths on the cisplatin/
vinorelbine arm and one on the cisplatin/pemetrexed arm.
Although this study has a very small number of patients, it
does help support the option of cisplatin/pemetrexed in the
current ECOG 1505 study that hopefully will lead to more
prospective data in the older population.
In this regard, it is important that we support enrollment
of patients on ECOG 1505 across all eligible age populations.
There is currently no large, randomized age-specific adjuvant trial to provide the robust database we need to better
inform our patients and ourselves of the risk/benefit ratio of
treatment of this population. However, by enrolling older
eligible patients in the current adjuvant trial, we will have
the opportunity to develop a significant database of toxicities
and long-term outcomes in the older patients compared with
younger patients. An interim report on the demographics
and toxicities of E1505 was reported at ASCO 2011.19
Although the median age is 61 to 62, it is notable that a
substantial percentage of patients were in their seventies,
with patients enrolled up to ages 84 and 86. This trial is
comparing standard cisplatin chemotherapy doublets with
or without the addition of 1 year of bevazucimab therapy.
Differences in the bevazucimab arm in terms of toxicity
include higher rates of neutropenia, as well as hypertension
and proteinuria. It is also important to recognize that across
all age groups, grade 5 toxicities of these regimens occurred
in 2.2% of the control arm and 3.3% in the bevazucimab arm.
This speaks to the importance of appropriate patient selection, careful follow-up, and management with state-of-theart supportive care to minimize deaths from toxicity in this
potentially curative setting.
Compared with breast cancer, colon cancer, and other
areas where adjuvant chemotherapy has been a standard
for decades, the incorporation of adjuvant chemotherapy is
relatively recent in the therapeutic history of NSCLC. The
toxicities of our current cytotoxic regimens, along with the
vulnerability of our older patientswith and without comorbid diseaseneed better definition than is currently available. Subset analyses of randomized clinical trials and
meta-analyses, as well as observational studies, do provide
direction for the practicing oncologist regarding older patients and the treatment options in the setting of early stage
NSCLC. Exciting advances with targeted therapeutic approaches based on the molecular biology of the cancer have
occurred in patients with advanced stage lung cancer. These
same approaches clearly need to be studied prospectively in
patients with early stage lung cancer across all age groups.
The potential benefit of reduced toxicity of these agents may
be particularly important in the older population. In the
meantime, the principles learned from other populations
treated with adjuvant chemotherapy should be heeded.
There is no evidence currently for nonplatinum-based chemotherapy in the adjuvant setting, although the substitution of carboplatin for cisplatin may be an appropriate option
for some older patients. Although evidence supports the use
of adjuvant chemotherapy in appropriate patients older
than age 70, there is insufficient evidence to recommend
OLDER PATIENTS WITH LUNG CANCER
adjuvant chemotherapy at this point to patients age 80 and

older. The potential importance of delivering standard-dose
therapy compared with potential myelotoxicity and compli-
cations in the older postoperative patient with comorbid

disease demands thoughtful and proactive patient management and supportive care.
Author
Jeffrey Crawford
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Agennix; Amgen;
Genentech
Research
Funding
Expert
Testimony
Other
Remuneration
Agennix; Amgen
REFERENCES
1. Ries LAG, Melbert D, Krapcho M, et al. (eds.). SEER Cancer Statistics
Review, 19752005. Bethesda, MD: National Cancer Institute; 2005.
2. Karnofsky D, Burchenal J. The Clinical Evaluation of Chemotherapeutic
Agents in Cancer. In MacLeod CM (ed). Evaluation of Chemotherapeutic
Agents. New York, NY: Columbia University Press, 1949; 196.
3. Weiss J, Langer C. NSCLC in the elderlythe legacy of therapeutic
neglect. Current Treatment Options in Oncology. 2009;10:180-194.
4. Cheitlin M. Cardiovascular physiology-changes with aging. Am J Geriatr
Cardiol. 2003;12:9-13.
5. Muhlberg W, Platt D. Age-dependent changes of the kidneys: Pharmacological implications. Gerontology. 1999; 45:243-253.
6. Dees E, OReilly SN, Goodman S, et al. A prospective pharmacologic
evaluation of age-related toxicity of adjuvant chemotherapy in women with
breast cancer. Cancer Invest. 2000;18:521-529.
7. Fletcher C, Peto R. The natural history of chronic airflow obstruction.
BMJ. 1977;1:1645-1648.
8. Rajdev L, Keller S. Surgery for Lung Cancer in Elderly Patients. In
Govindan R (ed). ASCO Educational Book. Alexandria, VA: American Society
of Clinical Oncology, 2006; 463-467.
9. Rivera C, Falcoz PE, Bernard A, et al. Surgical management and
outcomes of elderly patients with early stage non-small cell lung cancer: a
nested case-control study. Chest. 2011;140:874-880.
10. Gray S, Landrum M, Lamont E. Improved outcomes associated with
higher surgery rates for older patients with early stage non-small cell lung
cancer. Cancer. Epub 2011 Jul 28.
11. Mountain CF. Revisions in the international system for staging lung
cancer. Chest. 1997;111:1710-1717.
12. Pignon J, Tribodet H, Scagliotti G, et al. Lung adjuvant cisplatin
evaluation: A pooled analysis by the LQCE Collaborative Group. J Clin Oncol
2008;26:3552-3559.
13. Pepe C, Hasan B, Winton T, et al. Adjuvant vinorelbine and cisplatin in
elderly patients: National Cancer Institute of Canada and Intergroup Study
JBR. 10. J Clin Oncol. 2007;25:1553-1561.
14. NSCLC Meta-analyses Collaborative Group, Arriagada R, Auperin A,

et al. Adjuvant chemotherapy, with or without postoperative radiotherapy, in
operable non-small-cell lung cancer: two meta-analyses of individual patient
data. Lancet. 2010;375:1267-1277.
15. Wisnivesky J, Smith C, Packer S, et al. Survival and risk of adverse
events in older patients receiving postoperative adjuvant chemotherapy for
resected stages II-IIIA lung cancer: Observational cohort study. BMJ. 2011;
343:d4013.
16. Cuffe S, Booth C, Peng Y, et al. Adoption of adjuvant chemotherapy
(ACT) for non-small cell lung cancer (NSCLC) in the elderly: A populationbased outcomes study. J Clin Oncol. 2011;29(suppl; abstr 7012)
17. Gu F, Strauss G, Wisnivesky J. Platinum-based adjuvant chemotherapy (ACT) in elderly patients with non-small cell lung cancer (NSCLC) in the
SEER-Medicare database: Comparison between carboplatin- and cisplatinbased regimens. J Clin Oncol. 2011;29(suppl; abstr 7014).
18. Strauss G, Herndon J, Maddeus Ma, et al. Adjuvant paclitaxel plus
carboplatin compared with observation in stage IB non-small-cell lung cancer:
CALGB 9633 with the Cancer and Leukemia Group B, Radiation Therapy
Oncology Group and North Central Cancer Treatment Group Study Groups.
J Clin Oncol. 2008;26:5043-5051.
19. Wakelee H, Dahlberg S, Keller D, et al. Interim report of on-study
demographics and toxicity from E1505, a phase III randomized trial of
adjuvant (adj) chemotherapy (chemo) with or without bevacizumab (B) for
completely resected early-stage non-small cell lung cancer (NSCLC). J Clin
Oncol. 2011;29(suppl; abstr 7013)
20. Scagliotti GV, Parikh P, von Pawel J, et al. Phase III study comparing
cisplatin plus gemcitabine with cisplatin plus pemetrexed in chemotherapynave patients with advanced stage non-small cell lung cancer. J Clin Oncol.
2008;26:3543-3551.
21. Kreuter M, Vansteenkiste J, Fischer J Randomized phase II trial on
refinement of early-stage NSCLC adjuvant chemotherapy with cisplatin and
pemetrexed (CPx) versus cisplatin and vinorelbine (CVb): TREAT. J Clin
Oncol. 2011;29(suppl; abstr 7002)
319
TREATMENT OF OLDER PATIENTS WITH

ADVANCED CANCER: BALANCING EFFICACY
WITH TOXICITY (eQ&A)
CHAIR
Supriya Gupta Mohile, MD, MS
University of Rochester Medical Center
Rochester, NY
SPEAKERS
Heidi D. Klepin, MD, MS
Wake Forest University School of Medicine
Winston-Salem, NC
Arati V. Rao, MD
Durham, NC
Considerations and Controversies in the

Management of Older Patients with
Advanced Cancer
By Supriya Gupta Mohile, MD, MS, Heidi D. Klepin, MD, MS, and Arati V. Rao, MD
Overview: The incidence of cancer increases with age. Oncologists need to be adept at assessing physiologic and
functional capacity in older patients in order to provide safe
and efficacious cancer treatment. Assessment of underlying
health status is especially important for older patients with
advanced cancer, for whom the benefits of treatment may be
low and the toxicity of treatment high. The comprehensive
geriatric assessment (CGA) is the criterion standard for evaluation of the older patient. The combined data from the CGA
can be used to stratify patients into categories to better
predict risk for chemotherapy toxicity as well as overall
outcomes. The CGA can also be used to identify and follow-up

on possible functional consequences from treatment. A variety of screening tools might be useful in the oncology practice
setting to identify patients who may benefit from further
testing and intervention. In this chapter, we discuss how the
principles of geriatrics can help improve the clinical care of
older adults with advanced cancer. Specifically, we discuss
assessing tolerance for treatment, options for chemotherapy
scheduling and dosing for older patients with advanced cancer, and management of under-recognized symptoms in older
patients with cancer.
was associated with comorbid conditions.4 In communitydwelling older individuals, the prevalence of polypharmacy
ranges from 15.6% to 94.3%.5 Polypharmacy can increase
the risk of adverse events from chemotherapy. Weight loss
is a marker for declining nutritional status and is often
observed in the geriatric population, particularly in people
who are frail.6 In studies of community-dwelling older people, there was a two-fold increased risk of death among
people who had lost 5% of their body weight.7 Approximately
20% of community-dwelling older adults screen positively
for some degree of cognitive disorder. The presence of cognitive disorders, particularly more advanced disease, may
limit life expectancy. Because cognitive issues are common
in older adults, screening for impairment prior to initiating
treatment is necessary in order to appropriately evaluate
whether patients have capacity for informed consent. Additionally, patients with cognitive disorders may have more
difficulty reporting treatment-related side effects. Depression and social isolation are important prognostic factors in
older patients undergoing treatment for cancer.
PPROXIMATELY 60% of all cancers and 70% of

cancer-related deaths occur in persons age 65 and
older.1 It is uncertain whether there is a benefit to initiating
treatment because data are lacking on efficacy and safety of
the proposed treatment in older patients with health issues
other than cancer. The Comprehensive Geriatric Assessment (CGA) is a tool used by geriatricians to assess functional status, comorbidity, cognition, social support system,
nutrition, and medication use. Results from the CGA can aid
oncologists in predicting outcomes and selecting appropriate
treatment regimens for their patients. The CGA can also
help identify and follow-up on symptoms in older patients
that can affect quality of life. In this chapter, we discuss the
CGA and its role in the care of older patients with advanced
cancer as well as treatment dosing considerations and management of under-recognized cancer-related symptoms in
older patients.
CGA
Although the commonly used Karnofsky Performance Status and Eastern Cooperative Oncology Group (ECOG) performance measures correlate with treatment toxicity, these
tools alone do not predict outcomes as well as the CGA in
the older population. The CGA, a compilation of standardized tools to assess geriatric domains, can help characterize
physiologic age and can detect unsuspected conditions
that may affect cancer treatment in more than 50% of older
patients.
Components of the CGA
Each domain within the CGA has been shown to predict

morbidity and mortality in community-dwelling older adults
(Table 1). Dependence for Activities of Daily Living and
Instrumental Activities of Daily Living (ADLs, IADLs) has
been shown to be predictive of mortality in geriatric oncology,2 and the incidence of ADL and IADL deficiencies is
higher for older patients with cancer than for age-matched
controls.3 The geriatrics literature supports the use of a
directly observed assessment of physical function in order to
assess the risk of falls and identify vulnerability in older
patients who may otherwise seem fit. The prevalence of
comorbidity increases with age and can affect survival of
patients with advanced cancer. In a study of 19,268 patients
with newly diagnosed cancer, decreased duration of survival
Prediction of Outcomes in Older Patients with Advanced Cancer
Similar to the situation with community-dwelling older

adults, several geriatric domains have been shown to influence survival for older patients with cancer. For example,
in the cancer setting, weight loss, malnutrition, and IADL
deficits before diagnosis have been associated with worse
overall survival rates.8 Kanesvaran and colleagues evaluated the effect of CGA domains on overall survival for
patients with advanced cancer and developed a prognostic
scoring system for survival for use by clinicians. Factors that
were independently associated with overall survival included low albumin level, ECOG performance status of 2 or
higher, positive geriatric depression screen, advanced stage
of disease, malnutrition, and older age.
From the Geriatric Oncology Program at the James Wilmot Cancer Center, University of
Rochester Medical Center, Rochester, NY; Wake Forest School of Medicine, Winston-Salem,
NC; Division of Geriatrics, Duke University Medical Center, Durham NC.
Address reprint requests to Supriya Gupta Mohile, MD, MS, Geriatric Oncology Program
at the James Wilmot Cancer Center, University of Rochester Medical Center, 601 Elmwood
Ave, Box 704, Rochester, NY 14642; email: supriya_mohile@urmc.rochester.edu.
1092-9118/10/1-10
321
MOHILE, KLEPIN, AND RAO

Table 1. Domains of Geriatric Assessment and Measurement Options
Domains
Definition
Measurement Options
Function/Physical
Performance
- Ability to carry out ones self care needs to live independently at home
- Ability to care for tasks that allow independence in the community
- Physical performance is an objective evaluation of mobility, balance, and fall risk.
Comorbidity/Pharmacy
- Chronic diseases may influence life expectancy and tolerance to cancer treatment.
Medications can increase risk of adverse events with cancer treatment.
Cognition
Cognitive deficits are common in older patients and may affect decision-making
capacity and interfere with cancer treatment.
Psychological Status
- Depression and anxiety are independently associated with adverse outcomes in

patients with cancer.
- Weight loss and anorexia affect tolerance to treatment and survival in older
patients with cancers.
Nutrition
Social Support
- Adequate social support is necessary for older patients to successfully undergo

treatment.
Two important studies have demonstrated the predictive

value of CGA domains for estimating the risk of severe
toxicity from chemotherapy. One of these studies was conducted by the Cancer and Aging Research Group at seven
institutions.9 A CGA was completed before chemotherapy
in 500 patients who were 65 or older and had cancer. Grade
35 toxicity related to the chemotherapy occurred in 53%
(50% grade 3, 12% grade 4, 2% grade 5). Risk factors for
grade 35 toxicity included age of 73 or older, cancer type
(gastrointestinal or genitourinary), standard dose, polyche-
KEY POINTS
322
Because the incidence of cancer increases with age,

oncologists need to be adept at assessing physiologic
and functional capacity in older patients in order to
provide safe and efficacious cancer treatment.
The combined data from the Comprehensive Geriatric Assessment can be used to stratify older patients
into categories to better predict risk for chemotherapy toxicity as well as overall outcomes.
Older adults frequently receive therapy of a lower
dose intensity, in part because of concerns regarding
toxicity.
Modifications of chemotherapy dose intensity and
treatment schedules for older adults considered to be
unfit require attention in clinical trials to inform the
risks and benefits of therapy.
Supportive care to prevent cancer-related symptoms
is essential for the care of older patients with cancer.
Cachexia and sarcopenia, chemotherapy-associated
peripheral neuropathy, and cancer-related fatigue
are symptoms that are under-recognized in older
patients and warrant further study.
Activities of Daily Living

Instrumental Activities of Daily Living
History of falls
Timed Up and Go
Short Physical Performance Battery
Handgrip testing
Charlson Comorbidity Scale
Cumulative Illness Scale-Geriatrics
Comorbidity count and severity
Medication Count
Beers Criteria
Mini-Mental Status Examination
Blessed-Orientation-Memory Scale
Short Portable Mental Status Questionnaire
Montreal Cognitive Assessment
Geriatric Depression Scale
Hospital Anxiety and Depression Scale
Mini-Nutritional Assessment
Weight loss
Body Mass Index
Needs assessment of financial capabilities,
transportation, and caregiver status.
Medical Outcomes Survey Social Support
motherapy, fall within the past 6 months, assistance needed

for IADLs, and decreased social activity. A risk stratification
schema (number of risk factors: percentage of grade 35
toxicity) was developed. Further validation of this schema is
in progress.
In the second study, use of the Chemotherapy Risk Assessment Scale for High-Age Patients (CRASH) was evaluated in more than 500 patients who were age 70 or older.10
A CGA was completed before chemotherapy was started.
Toxicity of the chemotherapy regimen was adjusted using an
index to estimate the average per-patient risk of chemotherapy toxicity (the MAX2 index): severe toxicity occurred in
64% of patients. The best model for hematologic toxicity
included IADL score, serum lactate dehydrogenase level,
diastolic blood pressure, and chemotherapy toxicity. The
best predictive model for nonhematologic toxicity included
performance status, Mini-Mental Status score, MiniNutritional Assessment score, and chemotherapy toxicity.
Information from two-thirds of the patients was used to
develop the risk stratification scheme, and the tool was
validated in the remaining one-third.
Overall, these predictive risk stratification schemes allow
clinicians to identify patients at highest risk for chemotherapy toxicity and can be used in future research to apply
interventions to reduce chemotherapy toxicity in vulnerable
older populations.
Models of Care of the Older Patient with
Advanced Cancer
Despite recent studies demonstrating the feasibility of the

CGA in oncology, adoption as the standard of care has been
slow because of lack of resources, difficulties with interpretation of results, and challenges in implementing targeted
interventions in specialty clinic settings such as oncology. A
short, simple, validated screening procedure could quickly
identify patients who are at risk for morbidity or death in
the oncology clinic. Partnering with geriatricians or a geri-
OLDER PATIENTS WITH ADVANCED CANCER
atric oncology consultative service can help provide geriatric

assessment.
For example, review of medication usage by a pharmacist

can decrease suboptimal prescribing and potentially lead to
a decrease of adverse drug events.22
Screening for Impairments in the Oncology Clinic
Several screening tools have been evaluated in oncology

clinical care and research. The Vulnerable Elders Survey-13
(VES-13) is a self-administered survey that consists of one
question for age and an additional 12 items assessing
self-rated health, functional capacity, and physical performance.11 In 400 community-dwelling older adults, increasing VES-13 scores strongly predicted death and functional
decline.12 In a national cohort, a high proportion (45.8%) of
elders with a history of cancer also scored as vulnerable
on the VES-13.3 Luciani and colleagues conducted a study in
a population of 419 patients age 70 and older with a history
of any type of solid or hematologic cancer;13 the sensitivity
and specificity of VES-13 was 87% and 62%, respectively,
compared with that of the CGA.
The Groningen Frailty Indicator (GFI) is a 15-item survey
that includes questions focusing on mobility/physical fitness,
vision/hearing, nutrition, comorbidity, cognition, and psychosocial health.14 The GFI score has been shown to correlate moderately with CGA results.15 In a study by Aaldriks
and colleagues, the mortality rate after initiation of chemotherapy was increased for patients with higher baseline GFI
scores.16 The GFI has also been shown to be predictive of
outcome in older patients with non-small cell lung cancer
treated with platinum-based doublet chemotherapy.17
The G8 is an eight-item questionnaire designed to assess
domains of nutrition, mobility, cognitive deficit, polypharmacy, age, and self-perceived health status. Soubeyran and
colleagues recommend a score of 14 as a predictor of CGA
deficits (90% sensitivity and 60% specificity).18 In one study,
sensitivity of the G8 for identifying deficits in older patients
with cancer was found to be superior to that of the VES-13.18
Because these results have not been consistent, screening
tools should not serve as a substitute for a full CGA.
However, if a full CGA cannot be undertaken, screening
tools may serve as a way to capture some important information within geriatric domains.
Multidisciplinary Approach to Care
A few studies have demonstrated that combining geriatric

and oncologic approaches can affect treatment decision making for patients with advanced cancer. In the ELCAPA
study, a geriatrician performed an extensive CGA and
proposed a geriatric intervention plan for older patients with
cancer.19 After the CGA, the initial cancer treatment plan
was modified for 78 (20.8%) of 375 patients, usually to
decrease treatment intensity. In another study of 161 patients (more than 50% of whom had advanced cancer),
cancer treatment was changed in 49% of patients.20 In a
pilot study, Horgan and colleagues demonstrated that although most eligible older patients were not referred for
geriatric assessment, when such assessment was done, the
results guided initial decision making.21 More data are
necessary to determine whether CGA and interventions can
improve outcomes. It is likely that support from multidisciplinary expertise, including social work, physical therapy,
occupational therapy, and nutrition can help develop CGAguided interventions for an at-risk older adult with cancer.
Dosing Considerations for Older Patients with

Advanced Cancer
Age-related physiologic changes and comorbid disease can

increase the risk of chemotherapy toxicity. The balance
between risk of treatment and potential benefit is particularly challenging when treating with noncurative intent.
With respect to treatment characteristics, both standard
dosing of chemotherapy and polychemotherapy have been
associated with increased toxicity among older patients with
cancer.9 Dose modification, therefore, is one consideration
for treatment planning to minimize the negative consequences of toxic therapies. Given the paucity of clinical trial
data specific to older patients, there are more questions than
answers regarding dose-modification strategies, particularly
in the setting of first-line treatment for patients with advanced cancer.
How Common Is Dose Modification for Older Adults
with Advanced Cancer?
Although several large population-based studies have

shown increased age as a risk factor for lower relative-dose
intensity during adjuvant chemotherapy, few data have
been collected in the metastatic setting. In a retrospective
single-insitution study of older patients (mean age, 74.6
years, ECOG score, 0 2) with advanced cancer, 44% had a
dose modification either at initiation of treatment or because
of toxicity.23 Gajra and colleagues studied factors associated
with primary reduction of chemotherapy dosing during the
first course of treatment among older adults receiving palliative chemotherapy in a secondary analysis of a multisite
observational study.24 Almost one-third (29%) of the 319
patients had a primary reduction of chemotherapy dosing.
Older age, a primary diagnosis of lung cancer, and comorbid
conditions were the factors independently associated with a
reduction in this cohort.
Indications for Dose Adjustment Related to
Aging Physiology
Physiologic changes associated with aging have implications for chemotherapy toxicity among older adults. Aging is
associated with decreased intestinal absorption, changes in
volume of distribution, decreased hepatic metabolism, and
impaired renal excretion. The degree to which these changes
have clinical significance can vary greatly within an older
population. Among the changes, a change in renal function
is the most well described and needs to be accounted for
in dosing considerations. Many chemotherapy drugs are
cleared through the kidneys and it is well-documented that
serum creatinine can provide a substantial underestimation
of renal function in older adults. A creatinine clearance
should be calculated for all older adults before chemotherapy is initiated, to inform dose adjustment for drugs cleared
through the kidneys. Multiple guidelines provide specific
recommendations for renal dose adjustment, including a
position paper published by the International Society of
Geriatric Oncology (SIOG).25
323

What Is the Benefit Versus Cost of Dose Modification
in the Metastatic Setting? Clinical Trial Evidence in
Colorectal Cancer
Depending on the rationale for dose modification, treatment efficacy may be compromised by an effort to decrease
toxicity. In contrast, quality of life may be substantially
improved by avoiding treatment toxicity. In general, guideline organizations such as the National Comprehensive
Cancer Network (NCCN) do not recommend dose attenuation during first-line therapy based on chronologic age if
the patient is fit. Many patients seen in clinical practice,
however, are less fit than those enrolled on trials, making
the judgment regarding risks and benefits of standard
treatment much more complicated.
A recently published trial addresses, in part, the issue of
dose reduction of first-line treatment. This study, by Seymour and colleagues, involved the use of a 22 factorial
design to randomly assign 459 frail (considered by their
physician to be unfit for full-dose combination chemotherapy) older adults with metastatic colorectal cancer to one of
four first-line systemic therapies using an attenuated starting dose (80% of standard).26 The study design allowed for
escalation to full-dose therapy after 6 weeks if tolerated.
The four treatments were 48-hour intravenous fluorouracil
(5FU), oxaliplatin plus 5FU, capecitabine alone, or oxaliplatin plus capecitabine. In addition to a standard quality-oflife questionnaire, a composite outcome of overall treatment
utility was incorporated in an attempt to capture patient
and physician satisfaction with the outcome of each treatment decision.
The median age of the patients was 74, and 13% were
older than 80. The primary reason the patients were considered unfit for standard therapy was frailty or older chronologic age. During protocol treatment, few patients had dose
escalation and 49% required additional dose reduction (below the 80% of the standard dose). There was a trend toward
benefit with use of oxaliplatin in this attenuated dosing
schedule, although the primary outcome was not significant (median progression-free survival, 5.8 months vs. 4.5
months; p 0.07). No substantial increase in grade 3
toxicity was associated with use of oxaliplatin, but oxaliplatin was associated with increased overall treatment utility,
suggesting a palliative benefit. Alternatively, capecitabine
was equivalent in efficacy to 5FU but offered no benefit in
quality of life and was also associated with a substantial
increase in toxicity. This study provides evidence for treatment of unfit older adults with metastatic colorectal cancer
with an attenuated chemotherapy regimen and introduces
additional outcome measures to help quantify the palliative
benefit a patient may receive from therapy.
In addition to dose reduction, questions regarding optimal
dosing schedules, including the option for treatment break,
for older patients are also common. Again, few studies have
addressed this issue specific to older patients, yet older
patients are most likely to have debilitating consequences
from ongoing therapy. For colorectal cancer, some data
can be extrapolated from existing studies. For patients fit
enough to receive combination chemotherapy with an
oxaliplatin-based regimen (FOLFOX), data from randomized trials support de-escalation to 5FU maintenance after
3 months of oxaliplatin. Although a fraction of patients
enrolled on these studies27,28 were older than 75, there is
324
suggestion of similar benefit compared with younger patients when this strategy is used.29 Additionally, data from
two randomized trials suggest that selected patients may
retain the efficacy of first-line combination therapy with
oxaliplatin- or irinotecan-based regimens despite prescribed
chemotherapy-free intervals.30,31 Although the evidence is
limited, it is reasonable to incorporate these data into
treatment planning for older patients in an attempt to
minimize the negative consequences of therapy on quality of
life.
Supportive Care Interventions to Maximize Treatment
Efficacy in Older Patients with Advanced Cancer
Several organizations including NCCN and ASCO have

developed guidelines to help manage common side effects.
We focus here on three areas that cause substantial distress
for older patients with cancer and require further research:
cachexia and sarcopenia, chemotherapy-associated peripheral neuropathy and falls, and cancer-related fatigue.32,33
Cancer Cachexia and Sarcopenia
Sarcopenia is the progressive generalized loss of skeletal

muscle mass, strength, and function. Cachexia has no uniform definition and is a complex metabolic syndrome that
is characterized by weight loss of more than 10%, reduced
food intake (1,500 kcal/d), and systemic inflammation
(C-reactive protein level 10 mg/L).34 It is estimated that
50% of people older than 80 have sarcopenia. Half of all
patients with cancer lose some body weight; one third lose
more than 5% of body weight, and as many as 20% of all
cancer deaths are caused directly by cachexia.34
Figure 1 illustrates how the pathophysiology of cancer
cachexia and sarcopenia are intertwined.35 These syn-
Fig 1. Interaction and consequences between elderly host-tumortherapy.

Abbreviations: ADLs, activities of daily living.

Table 2. Characteristics of Oncology Agents that Cause Peripheral Neuropathy
Agent
Incidence
Taxanes: paclitaxel,
docetaxel
36%
Platinum agents: cisplatin,

carboplatin, oxaliplatin
30% cisplatin
6080%: oxaliplatin
20%: carboplatin when
combined with
paclitaxel
31% with vincristine
Vinca alkaloids: vincristine,

vinblastine, vindesine,
vinorelbine, vinflunine
Nucleoside analogs:
cytarabine, gemcitabine
110%
Bortezomib
30%
Thalidomide
2040%
Bevacizumab
Case reports
Type and Symptoms
Axonal
Tingling and numbness within 24 hr of infusion
Motor neuropathy can cause foot drop
Acute neuropathic pain
Pure sensory neuropathy
Autonomic symptoms: dizziness, impotence, orthostatic
hypotension
High-frequency hearing loss with cisplatin
Oropharyngeal dysesthesia with oxaliplatin
Sensorimotor neuropathy
Symptoms usually seen within 3 mos of therapy
Tingling and numbness in hands and feet
Muscle weakness and leg cramps
Paralytic ileus and megacolon
Demyelinating polyneuropathy
Progressive monophasic course within 23 wk after
beginning of treatment
Severe motor weakness
Quadriparesis requiring ventilatory support
Predominantly sensory and distal nerves
Length dependent
Axonal loss present
Tingling and numbness in hands and feet
Muscle cramps
Severe optic neuropathy
dromes are associated with increased levels of proinflammatory cytokines that have a direct effect on muscle
metabolism and anorexia. One of the most prominent mechanisms is the anabolic resistance of older muscles to postprandial amino acid loading, which leads to a negative
whole-body protein balance.36 In turn, this negative balance
leads to hypermetabolism with increased resting energy
expenditure. The consequences of sarcopenia and cachexia
in older adults include fatigue, depression, decreased mobility, depression, and falls.
To date, no clinically applied regimen has been completely
successful in reversing cancer-associated muscle or weight
loss. A Cochrane review demonstrated that megestrol acetate (Megace) improved appetite and weight gain.37 However, fat mass rather than lean body mass increased and
there was no benefit in the quality of life. Prednisolone and
dexamethasone have been tested in randomized trials and
both have improved appetite and well-being compared with
placebo.38 Steroids and megestrol acetate are not recommended for long-term use because of such side effectsas
venous thromboembolism and adrenal insufficiency. Randomized trials of nutritional supplements have shown an
improved net food intake and increased physical activity but
no benefit39 in terms of lean body mass or survival. The
authors of another Cochrane review concluded that there
was insufficient data to establish whether eicosapentaenoic
acid (EPA) was better than placebo and that there was no
evidence that EPA improves symptoms when combined with
megestrol acetate.40 In one large study,41 patients were
randomly assigned to one of the following five arms: medroxyprogesterone acetate, EPA, l-carnitine, thalidomide,
or medroxyprogesterone acetate plus EPA plus l-carnitine
plus thalidomide. Interim analysis of data for 125 patients
showed a substantial improvement of fatigue in the
l-carnitine arm and the combination treatment arm. Angio-
Characteristics
Dose dependent; i.e., based on treatment schedule,

infusion duration, and cumulative dose
- Amount of platinum binding to DNA is comparable to or

exceeds levels known to be cytotoxic to tumor cells
- cumulative dose for cisplatin: 400500 mg/m2
- cisplatin-induced peripheral neuropathy is irreversible in
3050% of patients
- Lifelong sequelae in survivors
- Vincristine and Vindesine cause worse PN than
Vinblastine and Vinorelbine
- Cap Vincristine dose at 2 mg
- Typically seen with high doses of cytarabine (3 gm/m2)
which can also cause central cerebellar neuropathy
- Reversible
- Dose and frequency related
- Dose dependent
- Increases with age
- Concurrent radiation may have been a risk factor
tensin converting enzyme-inhibitors, cytokine inhibitors

(infliximab or anti-tumor-necrosis factor), myostatin antagonists, peroxisome-proliferator-activated-receptor- (PPAR-)
agonists, selective androgen-receptor modulators (ostarine),
and gherelin agonists are all currently being evaluated
in clinical trials for the management of sarcopenia and
cachexia. Resistance exercise training is an effective intervention that augments muscle mass and strength and attenuates the development of sarcopenia in older patients.
Chemotherapy-Associated Peripheral Neuropathy and Falls
The incidence of cancer treatment-related peripheral neuropathy is variable and ranges from 30% to 40%. The
neuropathy usually affects distal sites first, and as cumulative doses increase, symptoms progress in severity. Sensory
symptoms and signs typically develop before motor symptoms, and pain will develop in a subset of patients. More
severe and persistent neuropathy may develop in patients
with pre-existing neuropathy (related to diabetes or vitamin
B12 deficiency). Table 2 lists characteristics of oncology
agents that can cause treatment-related peripheral neuropathy.42 Interestingly, one study demonstrated that patients
age 65 or older do not have greater risk of peripheral
neuropathy, and advanced age is not associated with increased severity of the disorder.43 However, peripheral neuropathy is often only partially reversible and can thus cause
substantial long-term morbidity in older patients. The risk
of falls increases with each cycle of chemotherapy and is
highest for patients with worse symptoms from peripheral
neuropathy.44 These patients had more severe muscle weakness, loss of balance, and a higher interference with walking
or driving. Physical performance measures from the CGA
can help identify patients and aid in the development of
interventions for patients with cancer treatment-related
peripheral neuropathy who are at most at risk for falling.
325

Table 3. Etiology of Cancer-Related Fatigue
Peripheral component: negative energy imbalance
Cancer itself or therapy for cancer
Treatment for cancer: chemotherapy, radiation therapy, surgery, hormone
therapy, targeted kinase inhibitors
Systemic infections
Hypothyroidism
Anemia
Malnutrition
Metabolic abnormalities
Sleep disorders
Psychological factors: depression, anxiety
Central component
Hyperactivity of the hypothalamic-pituitary-adrenal axis from stress (i.e.,
increase in cortisol or corticotrophin-releasing factor
Decreased gonadotropin levels
Increase in the numbers of circulating T-lymphocytes
Increased interleukin-1 receptor antagonist
Increased soluble tumor necrosis factor receptor type II
Increased neopterin levels
Currently, no proven pharmacologic treatments are available for established chemotherapy-associated peripheral
neuropathy. Discontinuation of the causative agent or dose
reductions may help. Pharmacologic agents that have been
tried include topical amitriptyline with ketamine, topical
lidocaine, selective serotonin norepinephrine uptake inhibitors, and antiepileptics. Because chemotherapy-associated
peripheral neuropathy is only partially reversible and
maybe permanent, several studies have focused on prevention of the disorder.45 Calcium and magnesium infusions
and glutathione have been shown to prevent neurotoxicity
from oxaliplatin and cisplatin, respectively. Clinical trials
involving patients receiving oxaliplatin or paclitaxel have
demonstrated that glutamine substantially reduces neuropathy symptoms along with decreased incidence of motor
weakness and gait disturbances.
Cancer-Related Fatigue
The NCCN defines cancer-related fatigue as a distressing

persistent subjective sense of tiredness or exhaustion that is
not proportional to recent activity and interferes with usual
functioning. Cancer-related fatigue may be an early symptom of malignant disease and is reported by as many as 40%
of patients at the time of diagnosis. In addition, as many as
90% of patients receiving radiation therapy and as many as
80% of patients receiving chemotherapy experience fatigue.
One study demonstrated no association between fatigue and
age.45 A number of studies found an association between
hemoglobin levels, pain, and nonpain symptoms with fatigue. In the Health and Retirement Study, 67% of the
patients with cancer were 65 and olderand approximately
50% had one or two coexisting medical conditions.46 Patients
with cancer had a substantially higher risk for fatigue,
depression, and pain. Table 3 provides information on the
etiology and possible pathophysiology of cancer-related fatigue.47
326
To date, there is no U.S. Food and Drug Administration

(FDA) approved drug for the treatment of cancer-related
fatigue. [In addition, there are no studies that have been
conducted to study CRF specifically in elderly patients with
cancer.] If fatigue is secondary to low hemoglobin, erythropoietin stimulating agents (ESA) may be used to treat the
underlying anemia. However, ESAs must be used with
caution given the risk of thromboembolic disease, and they
continue to be contraindicated in patients receiving myelosuppressive therapy for curative cancers. A randomized
study of methylphenidate (target dose, 54 mg/d) in 148 adult
patients with cancer did not significantly improve cancerrelated fatigue (p 0.35).48 A subset analysis suggested
that patients with more severe fatigue and/or with more
advanced disease did have some improvement in fatigue.
Patients in the methylphenidate arm reported more nervousness and appetite loss. Several studies have shown no
benefit of selective serotonin reuptake inhibitors (sertraline
or paroxetine) for the treatment of cancer-related fatigue in
patients with cancer. Modafinil, a wakefulness drug, improved severe cancer-related fatigue in a large randomized
study.49
Nonpharmacologic interventions have also been studied
for their effect on cancer-related fatigue.50 A Cochrane
review demonstrated that exercise improves cancer-related
fatigue during and after cancer therapy.51 Caution in developing the exercise prescription is recommended for patients
with bone metastases or myelosuppression andpatients who
are febrile. Exercise interventions included moderately intense (55% to 75% of heart rate) aerobic exercise (e.g.,
walking, cycling), ranging from 10 90 minutes 3 to 7 days
per week. Psychological and behavioral interventions47 that
have been evaluated for cancer-related fatigue include support groups, yoga, and cognitive behavioral interventions.
These interventions have improved cancer-related fatigue
and vitality in most studies. These interventions also can
be combined with exercise for maximizing benefit, as was
demonstrated in the GROUP-HOPE trial.52
Conclusion
The incidence of cancer increases with age. An older

patients chronologic age does not always reflect his or her
overall health status. Therefore, oncologists need to be adept
at assessing physiologic and functional capacity in older
patients. The CGA is the criterion standard for evaluation of
an older patient. The combined data from the CGA can be
used to stratify patients into risk categories to better predict
chemotherapy toxicity and survival. More research is
needed on the best ways to incorporate the CGA into
decision making for cancer treatment, implement novel
treatment dosing options, and develop interventions to improve symptoms of sarcopenia, cachexia, chemotherapyassociated peripheral neuropathy, and cancer-related
fatigue in older patients with cancer.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Supriya Gupta Mohile*

Heidi D. Klepin*
Arati V. Rao*
REFERENCES
1. Smith BD, Smith GL, Hurria A. Future of Cancer Incidence in the
United States: Burdens Upon an Aging, Changing Nation. J Clin Oncol.
2009;27:2758-2765.
2. Wedding U, Rohrig B, Klippstein A, et al. Age, severe comorbidity and
functional impairment independently contribute to poor survival in cancer
patients. J Cancer Res Clin Oncol. 2007;133:945-950.
3. Mohile SG, Xian Y, Dale W, et al. Association of a cancer diagnosis with
vulnerability and frailty in older Medicare beneficiaries. J Natl Cancer Inst.
2009;101:1206-1215.
4. Piccirillo JF, Tierney RM, Costas I, et al. Prognostic importance of
comorbidity in a hospital-based cancer registry. JAMA. 2004;291:2441-2447.
5. Buck MD, Atreja A, Brunker CP, et al. Potentially inappropriate
medication prescribing in outpatient practices: prevalence and patient characteristics based on electronic health records. Am J Geriatr Pharmacother.
2009;7:84-92.
6. Fried LP, Kronmal RA, Newman AB, et al. Risk factors for 5-year
mortality in older adults: the Cardiovascular Health Study. JAMA. 1998;279:
585-592.
7. Diehr P, Bild DE, Harris TB, et al. Body mass index and mortality in
nonsmoking older adults. Am J Pub Health. 1998;88:623-629.
8. Kanesvaran R, Li H, Koo KN, Poon D. Analysis of prognostic factors of
comprehensive geriatric assessment and development of a clinical scoring
system in elderly Asian patients with cancer. J Clin Oncol. 2011;29:36203627.
2011;29:3457-3465.
10. Extermann M, Boler I, Reich RR, et al. Predicting the risk of chemotherapy toxicity in older patients: the Chemotherapy Risk Assessment Scale
for High-Age Patients (CRASH) score. Cancer. Epub 2011 Nov 9.
11. Saliba D, Elliott M, Rubenstein LZ, et al. The Vulnerable Elders
Survey: a tool for identifying vulnerable older people in the community. J Am
Geriatr Soc. 2001;49:1691-1699.
12. Min LC, Elliott MN, Wenger NS, Saliba D. Higher vulnerable elders
survey scores predict death and functional decline in vulnerable older people.
J Am Geriatr Soc. 2006;54:507-511.
13. Luciani A, Ascione G, Bertuzzi C, et al. Detecting disabilities in older
patients with cancer: comparison between comprehensive geriatric assessment and vulnerable elders survey-13. J Clin Oncol. 2010;28:2046-2050.
14. Steverink NSJ, Schuurmans H, et al. Measuring frailty: developing
and testing the GFI (Groningen Frailty Indicator). The Gerontologist. 2001;
41(Special Issue 1):236.
15. Schrijvers D, Baiter A, Vos MD, et al. Evaluation of the Groningen
Frailty Index (GFI) as a screening tool in elderly patients (pts): an interim
analysis. 2010 ESMO; Abstract 566PD.
16. Aaldriks AA, Maartense E, le Cessie S, et al. Predictive value of
geriatric assessment for patients older than 70 years, treated with chemotherapy. Crit Rev Oncol Hematol. 2011;79:205-212.
17. Biesma B, Wymenga AN, Vincent A, et al. Quality of life, geriatric
assessment and survival in elderly patients with non-small-cell lung cancer
treated with carboplatin-gemcitabine or carboplatin-paclitaxel: NVALT-3 a
phase III study. Ann Oncol. 2011;22:1520-1527.
18. Soubeyran P, Bellera C, Goyard J, et al. Validation of the G8 screening
tool in geriatric oncology: The ONCODAGE project. J Clin Oncol. 2011;29
19. Caillet P, Canoui-Poitrine F, Vouriot J, et al. Comprehensive geriatric
assessment in the decision-making process in elderly patients with cancer:
ELCAPA study. J Clin Oncol. 2011;29:3636-3642.
20. Chaibi P, Magne N, Breton S, et al. Influence of geriatric consultation
with comprehensive geriatric assessment on final therapeutic decision in
elderly cancer patients. Crit Rev Oncol Hematol. 2011;79:302-307.
21. Horgan AM, Leighl NB, Coate L, et al. Impact and feasibility of a
comprehensive geriatric assessment in the oncology setting: a pilot study.
Am J Clin Oncol. Epub 2011 Mar 17.
22. Hanlon JT, Weinberger M, Samsa GP, et al. A randomized, controlled

trial of a clinical pharmacist intervention to improve inappropriate prescribing in elderly outpatients with polypharmacy. Am J Med. 1996;100:428-437.
23. Luciani A, Marussi D, Ascione G, et al. Do elderly cancer patients
achieve an adequate dose intensity in common clinical practice? Oncology.
2006;71:382-387.
24. Gajra A, Hardt W, Tew W, et al. Primary dose reduction (PDR) of
chemotherapy (chemo) in patients (Pts) older than age 65 with advanced
cancer (Ca) and toxicity outcomes. J Clin Oncol. 2011:9037.
25. Lichtman SM, Wildiers H, Launay-Vacher V, Steer C, Chatelut E,
Aapro M. International Society of Geriatric Oncology (SIOG) recommendations for the adjustment of dosing in elderly cancer patients with renal
insufficiency. Eur J Cancer. 2007;43:14-34.
26. Seymour MT, Thompson LC, Wasan HS, et al. Chemotherapy options
in elderly and frail patients with metastatic colorectal cancer (MRC
FOCUS2): an open-label, randomised factorial trial. Lancet. 2011;377:17491759.
27. Tournigand C, Cervantes A, Figer A, et al. OPTIMOX1: a randomized
study of FOLFOX4 or FOLFOX7 with oxaliplatin in a stop-and-Go fashion in
advanced colorectal cancera GERCOR study. J Clin Oncol. 2006;24:394-400.
28. Chibaudel B, Maindrault-Goebel F, Lledo G, et al. Can chemotherapy
be discontinued in unresectable metastatic colorectal cancer? The GERCOR
OPTIMOX2 Study. J Clin Oncol. 2009;27:5727-5733.
29. Figer A, Perez-Staub N, Carola E, et al. FOLFOX in patients aged
between 76 and 80 years with metastatic colorectal cancer: an exploratory
cohort of the OPTIMOX1 study. Cancer. 2007;110:2666-2671.
30. Adams RA, Meade AM, Seymour MT, et al. Intermittent versus
continuous oxaliplatin and fluoropyrimidine combination chemotherapy for
first-line treatment of advanced colorectal cancer: Results of the randomised
phase 3 MRC COIN trial. Lancet Oncol. 2011;12:642-653.
31. Labianca R, Sobrero A, Isa L, et al. Intermittent versus continuous
chemotherapy in advanced colorectal cancer: a randomised GISCAD trial.
Ann Oncol. 2011;22:1236-1242.
32. NCCN guidelines for supportive care. Version 2.2011 http://www.
nccn.org/professionals/physician_gls/f_guidelines.asp#suportive.
Accessed
March 15, 2012.
33. Basch E, Prestrud AA, Hesketh PJ, et al. Antiemetics: American
Society of Clinical Oncology Clinical Practice Guideline Update. J Clin Oncol.
2011;29:4189-4198.
34. Fearon K, Strasser F, Anker SD, et al. Definition and classification of
cancer cachexia: an international consensus. Lancet Oncol. 2011;12:489-495.
35. Morley JE, Baumgartner RN. Cytokine-related aging process. J Gerontol A Biol Sci Med Sci. 2004;59:M924-M929.
36. Cuthbertson D, Smith K, Babraj J, et al. Anabolic signaling deficits
underlie amino acid resistance of wasting, aging muscle. FASEB J. 2005;19:
422-444.
37. Berenstein EG, Ortiz Z. Megestrol acetate for the treatment of
anorexia-cachexia syndrome. Cochrane Database Syst Rev. 2005:CD004310.
38. Willox JC, Corr J, Shaw J, et al. Prednisolone as an appetite stimulant
in patients with cancer. Br Med J. 1984;288:27.
39. Fearon KC, Von Meyenfeldt MF, Moses AG, et al. Effect of a protein
and energy dense N-3 fatty acid enriched oral supplement on loss of weight
and lean tissue in cancer cachexia: a randomised double blind trial. Gut.
2003;52:1479-1486.
40. Dewey A, Baughan C, Dean T, et al. Eicosapentaenoic acid (EPA, an
omega-3 fatty acid from fish oils) for the treatment of cancer cachexia.
Cochrane Database Syst Rev. 2007:CD004597.
41. Mantovani G. Randomised phase III clinical trial of 5 different arms of
treatment on 332 patients with cancer cachexia. Eur Rev Med Pharmacol Sci.
2010;14:292-301.
42. Gutierrez-Gutierrez G, Sereno M, et al. Chemotherapy-induced peripheral neuropathy: Clinical features, diagnosis, prevention and treatment
strategies. Clin Transl Oncol. 2010;12:81-91.
43. Argyriou AA, Polychronopoulos P, Koutras A, et al. Is advanced age
327

associated with increased incidence and severity of chemotherapy-induced
peripheral neuropathy? Support Care Cancer. 2006;14:223-229.
44. Tofthagen C, Overcash J, Kip K. Falls in persons with chemotherapyinduced peripheral neuropathy. Support Care Cancer. 2012;20:583-589.
45. Prue G, Rankin J, Allen J, et al. Cancer-related fatigue: a critical
appraisal. Eur J Cancer. 2006;42:846-863.
46. Reyes-Gibby CC, Aday LA, et al. Pain, depression, and fatigue in
community-dwelling adults with and without a history of cancer. J Pain
Symptom Manage. 2006;32:118-128.
47. Rao AV, Cohen HJ. Fatigue in older cancer patients: etiology, assessment, and treatment. Semin Oncol. 2008;35:633-642.
48. Moraska AR, Sood A, Dakhil SR, et al. Phase III, randomized, doubleblind, placebo-controlled study of long-acting methylphenidate for cancerrelated fatigue. J Clin Oncol. 2010;28:3673-3679.
328
49. Jean-Pierre P, Morrow GR, Roscoe JA, et al. A phase 3 randomized,

placebo-controlled, double-blind, clinical trial of the effect of modafinil on
cancer-related fatigue among 631 patients receiving chemotherapy: a University of Rochester Cancer Center Community Clinical Oncology Program
Research base study. Cancer. 2010;116:3513-3520.
50. Mustian KM, Morrow GR, Carroll JK, et al. Integrative nonpharmacologic behavioral interventions for the management of cancer-related fatigue.
Oncologist. 2007;12 Suppl 1:52-67.
51. Cramp F, Daniel J Exercise for the management of cancer-related
fatigue in adults. Cochrane Database Syst Rev. 2008:CD006145.
52. Courneya KS, Friedenreich CM, Sela RA, et al. The group psychotherapy and home-based physical exercise (group-hope) trial in cancer survivors:
Physical fitness and quality of life outcomes. Psychooncology. 2003;12:357374.
RECENT CLINICAL HIGHLIGHTS IN

GYNECOLOGIC ONCOLOGY
CHAIR
Lynette Denny, MD, PhD
University of Cape Town
Rondebosch, South Africa
SPEAKERS
D. Scott McMeekin, MD
University of Oklahoma Health Sciences Center
Oklahoma City, OK
Nicoletta Colombo, MD
University of Milan-Bicocca
Milan, Italy
International Perspective on the Global

Advances in Gynecologic Oncology
By Lynette Denny, MD, PhD
Overview: The treatment of gynecologic cancer has evolved

over the years, with greater emphasis on tailored surgery and
reducing morbidity and mortality related to surgery, particularly in the management of vulvar and cervical cancer. The
addition of concurrent chemotherapy to radiation regimens
has improved survival of patients with cervical cancer in
developed countries; however, most women with cancer in
developing countries have advanced, untreatable disease and
minimal access to anticancer therapies. In the past 15 years
there has been intense research into alternatives to cervical
YNECOLOGIC ONCOLOGY includes a spectrum of

cancer that affects the vulva, vagina, cervix, uterus,
tubes, ovaries, peritoneum, and, in some European countries, breast. Cancers of the lower genital tract, most of
which are etiologically associated with infection with highrisk types of human papillomavirus (HPV), are more commonly found in developing compared with developed
countries. The inverse is true for cancers of the upper genital
tract. Over the years, substantial advances have been made
in the management of gynecological cancers. This article
explores those advances.
Vulvar Cancer
Vulvar cancer is the fourth most common gynecologic

cancer, accounting for 4% to 5% of all malignant tumors of
the female genital tract. Age standardized incidence rates
(ASIRs) vary from 0.1 per 100,000 population (Algeria) to
1.6 to 1.7 per 100,000 population (Zimbabwe and United
Kingdom, respectively).1 Vulvar cancer is usually diagnosed
in older women (mean age, 65 to 70 years), who often
have lichen sclerosus et atrophicus or differentiated vulvar
intraepithelial neoplasm, which is not related to HPV.
Increasingly, however, vulvar cancer is diagnosed in
younger women, which relates to persistent infection with
HPV and infection with HIV. Toki and colleagues2 found
the mean age of patients with HPV-related vulvar cancer to
be 55 years, whereas that of patients with nonHPVassociated vulvar cancer was 77 years. Eva and colleagues3
found that 85.7% of 70 women who were diagnosed as
having differentiated vulvar intraepithelial neoplasm had
concurrent, previous, or subsequent vulvar squamous cell
carcinoma. However, only 25% of women with HPV-related
undifferentiated vulvar intraepithelial neoplasm and 33%
of women with lichen sclerosus et atrophicus and squamous
cell hyperplasia had a history of developing invasive vulvar
cancer.
Treatment of vulvar cancer is primarily surgical. Up to the
late 1980s the most common procedure was a radical vulvectomy with en bloc resection of bilateral groin node lymphadenectomythe so-called butterfly incision pioneered by
Antoine Basset in 1912. In the 1940s, Frederick Taussig
adapted the Basset operation by performing separate groin
incisions, which resulted in improved survival rates, albeit
with very high operative morbidity and mortality rates. The
triple incision technique was introduced, and although no
randomized controlled trials have compared this technique
330
cytologic testing, particularly in low resourced regions but

also in an attempt to improve on cytologic testing in developed countries. Surgical staging in endometrial cancer has
enabled the use of adjuvant radiation to be individualized to
the patients particular risk factors for recurrence. The management of ovarian cancer, long stagnant since the introduction of platinum and paclitaxel as chemotherapeutic agents, is
set to change with the onset of molecular and genetic profiling
and the introduction of novel therapies.
with the butterfly incision, retrospective studies found no

inferior outcomes and a low risk of bridge recurrence
(2.4%).4,5 Compared with en bloc resection, rates of associated morbidity, such as wound breakdown and lymph drainage problems, were substantially lower with the triple
incision technique.6 However, this technique still involved
removal of external genitalia.
The introduction of radical wide local incision further
reduced operative morbidity, and this approach, which often
allowed preservation of anatomy, was not shown to be
inferior in terms of oncologic safety. Current guidelines
suggest that 1 cm of disease-free tissue around the invasive
lesion represents a safe margin; however, hard data on the
value of disease-free margins are lacking, and some recent
publications suggest that margins are irrelevant if the
entire lesion is excised.7 Radical wide local excision has
enabled a reasonably accurate definition of lateral compared
with central lesions. Central lesion refers to the ability to
remove the invasive lesion with 1 cm of disease-free tissue
without damaging or removing central structures (e.g., clitoris, urethra, and anus). Lateral lesion allows 1 cm of
disease-free tissue without compromising central structures. To preserve the central structures, it is possible to
perform primary chemoradiation and to only perform surgery for residual disease. Moore et al treated 58 women with
T3 or T4 tumors not amenable to surgical resection with
radiation (1.8Gy daily 32 fractions) plus weekly cisplatinum (40mg/m2), followed by surgical resection of residual
tumor or biopsy to confirm pathological complete response.
Forty of the 58 women completed treatment and 37/58 (64%)
women had a complete clinical response. Of these women,
34 underwent biopsy and 29 (78%) of these women had a
complete pathological response.8 Rogers et al reported on a
retrospective study of 50 women treated with chemoradiation for advanced vulvar cancer in Cape Town from 1982 to
2001. Only 14 (28%) of the women had a complete response,
29 (58%) had a partial response. Of the women who under-
From the Department of Obstetrics & Gynaecology, University of Cape Town/Groote

Schuur Hospital, Cape Town, South Africa.
Address reprint requests to Lynette Denny, MD, PhD, Department of Obstetrics &
Gynaecology, University of Cape Town/Groote Schuur Hospital, H45, Old Main Building,
Groote Schuur Hospital, Observatory 7925, Cape Town, South Africa; email: lynette.
denny@uct.ac.za.
1092-9118/10/1-10
GLOBAL ADVANCES IN GYNECOLOGIC ONCOLOGY
went post radiation surgery (7 patients), survival was significantly better.9

Systematic inguinofemoral lymphadenectomy comprises
resection of superficial inguinal lymph nodes and deep
femoral nodes. Although current guidelines recommend at
least 6 nodes per groin be removed, systematic lymphadenectomy is associated with substantial morbidity, such as
leg edema, lymphocysts, wound breakdown, and erysipelas.
To avoid these complications, the use of sentinel node
dissection in vulvar cancer has been investigated. Sentinel
nodes were studied in a large, prospective, multicenter study
(Groningen International Study on Sentinel nodes in Vulvar
cancer [GROINSS-V]), which included 403 women with
unifocal vulvar cancer stage I or II, tumor size less than
4 cm, stromal invasion greater than 1 mm, and clinically
negative lymph nodes.10 Lymphadenectomy was not performed in women with sentinel nodenegative cancer. Groin
recurrences occurred in 2.3% of patients, with a median
follow-up of 35 months. Overall disease-specific survival was
97% after 3 years, and morbidity was substantially reduced.
These results are equivalent to results of systematic inguinofemoral lymphadenectomy from an oncologic point of
view.
Vaginal Cancer
Primary cancer of the vagina is rare and accounts for

approximately 2% of all cancers of the female genital tract.
Most women are older than 60 years, and only 10% to 15%
are younger than 50 years. The most common type of vaginal
cancer is squamous carcinoma (80% to 90%) and adenocarcinoma (4% to 10%). Infection with HPV-16 is believed to be
an important etiologic factor. Prognosis is dependent on age,
histologic type, and tumor stage. The optimal treatment is
still controversial because of the rarity of the disease and
lack of appropriate clinical trials. Standard therapy is radiation, either alone or concurrently with platinum-based
chemotherapy. The advantage of radiation is the preservation of the vagina, although the vagina does not always
remain fully functional.
KEY POINTS
The modern management of vulvar cancer is to distinguish between central and lateral lesions and to
adapt surgery and radiation to reduce morbidity and
enhance quality of life.
Management of cervical cancer has not changed substantially in the past 10 to 20 years; however, approaches to cervical cancer prevention have
undergone a re-evaluation at both the primary and
secondary prevention levels.
The surgical staging of endometrial cancers has enabled the use of adjuvant therapies to be tailored to
the individual patient.
Current novel therapies undergoing trials bring hope
to the long stagnant progress in the management of
ovarian cancer.
Quality of life in the treatment of patients with
gynecological malignant tumors should inform all
treatment decisions.
Cervical Cancer
The most noteworthy advances in cervical cancer in the

past 15 years have been the exploration of alternative
approaches to cytologic testing and colposcopy for the prevention of cervical cancer in developing countries. In addition, the approach to screening in developed countries has
changed from conventional cytologic testing to the introduction of liquid-based cytologic and HPV DNA testing. In fact,
many molecular tests to increase the sensitivity and specificity for the detection of cervical cancer precursors are
currently undergoing clinical trials.
The potential utility of HPV DNA testing was shown in a
randomized trial conducted in South Africa.11 This trial of
more than 6,500 unscreened women, aged 35 to 65 years,
showed a substantial reduction of high-grade cervical cancer
precursors in women who screened positive with Hybrid
Capture 2 (Qiagen Inc, Gaithersburg, MD) for high-risk
types of HPV DNA and who underwent treatment with
cryotherapy, and this reduction was sustained for 36
months. After 36 months the decrease in the cumulative
detection of grade 2 or higher cervical intraepithelial neoplasia in the HPV treatment arm compared with the
control (delayed treatment) arm was 1.5% compared with
5.6% (difference, 4.1%; p 0.001). The difference, however,
in the visual inspection with acetic acid (VIA) treatment
group compared with the control arm was significantly less
(3.8% vs. 5.6%; difference, 1.8%; p 0.002). In India,
Sankaranarayanan and colleagues12 screened 131,746
healthy women who were randomly assigned to one of
four groups: screening by HPV testing, cytologic testing,
VIA, or no screening (standard of care). There was a
reduction in the numbers of advanced cancers and cancer
deaths in the HPV treatment group (hazard ratio, 0.47; 95%
CI, 0.32 to 0.69) compared with no reductions in the VIA or
the cytologic testing treatment groups. These studies and
others conducted in developed countries suggest that HPV
DNA testing has the potential to substantially reduce the
incidence of and mortality from cervical cancer. More technologically accessible and affordable tests for HPV DNA
testing are awaited for implementation in low-resource
settings.
In developed countries, HPV DNA testing has been recommended as a primary screen in women older than 30
years, with cytologic testing as a triage for women with
positive test results. Only women with positive results on
both tests would be referred for colposcopy and treatment.13
Other methods of triage include p16-INK4A overexpression
and testing for HPV-E6/7 messenger RNA, which are still
under study.
The trends in terms of cervical cancer treatment include
the following: (1) fertility-sparing surgery in young women
with early-stage disease; (2) laparoscopic radical hysterectomy and node dissection; (3) use of sentinel nodes to
prevent morbidity associated with systematic lymphadenectomy (still under study) and; (4) use of concurrent
cisplatinum-based chemotherapy concurrently with radiation, which has shown substantial improvement in overall
survival and disease-free survival in women with advanced
disease.
Although cervical cancer is a relatively rare disease in
developed countries with functional screening programs,
cervical cancer remains the most common cancer among
331
LYNETTE DENNY
women living in developing countries because of the lack of

effective screening programs. Most women with cancer in
developing countries have advanced and untreatable disease. In addition, palliative care remains out of reach of
most women in developing countries, leaving these women
to die painful and undignified deaths. Cervical cancer is the
one cancer that illustrates the gross inequities of health care
between rich and poor countries.
Endometrial Cancer
Worldwide, endometrial cancer is the most common cancer of the female genital tract and the seventh most common
cause of death from cancer in women in western Europe.
Yearly, approximately 7,406 cases are registered in the
United Kingdom, 88,068 in the European Union, and 40,102
in the United States.14 The median age of cancer occurrence
is in the sixth decade, and more than 90% of cases occur in
women older than 50 years. Approximately 5% of endometrial cancers are associated with Lynch syndrome II (hereditary nonpolyposis colorectal carcinoma syndrome); women
with this syndrome have a 30% to 60% risk of developing
endometrial cancer.
Type 1 endometrial cancer, or endometrioid adenocarcinoma, represents 80% of cancers, with serous carcinomas
being the prototype of type II uterine cancers. Clear cell
cancers are rare and comprise approximately 1% of endometrial adenocarcinomas.
Since 1988, the International Federation of Gynecology
and Obstetrics (FIGO) has recommended surgical staging of
endometrial cancer, which includes systematic pelvic and
para-aortic lymphadenectomy. FIGO adopted a new staging
system in 2009 in which the old FIGO stage 1a and 1B were
amalgamated into stage 1A and the old 1C became stage 1B.
In stage II cancer, the distinction between superficial and
deep stromal invasion was merged as stage II in which
stromal invasion is documented.
The role of systematic lymphadenectomy has been the
subject of numerous studies and much controversy. The
Adjuvant External Beam Radiotherapy in the Treatment of
Endometrial Cancer (ASTEC) trial15 randomized 1,408
women from 85 centers with histologically proven endometrial cancer to standard surgery (704 women with hysterectomy, bilateral salpingo-oophorectomy, peritoneal washings,
and palpation of para-aortic lymph nodes) or to standard
surgery plus lymphadenectomy (704 women). The study
found no evidence of benefit in terms of either overall or
recurrence-free survival for pelvic lymphadenectomy in
women with early endometrial cancer. This study was,
however, criticized for numerous protocol violations, and
some authors thought that it did not adequately assess the
role of effective lymphadenectomy or the role of individualized adjuvant radiotherapy in endometrial cancer.16
The current recommendation for new FIGO stage 1A
cancer is to perform standard surgery only, but for high-risk
cases (i.e., 50% invasion, grade 3 lesions), systemic lymphadenectomy should be performed. The value of this approach
is that approximately 30% of women with high-risk stage I
disease will not require adjuvant whole pelvic irradiation
because their cancer is node negative. Vaginal brachytherapy would still be indicated for women with high-risk
characteristics to reduce local recurrence.17
The value of adjuvant radiation for women with endometrial cancer localized to the uterus is still under scrutiny.
332
External beam whole pelvic radiation has been shown to

reduce the rate of locoregional recurrence in intermediaterisk endometrial cancer; however, a number of trials have
failed to demonstrate that radiation improves overall or
disease-specific survival. Post-Operative Radiotherapy in
Endometrial Cancer 2 (PORTEC-2), a randomized trial
comparing vaginal brachytherapy with external beam radiation, also did not find overall improved survival; however,
quality of life was better in women treated with vaginal
brachytherapy only.17
Ovarian Cancer
The ASIR of ovarian cancer in the European Union in

200814 was 13.4 per 100,000 women, with a mortality rate of
7.6 per 100,000. In the United Kingdom the ASIR was 17.5
per 100,000 women, with a mortality rate of 9.6 per 100 000.
In 2011 in the United States, there were 21,990 new cases of
ovarian cancer diagnosed and 15,640 deaths, making ovarian cancer the second most lethal cancer among women in
the United States.18 Approximately 80% of advanced stage
ovarian cancers have high-grade histologic features, and
these tumors can arise from the fimbrial end of the fallopian
tube, the peritoneal cavity, or the surface epithelium of the
ovary. Efforts to improve on the long-term results of primary
therapy (a combination of maximum effort surgery, including hysterectomy, bilateral salpingo-oophorectomy, infracolic omentectomy, and maximum debulking of all
macroscopic tumor to 2 cm of residual disease, followed by
chemotherapy using a combination of carboplatin and paclitaxel) through the addition of additional cytotoxic agents
have not been successful.
New and effective therapies for ovarian cancer have not
been reported since the introduction of platinum-based
drugs and paclitaxel. Neither these drugs nor second-line
therapies (e.g., gemcitabine) have produced substantial advances in overall survival.
Recently, however, two phase III clinical trials of bevacizumab, an angiogenesis inhibitor and vascular endothelial
growth factorneutralizing monoclonal antibody, have
shown improvements in progression-free survival but not
overall survival. Tumor angiogenesis, a process in which
cancers induce the formation of new blood and lymphatic
vasculature to enable proliferation, invasion, and growth of
metastasis, appears to be essential to disease progression in
women with epithelial ovarian cancer (EOC). Several antiangiogenic agents have been investigated in clinical trials,
most directed against vascular endothelial growth factor, a
central promoter of the initiation of angiogenesis. Two trials
(Gynecologic Oncology Group 0218 and ICON7)19,20 were
conducted in a total of 3,401 women with newly diagnosed
cancers, and the third (OCEANS)21 was conducted in 484
women with platinum-sensitive relapse. Although the data
from these studies are yet to mature, the results indicate a
substantial prolongation of progression-free survival when
administered with standard platinum-based combined chemotherapy followed by continuation of bevacizumab therapy.
Several vascular endothelial growth factor receptor tyrosine kinase inhibitors are under evaluation as either
monotherapy or combination therapy in recurrent ovarian
cancer. One of this class of drugs, cediranib has demonstrated in a phase II trial a clinical benefit rate of 30% and
median progression-free survival rate of 5.2 months.22
GLOBAL ADVANCES IN GYNECOLOGIC ONCOLOGY
A new class of anticancer drugs, known as PARP inhibitors, are also being evaluated. Repair pathways of DNA are
critical for the maintenance of genome integrity and the
response to DNA-damaging chemotherapy. Naturally
occurring or environmentally induced single-stranded
DNA breaks are generally repaired by the enzyme PARP.
In the presence of a PARP inhibitor, single-stranded breaks
may accumulate, leading to double-stranded breaks, usually
repaired through separate molecules involved in the
process known as homologous recombination repair (HRR).
In cells with defective HRR, accumulated double-stranded
breaks generally lead to apoptosis. It has been demonstrated
that EOC in women with hereditary breast-ovarian cancer
syndrome harbor such defects in HRR because of the
germline mutations in BRCA1 and BRCA2. Kaye and colleagues reported on a phase II randomized trial of a PARP
inhibitor known as olaparib compared with placebo in
women with recurrent unselected serous EOC who were in
partial or complete response after the last platinumcontaining chemotherapy regimen.23 The drug was well
tolerated, and results showed a substantially prolonged
progression-free survival with a 65% reduction in progression in the olaparib group. Survival data are not yet available.
Other agents in clinical trials include a therapeutic antifolate receptor antibody (farletuzamab), which may block
folate receptor signal transduction and promote immunogenicity. Folate receptor is relatively absent in normal
tissues and transduces a mitogenic signal on binding to
circulating folate. EC145 is a synthetic agent composed of
folate covalently linked to a highly toxic vinca alkaloid. Once
internalized into the cell, EC145 binds to folate receptor

and the vinca alkaloid is activated.
The use of PEGylation to alter the pharmacokinetics
and dynamics of cytotoxic agents is another advance. A
novel pegylated topoisomerase I inhibitor has demonstrated
single-agent response rates up to 20% for platinum-resistant
EOC in two trials.24,25 Resistance to platinum agents is a
major obstacle to therapeutic outcomes in many patients.
Recent studies suggest that methylation of promotor
regions of genes normally required to induce apoptosis may
be one of the causes of platinum resistance. DNA methyltransferase inhibitors may reverse this process. Two studies
have produced some preliminary evidence that demethylation agents can partially restore sensitivity to platinum
therapy.26,27
Conclusion
The understanding of the natural history of gynecologic

malignant tumors, combined with some well-conducted
randomized clinical trials, has markedly advanced the
treatment of women with gynecologic cancers during the
past 10 to 20 years. The biggest change has occurred in
cervical cancer prevention rather than management of cervical or other gynecologic cancers. Increasingly, targeted
therapies are showing efficacy in ovarian cancers, and for
others, more rational and less mutilating surgery is being
performed. The role of adjuvant therapies in endometrial
cancer still needs to be clarified, but in all cancers of the
female genital tract, individualization of treatment along
with oversight of a multidisciplinary team will most likely
improve outcomes.
Author
Lynette Denny
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
GlaxoSmithKline;
Merck
Research
Funding
GlaxoSmithKline;
Merck
Expert
Testimony
Other
Remuneration
REFERENCES
1. Parkin DM, Bray F. Chapter 2: the burden of HPV-related cancers.
Vaccine. 2006;24 S3:S3/11-S3/25.
2. Toki T, Kurman RJ, Park JS, et al. Probably nonpapillomavirus etiology
of squamous cell carcinoma of the vulva in older women: a clinicopathologic
study using in situ hybridization and polymerase chain reaction. Int J
Gynecol Pathol. 1991;10:107-125.
3. Eva LJ, Ganesan R, Chan KK, et al. Differentiated-type vulval intraepithelial neoplasia has a high-risk association with vulval squamous cell
carcinoma. Int J Gynecol Cancer. 2009;19:741-744.
4. Hacker NF, Leuchter RS, Berek JS, et al. Radical vulvectomy and
bilateral inguinal lymphadenectomy through separate groin incisions. Obstet
Gynecol. 1981;58:574-579.
5. Siller BS, Alvarez RD, Conner WD, et al. T2/3 vulva cancer: a casecontrol study of triple incision versus en bloc radical vulvectomy and inguinal
lymphadenectomy. Gynecol Oncol. 1995;57:335-339.
6. Lin JY, DuBeshter B, Angel C, et al. Morbidity and recurrence with
modifications of radical vulvectomy and groin dissection. Gynecol Oncol.
1992;47:116-121.
7. Woelber L, Choschzick M, Eulenburg C, et al. Prognostic value of
pathological resection margin distance in squamous cell cancer of the vulva.
Ann Surg Oncol. 2011;18:3811-3818.
8. Moore DH, Ali S, Koh W-J, Michael H, et al. A phase II trial of radiation
therapy and weekly cisplatin chemotherapy for the treatment of locallyadvanced squamous cell carcinoma of the vulva: a gynecologic oncology group
study. Gynecol Oncol. 2012;124:529-533.
9. Rogers LJ, Howard B, Van Wijk L, et al. Chemoradiation in Advanced

Vulval Cancer. Int J Gynecol Cancer. 2009;19:745-751.
10. Van der Zee AG, Oonk MH, De Hullu JA, et al. Sentinel node dissection
is safe in the treatment of early-stage vulvar cancer. J Clin Oncol. 2008;26:
884-889.
11. Denny L, Kuhn L, Hu C, et al. Human papillomavirus-based cervical
cancer prevention: long-term results of a randomized screening trial. J Natl
Cancer Inst. 2010;102:1-11.
12. Sankaranarayanan R, Nene BM, Shastri SS, et al. HPV screening for
cervical cancer in rural India. N Engl J Med. 2009;360:1385-1394.
13. Guglielmo R, Meijer CJLM. HPV screening: available data and recommendations for clinical practice. Curr Cancer Ther Rev. 2010;6:104-109.
14. International Agency for Research on Cancer. GLOBOCAN 2008.
http://globocan.iarc.fr. Accessed January 2012.
15. ASTEC study group, Kitchener H, Swart AM, et al. Efficacy of systematic pelvic lymphadenectomy in endometrial cancer (MRC ASTEC trial): a
randomised study. Lancet. 2009;373:125-136.
16. Creasman WT, Mutch DE, Herzog TJ. ASTEC lymphadenectomy and
radiation therapy studies: are conclusions valid? Gynecol Oncol. 2010;116:
293-294.
17. Nout RA, Smit VT, Putter H, et al. Vaginal brachytherapy versus pelvic
external beam radiotherapy for patients with endometrial cancer of highintermediate risk (PORTEC-2): an open-label, non-inferiority, randomised
trial. Lancet. 2010;375:816-823.
18. Siegel R, Ward E, Brawley O, et al. Cancer statistics, 2011; the impact
333
LYNETTE DENNY
deaths. CA Cancer J Clin. 2011;61:212-236.
19. Burger RA, Brady MF, Bookman MA, et al. Phase III trial of bevacizumab (BEV) in the primary treatment of advanced epithelial ovarian cancer
(EOC), primary peritoneal cancer (PPC) or fallopian tube cancer (FTC): a
Gynecologic Oncology Group study. J Clin Oncol. 2010:28;18s (suppl; abstr
LBA1).
20. Perren TS, Pfisterer J, Ledermann J, et al. ICON7: a phase III
Gynaecologic Cancer Intergroup (GCIG) trial of adding bevacizumab to
standard chemotherapy in women with newly diagnosed epithelial ovarian,
primary peritoneal or fallopian tube cancer. Ann Oncol. 2010:21 (suppl 8;
abstr LBA4).
21. Aghajanian C, Finkler NJ, Rutherford T, et al. OCEANS: a randomized
double -blinded, placebo-controlled phase III trial of chemotherapy with or
without bevacizumab (BEV) in patients with platinum-sensitive recurrent
epithelial ovarian (EOC), primary peritoneal (PPC), or fallopian tube cancer
(FTC). J Clin Oncol. 2011:29;TK (suppl; abstr LBA5007).
22. Matulonis UA, Berlin S, Ivy P, et al. Cediranib, an oral inhibitor of
vascular endothelial growth factor receptor kinases, is an active drug in
334
recurrent epithelial ovarian, fallopian tube, and peritoneal cancer. J Clin

Oncol. 2009;27:5601-5606.
23. Ledermann JA, Harter P, Gourley C, et al. Phase II randomized
placebo-controlled study of olaparib (AZD2281) in patients with platinumsensitive relapsed serous ovarian cancer. J Clin Oncol. 2011;29; TK (suppl;
abstr 5003).
24. Vergote IB, Micha JP, Pippitt CH Jr, et al. Phase II study of NKTR-102
in women with platinum-resistant/refractory ovarian cancer. J Clin Oncol.
2010;28:15s (suppl; abstr 5013).
25. Garcia AA, Vergote IB, Micha JP, et al. The role of NKTR-102 in
women with platinum resistant/refractory ovarian cancer and failure on
pegylated liposomal doxorubicin (PLD). J Clin Oncol. 2011;29 (suppl; abstr
5047).
26. Fu S, Hu W, Iyer R, et al. Phase 1b-2a study to reverse platinum
resistance through use of a hypomethylating agent, azacitidine, in patients
with platinum-resistant or platinum-refractory epithelial ovarian cancer.
Cancer. 2011;117:1661-1669.
27. Matei D, Shen C, Fang F, et al. A phase II study of decitabine and
carboplatin in recurrent platinum (Pt)-resistant ovarian cancer (OC). J Clin
Oncol. 2011;29;TK (suppl; abstr 5011).
The European Society of Gynaecological

Oncology: Update on Objectives and
Educational and Research Activities
By Renata Brantnerova, Ranjit Manchanda, MD, and Nicoletta Colombo, MD
Overview: The European Society of Gynaecological Oncology

(ESGO) is the principal European society contributing to the
study, prevention, and treatment of gynecologic cancers.
Founded in 1983, ESGO has more than 1,300 members in more
than 40 European countries and worldwide who benefit from
ESGOs innovative education and research initiatives and
networking opportunities. ESGO objectives have been recently
identified through a strategic planning process and include
education, care, research, collaboration, awareness, and sustainability. As a leading gynecologic oncology society, ESGO
holds biennial meetings where experts meet to discuss latest
advances in gynecologic treatment and care. The 17 th International Meeting of ESGO (ESGO 17) proved to be a resounding success, with 2,700 delegates and speakers who gathered
from around the world in the cultured city of Milan, Italy. The
structure of the congress included keynote lectures, debates,
OUNDED IN 1983 in Italy as a forum for professionals

dedicated to the care of women with gynecologic cancers, the European Society of Gynaecological Oncology
(ESGO) will soon celebrate 30 years of existence.
The mission of ESGO is to improve the health and wellbeing of European women with gynecologic (genital and
breast) cancers through prevention, excellence in care, and
high-quality research and education.
Looking back over the last decade, our membership has
increased rapidly from being an exclusive club of about 150
members in 2002 into a truly pan-European and international community of multidisciplinary professionals: ESGO
had 1,330 members in 2011, with an additional 240 new
members joining in the last quarter of the year. ESGO is a
young society: In 2011, 37% of members declared to be
younger than age 45 and 25% of all members were part of
the European Network of Young Gynaecologic Oncologists
(ENYGO), which is the ESGOs European platform for trainees and professionals younger than age 40. ESGO cares about
members from all regions of Europe: Members from lowerincome countries form 27% of ESGO active membership.
European Challenges in Gynecologic Oncology
There are approximately 630,000 women diagnosed with

gynecologic cancers in Europe every year (including breast
cancer). Patterns of incidence and mortality are different
from region to region, and there are continuing inequalities
both among and within European countries in terms of
cancer risk, detection, and treatment. The estimated incidence for cervical cancer is 54,517 new cases/year with
24,874 annual deaths related to this disease.1 Cervical
cancer is the seventh most common cancer among European
women of all ages, and the second most common cancer in
women ages 15 to 39 years.2 Rates are highest among
women in countries of the former Soviet Union, who have
twice the risk of dying as a result of cervical cancer as
Western European women. Women in Romania are 11 times
more likely than Finnish women to die as a result of cervical
cancer.3
Ovarian cancer is the fifth most common cancer among
state-of-the art sessions, and focused sunrise sessions, together with oral and poster presentations and satellite symposia sponsored by pharmaceutical companies. For the first
time, during ESGO 17 the Society organized a seminar for
European patient groups with an interest in gynecologic
cancers with the aim of facilitating different patientrelated
activities across Europe. Moreover, The European Network of
Young Gynaecologic Oncologists (ENYGO), the European Network of Gynaecologic Oncology Trial Groups (ENGOT), and
the European Network of Translational Research in Gynaecological Oncology (ENTRIGO) had their own section during
ESGO 17. ESGO also holds numerous workshops throughout
the calendar year and provides clinical and research grants,
online educational materials, webcasts, and numerous networking opportunities
women in Europe, with more than 65,000 new cases and

41,000 deaths annually. Rates are highest in Eastern and
Northern Europe and lowest in Southern Europe.1
Uterine cancer is the fourth most common female cancer
in Europe, with 88,000 new cases/year and 22,000 deaths.
The highest incidence is observed in Eastern and Northern
Europe, and lowest in southern regions.1
Breast cancer is the most common cancer in women in
Europe, with over 425,000 new cases/year. Breast cancer is
responsible for more deaths among European women than
any other cancer, but survival has improved thanks to
screening and better treatments. However, survival is
higher in Western than in Eastern Europe, with further
differences within Western Europe: lower in the United
Kingdom and Denmark than in Finland, Sweden, France,
Italy, and the Netherlands.2 These statistics outline how
there are unacceptable inequities in access to information,
prevention, screening, treatment and care across Europe.
Patterns of incidence and mortality are different from region
to region, often reflecting a lack of awareness and access to
appropriate services. ESGO is committed to help women
with gynecological cancer across Europe to obtain accurate,
reliable and timely information about their disease, to understand treatment options and to have access to the best
possible care.
ESGO Objectives
Considering the changing external and internal environment, as well as global and medical trends affecting the field
of gynecologic oncology, ESGO conducted strategic planning
From the European Society of Gynaecological Oncology, the UCL Elizabeth Garrett
Anderson Institute for Womens Health, and the University of Milan Bicocca, European
Institute of Oncology.
Address reprint requests to Nicoletta Colombo, MD, University of Milan Bicocca,
European Institute of Oncology, Via Ripamonti 435, 20141 Milan, Italy; email:
Nicoletta.colombo@ieo.it.
1092-9118/10/1-10
335
BRANTNEROVA, MANCHANDA, AND COLOMBO
during 2011. The aims were to analyze the current positioning of the Society; define its future as Europes leader in
education, research, and care in the field of gynecologic
cancers; and ensure ESGOs continuous development as an
active resource of knowledge, information, collaboration and
networking for its members and all professionals in the field.
Having in mind the ESGO mission to improve the health
and well-being of European women with genital and breast
cancers, core fields of interest were defined and general
objectives were established: 1) educationto provide high
quality educational activities and improve training in
gynaecologic oncology in Europe; 2) careto establish multidisciplinary standards of care and act as the European
authority in the field; 3) researchto support platforms for
collaborative clinical, translational, and basic research in
gynecologic cancers; 4) collaborationto promote collaboration among scientific societies, health care professionals,
patient organizations, business, industry, and governmental
bodies; 5) awarenessto raise public and governmental
awareness of gynecologic cancers and their prevention and
treatment; and 6) sustainabilityto ensure ESGOs sustainability and positioning as the leading European professional
medical society in its field.
Education and Outreach
One of ESGOs primary missions is to provide high-quality

educational activities and improve training in gynecologic
oncology in Europe.
The highlight of the educational activities is organizing
the ESGO international bi-annual meeting, which has a
main focus on education, but with an ongoing trend to
present science and research. The ESGO 17 proved to be a
resounding success, with 2,700 delegates and speakers who
gathered from around the world in the cultured city of
KEY POINTS
336
The European Society of Gynaecological Oncology

(ESGO) is the principal European society contributing to the study, prevention, and treatment of gynecological cancers.
ESGO objectives were recently revisited through a
strategic planning process and include education,
care, research, collaboration, awareness, and sustainability.
The highlight of the educational activities is ESGOs
international bi-annual meeting, which has a main
focus on education, but with an ongoing trend to
present science and research.
During ESGO 17 the Society organized a seminar for
European patient groups with an interest in gynecologic cancers with the aim of facilitating different
patient-related activities across Europe.
The European Network of Young Gynaecologic Oncologists (ENYGO) is a network for young physicians
and trainees in gynecologic oncology and related
subspecialties that represents the needs and aspirations of all trainees involved in the study, prevention,
and treatment of gynecologic cancers.
Milan, Italy. The structure of the congress included keynote

lectures, debates, state-of-the art sessions, and focused sunrise sessions, together with oral and poster presentations
and satellite symposia sponsored by pharmaceutical companies.
The keynote lectures were presented by Prof. Umberto
Veronesi and Prof. Rene Bernards. Prof. Veronesi focused on
the progress in breast cancer management, highlighting his
lifetime achievements on the quality of life QOL of women
with breast cancer. By proposing a new paradigm from
maximum tolerable treatment to minimum effective treatment, he was able to completely modify the care of patients
with breast cancer: going from the anatomic concept of
cancer spread to the biologic one, the therapy for these
patients changed from the mutilated radical mastectomy to
the conservative quadrantectomy, from the complete lymph
node dissection to the sentinel-node biopsy and more recently from the lengthy external radiotherapy to the intraoperative radiotherapy. All these achievements led to a
substantial improvement in the QOL of women with breast
cancer.
Prof. Bernards highlighted that fact that only 22% of
patients with cancer derive substantial benefit from their
treatment, mainly because of the intra- and intertumor
heterogeneity. To answer the basic questions of whom to
treat and in what way, a more sophisticated approach is
needed. Molecular diagnostics has the potential to replace
the microscope in identifying prognostic and predictive
markers that should guide a more personalized cancer
treatment.
One of the main highlights of the plenary session included
the presentation of the results of the secondary end point
QOL of an international study which randomly assigned
patients with recurrent ovarian cancer to receive maintenance olaparib or placebo.4 There was no statistically significant difference in the QOL scales between treatment arms.
The median time to worsening QOL was 2.8 months for
olaparib compared with 3.7 months for placebo, and 70%
of patients receiving olaparib experienced nausea compared
with 43% receiveing placebo (hazard ratio [HR] 2.18).
Results of a phase II study combining olaparib with
pegylated liposomal doxorubicin (PLD) in patients with
advanced solid tumors (28 ovarian cancer, 13 breast cancer,
three others) who had received at least three prior chemotherapy regimens were presented by Del Conte and colleagues during the ovarian oral session.5 The maximum
tolerated dose was not reached. Common adverse events
(AEs) of any grade included stomatitis (73%), nausea (71%),
asthenia (57%), pyrexia (43%), anorexia (41%), vomiting
(41%), cough (39%), neutropenia (27%) and palmar-plantar
erythrodysesthesia (25%). Serious AEs occurred in 27% of
patients, with pneumonitis being most common (n 3). Two
patient deaths were considered possibly related to study
therapy. Treatment with olaparib and PLD resulted in two
(5%) complete responses (CRs), 11 (25%) partial responses
(PRs), and 13 (30%) incidents of stable disease (SD). Of
the 13 patients who achieved a PR or better, 11 were BRCA
mutation positive and two were BRCA mutation negative or
unknown.
Dr. Fischerova received the award for the best poster
presentation for her work, The Role of Ultrasound in
Planning Fertility Sparing Surgery and Individual Treatment in Early Stage Cervical Cancer. She analyzed the
ESGO EDUCATIONAL AND RESEARCH ACTIVITIES
accuracy of ultrasound in predicting tumor size, parametrial

and lymph node involvement, distance between the tumor
and the internal cervical os, and depth of stromal invasion
in 83 patients undergoing radical hysterectomy for stage IB
cervical cancer. Overall sensitivity of ultrasound imaging in
the evaluation of parameters required for fertility-sparing
surgery was 68.4%. The lower sensitivity mainly resulted
from lack of detection of lymph-node involvement.6
The oral translation research award was assigned to
Dr. Lai7 for her presentation, Human Ovarian Cancer
Stem Cells Can Be Efficiently Killed by Gamma Delta
T-Lymphocytes. A subset of cancer stem cells (CSCs) from
SKOV3 cell line were cocultured with gamma delta T cells.
Their proliferation rate decreased to 40% after 48 hours of
exposure. The gamma delta T cells increased the sensitivity
of SKOV3 CSCs to chemotherapeutic drugs. Dr. Lai also
found that gamma delta T cells induced G2/M phase cellcycle arrest and subsequent apoptosis in SKOV3 CSCs.
Moreover, xenograft mouse models demonstrate that
gamma delta T cells dramatically reduced the tumor burden
in vivo.
Finally, the prize for the best oral presentation in the
young professionals session was awarded to A Risk Model
for Secondary Cytoreductive Surgery in Recurrent Ovarian
Cancer: An Evidence-Based Proposal for Patient Selection.8
Individual data on 1,075 patients with recurrent ovarian
cancer undergoing secondary cytoreductive surgery were
pooled and analyzed. Complete secondary cytoreduction was
associated with six variables: International Federation of
Gynecology and Obstetrics (FIGO) stage (odds ratio [OR]
1.32), residual disease after primary cytoreduction (OR
1.69), progression-free interval (OR 2.27), Eastern Cooperative Oncology Group (ECOG) performance status (OR
2.23), cancer antigen 125 (OR 1.85), and ascites at
recurrence (OR 2.79). These variables were entered into
the risk model and assigned scores to identify low- and
high-risk categories. In external validation, the sensitivity
and specificity were 83.3% and 57.6%, respectively.
ESGO also provides online educational resources available on ESGO website: 1) webcasts of ESGO meeting sessions and e-Posters that extend the life of ESGO meetings
and bring them to members who could not attend; 2) an
online Video Library that provides a unique opportunity for
visual education; 3) e-Series, which are webcast presentations by internationally renowned speakers; 4) educational
DVDs; and 5) the Textbook of Gynaecological Oncology, a
unique publication that compiles articles from world-famous
authors in the field of gynecologic oncology, including surgery, radiotherapy, chemotherapy and imaging specialties.
Care
Committed to the cause of excellence in care, ESGO aims

to establish multidisciplinary standards of care and acts
as the European authority in the field. Therefore, together
with the European training standards, ESGO set the
multidisciplinary standards of care and provides supervision for certified training. Within the ESGO-European
Board and College of Obstetrics and Gynaecology (EBCOG)
Hospital Accreditation Program, ESGO has provided hospital recognition of excellence in training and care while
certifying training centers across Europe. Trainees trained
at ESGO-accredited centers become ESGO-certified gyneco-
logic oncologists. Currently, more than 35 European hospitals and training centers have been accredited.
Research
One of ESGOs ambitious goals is to become the voice of

European research in gynecologic cancers and the coordinating body of European clinical and translational research,
with the aim to improve the prediction, prevention, detection, and treatment of women-specific cancers.
In 2007, at ESGOs congress in Berlin, Germany, the
European Network of Gynaecological Oncological Trial
Groups (ENGOT) was founded. Currently, the network
consists of 17 European cooperative groups and national and
regional clinical trial units. A consensus statement on requirements for trials between academic groups and pharmaceutical companies has been produced and a roadmap for
clinical trials in Europe has been designed.9
To complement these existing activities and make further
advances, in 2009 ESGO formed the European Network
of Translational Research in Gynaecological Oncology
(ENTRIGO), a platform that brings together scientists,
clinical academics, and patients. In 2011, ESGO organized
the first ENTRIGO Translational Research Workshop,
which brought together the leading European scientists
in the fields of epidemiology, molecular carcinogenesis, personalized medicine and molecular imaging, pathology, and
cancer genetics and epigenetics. In a long-term perspective,
the aim of ENGOT is to run large European Unionfunded,
trans-European research programs.
Collaboration and Networking
The goal of ESGO is to be a platform promoting and

facilitating collaboration among scientific societies, health
care professionals, patient organizations, business, industry, and governmental bodies. Within this aim, ESGOs
biennial meetings give the best networking opportunities to
meet and discuss women-specific cancers, not only to ESGO
members but also to the wide community of medical professionals interested in these cancers.
Networking is rated by members as one of the most
important appeals of initiating or renewing ESGO membership. Task forces have been recently established within the
ESGO community to give members opportunities to exchange ideas and debate a wide range of related topics (e.g.,
cancer in pregnancy, fertility preservation, psycho-oncology,
surgery harmonization). Members are welcome to be involved in the association, use the ESGO newsletter and
website to promote their open projects, and search for
collaboration with their colleagues.
ESGO pays special attention to cooperation with sister
societies, with regional and international societies of gynecologic cancer and mainly with the International Society of
Gynecologic Cancer (IGCS), Society of Gynecologic Oncology
(SGO), and Asian Society of Gynecologic cancer (AGCS).
ESGO is a member and representative of gynecologic oncologic subspecialty of EBCOG and a member of European
Cancer Organisation (ECCO) which is the umbrella organization of all European cancer-related societies.
Awareness
One of ESGOs objectives is to raise public and governmental awareness of gynecologic cancers in Europe, and of
their prevention and treatment. The incidence of and mor-
337
BRANTNEROVA, MANCHANDA, AND COLOMBO
tality from gynecologic cancers are increasing in Europe,

and all women especially older womenare at risk. Trustworthy public and patient information and education are
lacking, especially in nonEnglishspeaking countries.
As the leading European professional society related to
cancer, ESGO shares the goal of patient groups to help
women living with gynecologic cancers to obtain accurate,
reliable, and timely information about their disease, to
understand treatment options and to have access to the best
possible care. For the first time, during ESCO 17 the Society
organized a seminar for European patient groups with an
interest in gynecologic cancers. Fifty participants, including
38 representatives of patient organizations, attended.
ESGO shares common goals with patient organizations,
and wishes to become a real facilitator of different patientrelated activities across Europe. The seminar at ESGO 17
was only the start of the partnership, and ESGO is committed to provide a continuing platform for patient organizations to meet and work together. Ten recommendations
coming out of this patient seminar have been approved by
the ESGO Council, and a small committee formed to take
these forward in 2012.
Focus on the Younger Generation: ENYGO
ENYGO is a network for young professionals and trainees

in gynecologic oncology and related subspecialties. ENYGO
was created in 2009 and is supported in its activities by
ESGO. ENYGO is the principal European network representing the needs and aspirations of all trainees involved in
the study, prevention, and treatment of gynecological cancers. It serves as a forum for promoting scientific and social
interaction, discussion, debate, and exchange of ideas and
views among trainees. ENYGO currently has approximately
400 members from 40 countries across Europe, with each
country having a national representative. ENYGO is represented on the ESGO council by its president and supported
by a vice president and an executive group.
ENYGO Activities at ESGO 17
ENYGO had its own session during the 17th ESGO

conference. Details of outcomes from this workshop include: 1) an introduction and retrospection about ENYGO by
Dr. Michaela Bossart and past-president Dr. Boris Vranes;
2) presentation of the initial results of the first Europe-wide
survey on training experiences of European gynecologic
oncology trainees by Dr. Ranjit Manchanda, which clearly
highlighted the differences in training across countries and
the importance and need for accredited training for trainees;
3) an update on Teaching the Teachers workshop presented
by Dr. Jurgen Piek, which highlighted the importance of
training teaching according to modern concepts and mechanisms such as the CanMEDS framework and specific, measurable, attainable, rewarding, time-bound (SMART) rules
to facilitate the phenomenon of lifelong learning; 4) a presentation on training systems in Europe by Dr. Murat
Gultekin; and 5) the launch and hard-copy/CD-ROM distribution of the second edition of the Textbook of Gynaecological
Oncology, which was a big success and greatly appreciated
by all trainees.
Subsequently, Dr. Ranjit Manchanda coordinated an Internet survey, to which 154 members responded. This
showed that surgical anatomy and laparoscopic surgery in
gynecologic oncology were the topics most members would
like to see covered by workshops. On the basis of these
results, ENYGO is facilitating the organization of a combined 2-day workshop focusing on surgical anatomy and
laparoscopic surgery in London in September 2012. After
the successful Teaching the Teachers (TTT) workshop in
Amsterdam, the Netherlands, in May 2011, the next TTT
workshop is proposed to be held in Macedonia in June 2012.
ENYGO will also help facilitate and organize a young
gynecologic oncologist session (YDSYoung Doctors Session) at the upcoming IGCS conference in Vancouver,
Canada.
Author
Renata Brantnerova*
Ranjit Manchanda
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Abbott
Laboratories
Nicoletta Colombo
Cancer
Research UK;
The Eve Appeal
REFERENCES
1. International Agency for Research on Cancer. Globocan 2008 Fast Stats
Factsheet: Europe. http://globocan.iarc.fr/factsheet.asp. Accessed February
24, 2012.
2. Ferlay J, Parkin DM, Steliarova-Foucher E. Estimates of cancer incidence and mortality in Europe in 2008. Eur J Cancer. 2010;46:765-781.
3. Arbyn M, Raifu AO, Autier P, et al. Burden of cervical cancer in Europe:
estimates for 2004. Ann Oncol. 2007;18:1708-1715.
4. Ledermann J, Harter P, Gourley C, et al. Phase 2randomized placebocontrolled study of olaparib(AZD2281) in patients with platinum-sensitive
relapsed serous ovarian cancer (PSR SOC) Int J Gynecol Cancer. 2011;21:S13.
5. Del Conte G, Sessa C, von Moos R, et al. Antitumor activity of olaparib
(AZD2281) and liposomal doxorubicin in previously treated ovarian cancer
patients. Int J Gynecol Cancer. 2011;21:S20.
338
6. Fischerova D, Freitag P, Zikan M, et al. The role of ultrasound in

planning fertility sparing surgery and individual treatment in early stage
cervical cancer. Int J Gynecol Cancer. 2011;21:S107.
7. Lai D, Wang F, Chen Y at al. Human ovarian cancer stem cells can be
efficiently killed by gamma dekta T-lymphocytes. Int J Gynecol Cancer.
2011;21:S45.
8. Tian WJ, Chi DS, Sehouli J, et al. A risk Model for secondary cytoreductive surgery in recurrent ovarian cancer: an evidence-based proposal for
patient selection. Int J Gynecol Cancer. 2011;21:S55.
9. Vergote I, Pujade-Lauraine E, Pignata S, et al. European Network of
Gynaecological Oncological Trial Groups requirements for trials between
academic groups and pharmaceutical companies. Int J Gynecol Cancer.
2010;20:476-478.
UPFRONT TREATMENT OF OVARIAN CANCER:

INTERNATIONAL CONSENSUS AND VARIATION
CHAIR
Gordon J. S. Rustin, MD, MSc
Mount Vernon Cancer Centre
Middlesex, United Kingdom
SPEAKERS
Deborah K. Armstrong, MD
Baltimore, MD
Keiichi Fujiwara, MD, PhD
Saitama Medical University International Medical Center
Hidaka, Japan
Biologicals in the Upfront Treatment of

Ovarian Cancer: Focus on Bevacizumab and
Poly (ADP-Ribose) Polymerase Inhibitors
By Rene Roux, MBBS, Martin Zweifel, MD, PhD, and Gordon J. S. Rustin, MD, MSc
Overview: Biologicals have made a major impact in the

management of several cancers, but have hitherto had a
negligible impact in ovarian cancer. Fortunately, ovarian cancer has been much more sensitive to cytotoxic chemotherapy
than many cancers, so treatments were still available. However, improvements are required as more than 80% of patients
who present with advanced ovarian cancer eventually will die
as a result of their disease.
NGIOGENESIS IS particularly relevant in ovarian

cancer.1,2 Vascular endothelial growth factor (VEGF)
is important in both ovarian physiology and pathology.
VEGF-A levels change cyclically with the menstrual cycle
and blockage of the pathway can suppress ovulation. In
ovarian cancer, intratumoral VEGF and VEGF receptor
(VEGFR) 2 expression and the carriage of VEGF gene
polymorphisms associated with increased VEGF excretion
are independent poor prognostic factors. Increased levels of
VEGF have been shown to be an independent prognostic
indicator of poor survival in early and advanced ovarian
disease. Tumor VEGF overexpression is associated with
tumor angiogenesis, malignant progression and metastasis,
ascites formation, and early recurrence and death as a result
of disease. In animal models, blockage of VEGF slows tumor
growth and inhibits ascites formation.
The most successful antiangiogenic agent developed to
date is bevacizumab, a humanized monoclonal antibody
which inhibits the binding of VEGF-A to its receptors,
VEGFR1 and VEGFR2. This inhibits the formation of new
tumor vessels and causes the remaining tumor vasculature
to regress and normalize which prevents tumor growth and
metastasis. In contrast to its activity in most other malignancies, bevacizumab has shown single-agent activity in
ovarian cancer. In phase II studies it demonstrated response
rates of 15.9% to 21%, a 6-month progression-free survival
(PFS) of 28% to 40% and an overall survival (OS) of between
10.7 and 16.9 months.3,4
Bevacizumab As First-Line Treatment in
Ovarian Cancer
Two pivotal studies using bevacizumab in the first-line

management of ovarian cancer have recently been reported:
ICON7 and GOG-0218,5,6 with details in Table 1. Both trials
showed similar improvement in PFS, with the ICON7 trial
showing a significant improvement in OS in a high-risk
subgroup (International Federation of Gynecology and Obstetrics [FIGO] stage III suboptimal debulked and FIGO
stage IV), of 36.6 compared with 28.8 months.5 It could be
postulated that an antivascular approach will be more
effective against larger tumors that require an additional
blood supply than microscopic deposits remaining after
optimal surgery. Major differences between the two trials
included a difference in dose and duration of bevacizumab.
GOG218 was a three-arm placebo-controlled trial whereas
ICON7 had two arms with no placebo and included some
stage I/II patients. Only in ICON7 will there be a reliable OS
340
The antiangiogenic antibody bevacizumab and the poly

(ADP-ribose) polymerase (PARP) inhibitor olaparib have recently been shown to improve progression-free survival of
patients with ovarian cancer with better hazard ratios in
certain groups than have been seen previously.
end point because there was virtually no cross-over of

patients between the arms.
Several other trials investigating the role of bevacizumab
in first-line treatment of ovarian cancer are currently recruiting patients. The phase IV ROSiA trial (NCT01239732)
is investigating the effect of adding 3-weekly bevacizumab
15 mg/kg to paclitaxel 175 mg/m2 3-weekly or 80 mg/m2
weekly and carboplatin area under the curve (AUC)
3-weekly, followed by bevacizumab maintenance 3-weekly
up to a total of 36 cycles in patients with previously untreated ovarian cancer. The four-arm phase III mEOC trial
(GOG-0241; NCT01081262) investigates the role of bevacizumab 15 mg/kg once every 3 weeks added to carboplatin
and paclitaxel or to oxaliplatin and capecitabine in metastatic mucinous ovarian cancer.
GOG-0252, a phase III three-arm trial (NCT00951496)
investigating intravenous vs. intraperitoneal paclitaxel and
platinum with the addition of bevacizumab 15 mg/kg in all
three arms, including maintenance treatment for 16 cycles
in previously untreated, optimally debulked stage III ovarian cancer. This will be the first trial investigating the role
of bevacizumab in combination with intraperitoneal chemotherapy.
Two trials are exploring the addition of bevacizumab to
dose-dense chemotherapy but unfortunately no trials are
comparing dose-dense with standard chemotherapy plus
bevacizumab. The improvement in both PFS (hazard ratio
[HR] 0.71) and OS (HR 0.75) from dose-dense paclitaxel
plus 3-weekly carboplatin compared with standard 3-weekly
paclitaxel and carboplatin compares favorably with the improvement seen with addition of bevacizumab.7 OCTAVIA is
a single-arm phase II trial of once every 3 weeks cycles of
treatment with bevacizumab 7.5 mg/kg added to paclitaxel
80 mg/m2 on days 1, 8, and 15 and carboplatin on day 1 of
each cycle. Bevacizumab may then continue to be administered as monotherapy until disease progression. Safety
results from the trial have recently been reported showing
that the addition of bevacizumab is well tolerated.8 GOG-
From the Mount Vernon Cancer Centre, Northwood, Middlesex, United Kingdom.
Address reprint requests to Gordon J.S. Rustin, MD, MSc, Department of Medical
Oncology, Mount Vernon Cancer Centre, Northwood, Middlesex HA6 2RN, UK; email:
grustin@nhs.net.
1092-9118/10/110
ANTIANGIOGENICS AND PARP INHIBITORS IN OVARIAN CANCER

Table 1. Summarized Results of the Three Phase III trials of Bevacizumab in Ovarian Cancer
Overall
No. of
Median Survival
Response Rate
Bevacizumab
Progression
Death
Trial
Patients
Setting
Trial Arms
GOG-02186
1873
First
(CP 6 Pl 5)
line
(mg/kg)
Progression
%
15
95% CI
HR
95% CI
HR
3 Pl 17
First
0.908* 0.795 to
.16*
Free
Overall
10.3
39.3
1.036 0.827 to .76
11.2
38.7
14.1
39.7
1.297
(CP 6 BEV5)
0.717* 0.625 to .001*
3 BEV 17
0.824
CP 6
reported
1.040
1528
95% CI
Not
CP 6 BEV 5
ICON75
(months)
Dose
7.5
48
1.152
0.87* 0.77 to
line
0.915 0.727 to .45
.04
0.85
0.99
(CPBEV) 6
67
3 BEV 12
11% to
.001
0.73* 0.60 to
(difference 28%
0.69 to
.11
1.04
.001
0.93
0.64
0.48 to
.002
0.85
22.4
Not yet
24.1
reached
14.5
28.8
18.1
36.6
19)
OCEANS9
484
Second
line,
platinum
(CGPl) 6
15
57
3 Pl 21
51% to
0.48
64%
0.39 to
.0001
0.61
0.751 0.537 to .094
8.4
1.052
Not yet
reached
or PD
sensitive (CGBEV) 6
3 BEV 21
78
73% to
.0001
12.4
84%
or PD
Abbreviations: BEV, bevacizumab; C, carboplatin; G, gemcitabine; HR, hazard ratio; P, paclitaxel; Pl, placebo.
* HR for progression or death.
All patients.
Restricted mean values.
High-risk patients.
0262 is a phase III trial of 3-weekly paclitaxel/carboplatin

compared with dose-dense weekly paclitaxel/3-weekly carboplatin with optional addition of bevacizumab 15 mg/kg in
both arms.
Bevacizumab As Second-Line Treatment in
Ovarian Cancer
The OCEANS trial was a randomized, double-blinded,

placebo-controlled phase III trial of carboplatin AUC 4
on day 1 and gemcitabine 1000 mg/m2 on days 1 and 8,
3-weekly with or without bevacizumab 15 mg/kg in patients
KEY POINTS
Bevacizumab increases progression-free survival

(PFS) by 2 to 4 months when added to first- or
second-line chemotherapy and continued as maintenance treatment in ovarian cancer.
Bevacizumab 7.5 mg/kg in a subgroup of patients
with stage III suboptimally debulked and stage IV
disease improved survival by 8 months.
Bevacizumab should be considered for patients who
are at high risk of having a short PFS because the
bevacizumab might delay symptoms and these patients with platinum-resistant disease are unlikely
to have an opportunity to receive bevacizumab as
part of second-line treatment.
Poly (ADP-ribose) polymerase (PARP) inhibitors have
shown response rates of up to 69% in early-phase
trials of a patient population with recurrent ovarian
cancer and enriched for patients with BRCA1/2
mutations.
with platinum-sensitive recurrent disease. Preliminary data

were presented at ASCO in 2011.9 PFS was significantly
increased in the bevacizumab arm (12.4 vs. 8.4 months).
The AURELIA (NCT00976911) trial, which reports this
year, evaluates the impact of adding bevacizumab (10 mg/kg
every 2 weeks or 15 mg/kg every 3 weeks) to either dosedense paclitaxel 80 mg/m2 weekly or topotecan 4 mg/m2 on
days 1, 8, and 15 of each 4-week cycle (or 1.25 mg/m2 on day
1 through 5 of each 3-week cycle) or liposomal doxorubicin
40 mg/m2 every 4 weeks. Only patients with platinumresistant ovarian cancer are eligible.
The GOG-0213 (NCT0056585) trial is investigating the
role of adding bevacizumab 15 mg/kg to six cycles of carboplatin AUC 5 and paclitaxel 175 mg/m2, followed by 3-weekly
bevacizumab until disease progression in platinum-sensitive
recurrent ovarian cancer. Patients are allowed to have
received prior bevacizumab, and this trial is also evaluating
the role of secondary cytoreductive surgery.10
Antiangiogenic Small-Molecule Tyrosine
Kinase Inhibitors
Randomized trials of three antiangiogenic tyrosine kinase

inhibitors in patients with ovarian cancer should present
results within the next 2 years.2 The AGO-Ovar 12/LUMEOvar 1 trial is evaluating intedanib in combination with
3-weekly paclitaxel and carboplatin, followed by intedanib
maintenance up to a total of 120 weeks in patients with
stage IIB to IV epithelial ovarian, fallopian tube, or primary
peritoneal cancer after prior tumor-debulking surgery. Pazopanib is being studied as maintenance therapy for up to 2
years, in a randomized, two-arm, placebo controlled, phase
III study in women with ovarian, fallopian tube, or primary
peritoneal cancer that has not progressed after completing
first-line chemotherapy for advanced ovarian cancer. The
ICON6 trial is investigating carboplatin and paclitaxel che-
341
ROUX, ZWEIFEL, AND RUSTIN

Table 2. Unresolved Questions Regarding the Use of Bevacizumab in Ovarian Cancer
Unresolved Question
Option 1/Comment
Option 2/Comment
What dose?
In combination or alone?
15 mg/kg 3 weekly licensed

With chemotherapy as improves response rate
For what duration?

Should evidence of progression
be sought?
Which first-line patients?
Which line of therapy?
Until progression
CA-125 every 6 weeks, and or CT every 3 months
will shorten therapy
All stage 4 and stage 3 with suboptimal surgery
First line
7.5 mg/kg 3 weekly probably as effective and cheaper

Only as maintenance (only addition of maintenance improved
PFS in GOG218)
For defined period-22 cycles
No CA-125 or scans will prolong bevacizumab and delay next
line of therapy
Try to select those likely to have short remission
Relapse
Abbreviations: CA, cancer antigen; CT, computed tomography; GOG, Gynecologic Oncology Group; PFS, progression-free survival.
motherapy with or without concurrent cediranib and also

the effect of maintenance cediranib in platinum-sensitive
relapsed ovarian cancer.
Reasons for Adding a New Treatment
There are many unresolved questions as to the optimal

use of bevacizumab in patients with ovarian cancer. It is a
very expensive therapy, so it is important to consider the
questions in Table 2 and the issues that follow when deciding whether to use this new drug.
Improved Survival
Survival curves from the GOG-0218 and ICON7 trials are

still falling, so it seems unlikely that there will be more
long-term survivors or cures than with cytotoxic chemotherapy alone. Results from the ICON7 trial show that there is
no improved survival for low-risk patients, but an almost
8-months-longer median survival for patients at high risk
for relapse. If the survival benefit from first-line use is
unclear and OCEANS shows improved survival in patients
after first relapse, a case could be made for reserving it for
relapse therapy.
Increased Response Rate
ICON7 and OCEANS both demonstrated an increased

response rate. A tumor response is likely to lead to resolution of cancer-related symptoms. This is of most relevance
to patients who have experienced relapse because most
patients receiving first-line treatment will have minimal
symptoms from their cancer after surgery. Patients who
do have cancer-related symptoms are usually candidates
now for neoadjuvant therapy. However, administering bevacizumab in this setting might raise concerns regarding
impaired wound healing after subsequent surgery.
Delaying Relapse
It is debatable whether the 2- to 4-month delay in PFS

justifies the ongoing treatment with 3-weekly infusions,
especially in view of the results from OCEANS showing that
the addition of bevacizumab to second-line treatment seems
to also delay PFS.
Less Toxic than Standard Therapy
Bevacizumab has been shown to add to toxicity of standard chemotherapy, especially hypertension, bleeding,
thromboembolic events, and gastric perforations. Administering bevacizumab as maintenance therapy reduces time
without therapy.
342
Rationale for PARP Inhibitors
Poly (ADP-ribose) polymerase (PARP), an enzyme discovered almost 50 years ago,11 is crucial for the repair of
single-strand DNA breaks (SSBs) via the base excision
repair (BER) pathway. Although it has no direct effect on
double-strand DNA breaks (DSBs),12 when a DNA replication fork comes across a persistent SSB (i.e., if PARP was
inhibited), the fork could collapse or form a DSB.13 Tumor
cells deficient in BRCA1/2 are unable to repair these DSBs
as a result of defects in homologous recombination. The
concept of two genetic mutations, that individually have
little effect but when combined are lethal, is known as
synthetic lethality14 and such an interaction exists between
BRCA1/2 and PARP. PARP inhibitors lead to selective death
of tumor cells that have hereditary or acquired BRCA1/2
(homozygous) mutations but have no selective effect on the
normal cells of BRCA carriers (heterozygous). Homologous
recombination defects are thought to occur in up to 50% of
high-grade serous carcinomas,15 not only through BRCA1/2
mutations but interestingly, via a BRCAness phenotype.
Recent microarray studies have identified a BRCAness gene
expression profile in patients with sporadic ovarian cancer
that corresponds with responsiveness to platinum-based
chemotherapy and to PARP inhibitors.16
Clinical Trials
A phase I trial investigated olaparib (AZD2281) in patients with solid tumors refractory to conventional chemotherapy and enriched for patients with BRCA1/2 mutations.
Partial responses were noted in 63% of patients (12 of 19)
including eight patients with ovarian cancer.17 Durable
responses (median, 28 weeks; range, 10 to 86) correlating
with an increased platinum-free interval were reported in
an expanded cohort of 50 patients from this trial. The overall
response rate (RR) was 69% in platinum-sensitive disease,
50% in platinum-resistant disease, and 27% in platinumrefractory disease. Dose-limiting toxicity was seen at 400 mg
and 600 mg twice daily, so the dose-expansion cohort received 200 mg twice daily with minimal toxicitymainly
fatigue and GI symptoms.
These encouraging results led to several phase II trials; a
single-arm, open-label sequential dosing trial in patients
with BRCA1/2 mutations reported clinical efficacy in both
dosing cohorts, but higher response rates were seen in those
receiving olaparib 400 mg twice daily compared with 100 mg
twice daily (RR, 33% and 12.5%, respectively).18 The toxicities were fatigue, sickness, and anemia but mostly grades 1
and 2. A randomized phase II trial comparing the efficacy of
olaparib at two dose levels with pegylated liposomal doxo-
ANTIANGIOGENICS AND PARP INHIBITORS IN OVARIAN CANCER
rubicin (PLD) in patients with BRCA1/2-mutated ovarian

cancer did not show a significant PFS advantage for either
olaparib dosing schedule (200 mg twice daily, 6.5 months;
400 mg twice daily, 8.8 months) compared with PLD (7.1
months; HR 0.88; p 0.66).19 Nonetheless, with comparable efficacy, replacing PLD with a tablet is an attractive
option.
The concept of BRCAness as described earlier herein led
to the investigation of whether PARP inhibitors were clinically active in patients with sporadic serous ovarian cancers.
A phase II trial of olaparib 400 mg twice daily in patients
with (n 17) and without (n 47) BRCA1/2 mutations
reported a response rate of 41% and 24%, respectively. The
better responses were again seen mostly in patients with
platinum-sensitive disease.20
A phase II randomized placebo-controlled trial of olaparib
monotherapy as maintenance treatment for patients with
relapsed platinum-sensitive serous ovarian cancer was reported at ASCO in 2011.21 PFS by Response Evaluation
Criteria in Solid Tumors (RECIST) criteria was significantly
longer in the olaparib group (n 136) compared with the
placebo group (n 129;8.4 vs. 4.8 months; HR 0.35; p
0.00001; Fig. 1). The PFS did not translate into an OS
benefit, which has led AstraZeneca (Wilmington, DE) to stop
pursuing this indication.
Besides those evaluating olaparib, clinical trials are underway with several other PARP inhibitors as monotherapy
in ovarian cancer and include MK4827 (Merck; Whitehouse
Station, NY), CEP-9722 (Cephalon, Frazer, PA), ABT-888
(Abbott Laboratories, Abbott Park, IL) and rucaparib (previously PF-01367338 or AG014699). Preliminary phase I
results for MK4827 have shown efficacy comparable with
olaparib22 and certainly warrant further investigation in
this setting. BMN-673 (BioMarin Pharmaceutical Inc., Novato, CA) is the most selective and potent PARP inhibitor
reported to date, and phase I data are awaited with interest.
Fig 1. Progression-free survival of patients in a phase II randomized placebo-controlled trial of olaparib monotherapy as maintenance treatment for patients with relapsed platinum-sensitive serous
ovarian cancer. Reprinted with permission. American Society of
Clinical Oncology. All rights reserved.21
Conclusion
Bevacizumab 7.5 mg/kg 3-weekly, should be considered

for patients who are at high risk of having a short PFS
because the bevacizumab might delay time to symptomatic
relapse, and it is this group of patients who seem to achieve
a substantial survival gain. For patients with optimally
debulked disease, the greatest benefit from bevacizumab is
more likely to be part of relapse therapy. The greatest
benefit from PARP inhibitors will be in patients with BRCA
mutations, but they are also likely to have a role in patients
with high-grade serous ovarian cancers. Myelosuppression
is likely to prevent PARP inhibitor administration in combination with chemotherapy. Because they have shown
single-agent activity and can delay progression, they have
the potential to be used as another line of relapse therapy or
as maintenance therapy after chemotherapy for relapse.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Rene Roux*
Martin Zweifel*
Gordon J. S. Rustin
AstraZeneca;
Roche
Boehringer
Ingelheim
REFERENCES
1. Campos SM, Ghosh S. A current review of targeted therapeutics for
ovarian cancer. J Clin Oncol. 2010;28:149362.
2. Mukherjee L, Rustin G. Antivascular therapy in gynaecological cancers.
In Kehoe S, Edmondson RJ, Gore M, et al (eds): Gynaecological Cancers:
Biology and Therapeutics. London: RCOG Press, 2011;121-137.
3. Burger RA, Sill MW, Monk BJ, et al. Phase II trial of bevacizumab in
persistent or recurrent epithelial ovarian cancer or primary peritoneal
cancer: a Gynecologic Oncology Group Study. J Clin Oncol. 2007;25:51655171.
4. Cannistra SA, Matulonis UA, Penson RT, et al. Phase II study of
bevacizumab in patients with platinum-resistant ovarian cancer or peritoneal
serous cancer. J Clin Oncol. 2007;25:5180-5186.
5. Perren TJ, Swart AM, Pfisterer J, et al. A phase 3 trial of bevacizumab
in ovarian cancer. N Engl J Med. 2011;365:2484-2496.
6. Burger RA, Brady MF, Bookman MA, et al. Incorporation of bevacizumab in the primary treatment of ovarian cancer. N Engl J Med. 2011;365:
2473-2483.
7. Katsumata N, Yasuda M, Takahashi F, et al. Dose-dense paclitaxel once

a week in combination with carboplatin every 3 weeks for advanced ovarian
cancer: a phase 3, open-label, randomised controlled trial. Lancet. 2009;374:
1331-1338.
8. Gonzalez-Martin A, Gladieff L, Stroyakovskiy D, et al. Front-line bevacizumab (BEV) combined with weekly paclitaxel (wPAC) and carboplatin (C)
for ovarian cancer (OC): safety results from the concurrent chemotherapy
(CT) phase of the OCTAVIa study. Presented at: European Multidisciplinary
Cancer Congress; 2011; Stockholm, Sweden (abstr 8002).
9. Aghajanian C, Finkler NJ, Rutherford T, et al. OCEANS: A randomized,
double-blinded, placebo-controlled phase III trial of chemotherapy with or
without bevacizumab (BEV) in patients with platinum-sensitive recurrent
epithelial ovarian (EOC), primary peritoneal (PPC), or fallopian tube cancer
(FTC). J Clin Oncol. 2011;29:45S (suppl; abstr LBA5007).
10. Coleman RL. Making of a phase III study in recurrent ovarian cancer:
the odyssey of GOG 213. Clin Ovarian Cancer. 2008;1:78-80.
11. Chambon P, Weill JD, Mandel P. Nicotinamide mononucleotide activa-
343
ROUX, ZWEIFEL, AND RUSTIN

tion of new DNA-dependent polyadenylic acid synthesizing nuclear enzyme.
Biochem Biophys Res Commun. 1963;11:39-43.
12. Noel G, Giocanti N, Fernet M, et al. Poly(ADP-ribose) polymerase
(PARP-1) is not involved in DNA double-strand break recovery. BMC Cell
Biol. 2003;4:7.
13. Haber JE. DNA recombination: the replication connection. Trends
Biochem Sci. 1999;24:271-275.
14. Ashworth A. A synthetic lethal therapeutic approach: poly(ADP) ribose
polymerase inhibitors for the treatment of cancers deficient in DNA doublestrand break repair. J Clin Oncol. 2008;26:3785-3790.
15. Press JZ, De Luca A, Boyd N, et al. Ovarian carcinomas with genetic
and epigenetic BRCA1 loss have distinct molecular abnormalities. BMC
Cancer. 2008;8:17.
16. Konstantinopoulos PA, Spentzos D, Karlan BY, et al. Gene expression
profile of BRCAness that correlates with responsiveness to chemotherapy and
with outcome in patients with epithelial ovarian cancer. J Clin Oncol.
2010;28:3555-3561.
17. Fong PC, Boss DS, Yap TA, et al. Inhibition of poly(ADP-ribose)
polymerase in tumors from BRCA mutation carriers. N Engl J Med. 2009;
361:123-134.
344
18. Fong PC, Yap TA, Boss DS, et al. Poly(ADP)-ribose polymerase inhibition: frequent durable responses in BRCA carrier ovarian cancer correlating
with platinum-free interval. J Clin Oncol. 2010;28:2512-2519.
19. Kaye SB, Lubinski J, Matulonis U, et al. Phase II, open-label, randomized, multicenter study comparing the efficacy and safety of olaparib, a poly
(ADP-ribose) polymerase inhibitor, and pegylated liposomal doxorubicin in
patients with BRCA1 or BRCA2 mutations and recurrent ovarian cancer.
J Clin Oncol. 2011;30:372-379.
20. Gelmon KA, Tischkowitz M, Mackay H, et al. Olaparib in patients with
recurrent high-grade serous or poorly differentiated ovarian carcinoma or
triple-negative breast cancer: a phase 2, multicentre, open-label, nonrandomised study. Lancet Oncol. 2011;12:852-861.
21. Ledermann JA, Harter P, Gourley C, et al. Phase II randomized
placebo-controlled study of olaparib (AZD2281) in patients with platinumsensitive relapsed serous ovarian cancer (PSR SOC). J Clin Oncol. 2011;29:
15s (suppl; abstr 5003).
22. Sandhu SK, Wenham, RM, Wilding G, et al. First-in-human trial of a
poly(ADP-ribose) polymerase (PARP) inhibitor MK-4827 in advanced cancer
patients (pts) with antitumor activity in BRCA-deficient and sporadic ovarian
cancers. J Clin Oncol 2010;28:15s (suppl; abstr 3001).
Intraperitoneal Treatment in Ovarian Cancer:

The Gynecologic Oncology Group Perspective
in 2012
By Deborah K. Armstrong, MD, Keiichi Fujiwara, MD, PhD, and Danijela Jelovac, MD
Overview: The peritoneal cavity is the major site of disease in

ovarian cancer. The peritoneal predominance of disease provides a rationale for administration of chemotherapy within
the peritoneal cavity. Intraperitoneal (IP) chemotherapy for
ovarian cancer has been studied rigorously for more than 30
years and has been reproducibly shown to improve the survival of patients with ovarian cancer. Three large randomized
trials of IP compared with intravenous (IV) therapy have
HE PERITONEAL cavity is the major site of disease in

ovarian cancer.1 Whereas ovarian cancer can spread
hematogenously or via the lymphatic system, the bulk of the
tumor will be found on peritoneal surfaces. This peritoneal
disease results from shedding of ovarian tumor cells into the
peritoneal cavity, circulation of these cells throughout the
abdomen and pelvis, and eventual implantation onto peritoneal surfaces. Viability of these cells and successful tumor
growth is further dependent on the development of sufficient
neovasculature to support cell survival and tumor growth.
This unique pattern of spread within the relatively accessible peritoneal cavity has led to attempts at surgical cytoreduction before administration of chemotherapy. Dating
back more than 30 years, nearly every study has demonstrated an inverse correlation between volume of tumor
remaining at the completion of initial surgery and overall
survival for patients with ovarian cancer.2 The peritoneal
predominance of ovarian cancer also provides a rationale
for administration of chemotherapy within the peritoneal
cavity.
When drugs are administered intravenously they are
immediately diluted in the blood. When the same drugs are
administered via the intraperitoneal (IP) route, the peritoneum can have sustained exposure to higher concentrations
of drugs for a more prolonged period of time, whereas normal
tissues such as the bone marrow may be relatively spared.
The pharmacologic advantage of administering a drug by
the peritoneal route can be quantified by the ratio of the
drug concentration (usually measured in area under the
curve [AUC]) in the peritoneal cavity to that in the plasma
after IP compared with IV injection. Table 1 shows this
pharmacologic advantage for some drugs commonly used in
ovarian cancer.3-5
The rate at which the peritoneal drug concentration decreases is a function of the volume of the fluid in the
peritoneal cavity, the surface area through which the drug
diffuses out of the cavity, the permeability of this surface,
and the difference in free drug concentration between the
cavity and the plasma. The clinical effectiveness of a drug
administered IP is affected not only by this pharmacologic
advantage but also by tumor penetration and distribution of
the drug within the peritoneal cavity. It is thus predicted
that tumor volume and the presence of substantial adhesive
disease will influence the efficacy of IP therapy.5
IP chemotherapy was first used in the 1950s for palliation
of ascites, primarily from colorectal carcinoma. In the 1970s
IP treatment became more feasible with the development of
demonstrated statistically significant improvement in clinical

outcome measures. Despite this, the IP approach has not
gained widespread acceptance in the treatment of ovarian
cancer. Here, we review reported, recently completed, and
ongoing trials of IP therapy in ovarian cancer including attempts to improve the tolerability and acceptance of this
proven approach.
permanent indwelling peritoneal catheters that allowed for

repetitive IP administration without requiring repeated
placement of temporary peritoneal catheters. Shortly after
this breakthrough, IP therapy began to be studied in ovarian
cancer. Following is a review of data from completed trials of
IP therapy in ovarian cancer, and a summary of ongoing and
recently completed trials and of novel and innovative approaches to IP therapy, with a particular focus on randomized phase III trials and studies from the Gynecologic
Oncology Group (GOG).
Randomized Trials of IP Versus IV Therapy in
Ovarian Cancer
Eight published comparative studies of IP compared with

IV administration for initial therapy of ovarian cancer were
the subject of a Cochrane meta-analysis in 2007.6 This
analysis of 1,819 women showed that they were less likely to
die if they received an IP component to the chemotherapy
(hazard ratio [HR] 0.79) and that the disease-free interval
was also significantly prolonged (HR 0.79). They did also
note that there may be greater serious toxicity with IP
therapy and that there is a potential for catheter-related
complications in patients receiving IP therapy.
In January 2006, the U.S. National Cancer Institute (NCI)
released a clinical announcement of IP therapy for ovarian
cancer.7 They evaluated data from eight trials and concluded that IP chemotherapy is beneficial for optimally
debulked stage III ovarian cancer. On average, for the eight
trials, IP therapy was associated with a 21.6% decrease in
the risk of death (HR 0.79) which is estimated to translate
into a 12-month increase in overall median survival. The
three largest trials, responsible for more than 75% of the
patients in this analysis, were conducted as U.S. cooperative
group trials and will be further discussed in this section.
The first of these trials, led by Southwest Oncology Group
(SWOG #8501) with the GOG (GOG #104), used IV cyclophosphamide (600 mg/m2) with either IP or IV cisplatin (100
mg/m2) administered every 3 weeks for six cycles.8 This
study showed a survival advantage for the group receiving
From the Johns Hopkins Sidney Kimmel Cancer Center, Baltimore, MD.
Address reprint requests to Deborah K. Armstrong, MD, Associate Professor of Oncology,
Associate Professor of Gynecology & Obstetrics, Johns Hopkins Sidney Kimmel Cancer
Center, 1650 Orleans Street, Room 190, Baltimore, MD 21231; email: darmstro@jhmi.edu.
1092-9118/10/1-10
345
ARMSTRONG, FUJIWARA, AND JELOVAC

Table 1. Pharmacologic Advantage for Intraperitoneal
Chemotherapy
Drug
Molecular Weight
Ratio of Drug AUC,

Peritoneal Cavity/Plasma
Cisplatin
Carboplatin
Topotecan
Docetaxel
Gemcitabine
Paclitaxel
300
371
458
862
300
854
12
1018
54
181
759
1,000
Abbreviation: AUC, area under the curve.

Adapted and modified from Markman 3 and Fujiwara et al. 4
IP cisplatin. In the 546 eligible patients, the estimated

median survival was substantially longer in the IP group (49
months) compared with the IV group (41 months). The HR
for the risk of death was 0.76 (p 0.02) in favor of IP
therapy. Although moderate to severe abdominal pain was
more frequent in the IP group, likely related to the catheter
and volume of infusate, grade 3/4 granulocytopenia and
tinnitus, clinical hearing loss, and grade 2 to 4 neuromuscular toxic effects were markedly more frequent in the IV
group. This is the purest of the IP trials to date as it used
the same drugs and same doses in both arms with only the
route of administration being different. Despite the benefits
observed in this trial, IP therapy was not widely embraced
by the gynecologic oncology community at the time of this
publication. This may be because of the publication in the
same year of the benefits of paclitaxel in patients with
suboptimally debulked disease.9
A second IP trial was also conducted by GOG (GOG #114)
and Southwest Oncology Group (SWOG #9227). In this trial
426 patients were randomly assigned to receive either IV
paclitaxel 135 mg/m2 over 24 hours followed by IV cisplatin
75 mg/m2 every 3 weeks for six cycles or IV carboplatin
(AUC 9) every 28 days for two cycles followed by a regimen
of IV paclitaxel 135 mg/m2 over 24 hours followed by IP
cisplatin at 100 mg/m2 every 3 weeks for six cycles (eight
total cycles of chemotherapy).10 This study demonstrated an
improved progression-free survival (median, 28 vs. 22
months; HR 0.78; p 0.01) and overall survival (median,
63 vs. 52 months; HR 0.81; p 0.05) in favor of the IP
KEY POINTS
346
Current recommendations for intraperitoneal (IP)

therapy are for patients with disease optimally debulked after initial surgery.
Three North American randomized phase III trials
comparing IP with intravenous therapy have demonstrated substantial improvement in progression-free
and overall survival.
Cisplatin is the mainstay of IP platinum treatment
but a recently completed trial (GOG 252) is the first to
use IP carboplatin in an arm of a randomized phase
III trial.
IP therapy after neoadjuvant therapy and interval
debulking is currently being tested.
IP therapy has not been rigorously tested for patients
with recurrent disease.
group. Toxicities grade 3 or worse, including neutropenia,

thrombocytopenia, and GI and metabolic toxicities, were
markedly more frequent in the IP group. As a result, 18% of
the patients on the IP arm received fewer than two courses
of IP therapy. Despite the substantial survival improvement
in this study, the gynecologic oncology community again did
not accept IP chemotherapy as standard treatment for ovarian
cancer. Many attributed the benefits seen in the experimental
arm to patients receiving eight cycles of therapy or to the
intensive carboplatin rather than the IP therapy.
The third trial was conducted by GOG (GOG #172). In this
study 417 eligible patients with optimally debulked stage III
ovarian cancer were randomly assigned to receive IV paclitaxel (135 mg/m2/24 hours) followed by IV cisplatin (75
mg/m2) or IV paclitaxel (135 mg/m2/24 hours) followed by IP
cisplatin (100 mg/m2), plus IP paclitaxel (60 mg/m2) on day
8.11 Treatments were repeated every 21 days for six cycles.
The median progression-free survival was 18.3 months in
the IV group and 23.8 months in the IP group (HR 0.77,
p 0.05). The median overall survival was 49.7 in the IV
group and 65.6 months IP group (HR 0.73, p 0.03). The
magnitude of improvement in median overall survival associated with IP/IV administration of chemotherapy is similar
to that observed with the introduction of either cisplatin or
paclitaxel. The 66-month median survival for the IP arm of
GOG 172 is the longest survival reported to date from a
randomized trial of advanced ovarian cancer.
Once again, the data in support of IP therapy have not
resulted in widespread acceptance of the IP approach. Many
argued that it was the additional drug delivered in the IP
arm of GOG 172 that resulted in the improved outcome. This
is somewhat circuitous logic in that those doses of drug
cannot be delivered intravenously. The IP arm was designed
to be intentionally intense, exploiting the ability to administer more drug per unit of time IP than can be delivered
with IV therapy.
There were substantially more patients with grade 3 and
4 leukopenia, thrombocytopenia, and GI toxicity, renal toxicity, neurologic toxicity, fatigue, infection, metabolic toxicity, and pain toxicity in the IP arm compared with the IV
arm. Because of these toxicities and/or catheter problems,
48% of patients in the IP arm received three or fewer IP
treatments, and only 42% patients received the planned six
cycles of IP therapy. Given that the study results are based
on an intent-to-treat analysis, modifications of the regimen
to improve tolerability could allow for improved completion
rates and possibly even better outcomes.
In a separate quality-of-life (QOL) analysis, patients who
received IP therapy had a worse QOL during therapy, but
there was no difference 1 year after completion of treatment.12 We and others have shown that the GOG 172 IP
regimen can be administered successfully with a reasonable
toxicity profile when patients are appropriately selected and
the therapy is performed at an experienced center with the
assistance of a skilled and dedicated support staff.13
Some have suggested that GOG 172 may overestimate the
benefit of IP therapy. Ozols and colleagues have reported on
a preliminary cross-trial analysis comparing the results of
IP therapy in GOG 172 with IV carboplatin/paclitaxel in
392 similarly staged patients in GOG #158,14 calculating
that instead of the 15.9-month improvement in median
overall survival, the difference may be substantially less (8.2
months) if carboplatin/paclitaxel had been the comparative
INTRAPERITONEAL TREATMENT IN OVARIAN CANCER
PFS: % Increase
OS: % Increase
25
20
Fig. 1. IP compared to IV chemotherapy Phase III
trials. The bars show the % increase in progression-free
survival (PFS, blue bars) and overall survival (OS, orange
bars) for IP compared with IV therapy.8,10,11 The red bar
shows the percent increase in OS for the IV arm of GOG
17210 with IV paclitaxel and carboplatin (TC) from GOG
protocol 158.13
15
10
5
0
GOG 104
arm. Historical, nonrandomized cross-trial comparisons
such as these lack the validity of those generated by prospective randomization and should not be relied upon for
generating credible conclusions. However, in response to the
assertions of Ozols et al., we performed a more rigorous
statistical analysis using this cross-trial comparison showed
that there remained a 19% improvement in overall survival
when the IP arm of GOG 172 was compared with the IV
paclitaxel/carboplatin arm of GOG 158.15 Figure 1 shows the
percentage improvement in progression-free and in overall
survival from these three trials and includes the overall
survival improvement comparing the IP arm of GOG 172
with the IV paclitaxel/carboplatin arm of GOG 158.
Recent Trials and New Approaches to IP Therapy
GOG protocol 252 recently closed to accrual. This trial

included an arm of IP carboplatin with weekly IV paclitaxel,
an arm with an IP cisplatin regimen that is a modification of
GOG 114
GOG 172
G158 TC c/w
G 172 IP OS
the IP arm of GOG 172, and an arm with IV weekly

paclitaxel and IV carboplatin. This is the first randomized
phase III trial to rigorously evaluate IP carboplatin. All
patients in this trial received bevacizumab. In addition, the
protocol enrolled a limited number of patients with suboptimal residual disease, which will allow for a preliminary
evaluation of the role of IP therapy in that setting.
Currently, the National Cancer Institute of Canada Clinical Trials Group, in conjunction with the Gynecologic Cancer Intergroup, in protocol OV.21 is examining the use of IP
platinum-taxane-based chemotherapy after neoadjuvant
chemotherapy and subsequent interval surgical debulking.
To address whether the less systemically toxic carboplatin
can be substituted for cisplatin IP, the first phase of the
study will have three arms: one IV only, and two IP-containing
regimens, one with carboplatin and one with cisplatin. At
completion of the first stage, one of the IP regimens will
proceed into a phase III comparison with the IV arm.
Author
Deborah K. Armstrong
Keiichi Fujiwara
Employment or
Leadership
Positions
Consultant or
Advisory Role
Eisai (I);
Genentech;
Genzyme;
Oncogenex
Amgen;
Boehringer
Ingelheim;
GlaxoSmithKline;
Zeria Pharma
Stock
Ownership
Honoraria
Research
Funding
Eisai (I);
Exelixis (I);
Morphotek
Expert
Testimony
Other
Remuneration
Bristol-Myers
Squibb;
Chugai Pharma;
Janssen
Oncology;
Nihonkayaku;
Sanofi; Taiho
Pharmaceutical
Danijela Jelovac*
REFERENCES
1. Jelovac D, Armstrong DK. Recent progress in the diagnosis and treatment of ovarian cancer. CA Cancer J Clin. 2011;61:183-203.
2. Bristow RE, Tomacruz RS, Armstrong DK, et al. Survival impact of
maximum cytoreductive surgery for advanced ovarian carcinoma during the
platinum-era: a meta-analysis of 6,885 patients. J Clin Oncol. 2002;20:12481259.
3. Markman M. Intraperitoneal chemotherapy. Semin Oncol. 1991;18:24854.
4. Fujiwara K, Armstrong D, Morgan M, et. al. Principles and practice of
intraperitoneal chemotherapy for ovarian cancer. Int J Gynecol Cancer.

2007;17:1-20.
5. Howell SB. Pharmacologic principles of intraperitoneal chemotherapy
for the treatment of ovarian cancer. Int J Gynecol Cancer. 2008;18:20-25
(suppl 1).
6. Jaaback K, Johnson N. Intraperitoneal chemotherapy for the initial
management of primary epithelial ovarian cancer. The Cochrane Database of
Systematic Reviews. 2006;1:CD005340.pub2.
7. NCI clinical announcement. http://ctep.cancer.gov/highlights/docs/clin_
347
ARMSTRONG, FUJIWARA, AND JELOVAC

annc_010506.pdf#searchintraperitonealchemotherapy. Accessed March 23,
2012.
8. Alberts DS, Liu PY, Hannigan EV, et al. Intraperitoneal cisplatin plus
intravenous cyclophosphamide versus intravenous cisplatin plus intravenous
cyclophosphamide for stage III ovarian cancer. N Engl J Med. 1996;335:19501955.
9. McGuire WP, Hoskins WJ, Brady MF, et al. Cyclophosphamide and
cisplatin compared with paclitaxel and cisplatin in patients with stage III and
stage IV ovarian cancer. N Engl J Med. 1996;334:1-6.
10. Markman M, Bundy BN, Alberts DS, et al. Phase III trial of standarddose intravenous cisplatin plus paclitaxel versus moderately high-dose carboplatin followed by intravenous paclitaxel and intraperitoneal cisplatin in
small-volume stage III ovarian carcinoma: an intergroup study of the Gynecologic Oncology Group, Southwestern Oncology Group, and Eastern Cooperative Oncology Group. J Clin Oncol. 2001;19:1001-1007.
348
11. Armstrong DK, Bundy B, Wenzel L, et. al. Intraperitoneal cisplatin and
paclitaxel in ovarian cancer. N Engl J Med. 2006;354:34-43.
12. Wenzel L, Huang HQ, Armstrong DK, et. al. Health-related quality of
life during and after intraperitoneal versus intravenous chemotherapy for
optimally debulked ovarian cancer: a Gynecologic Oncology Group study.
J Clin Oncol. 2007;25:437-443.
13. Guile MW, Horne AL, Thompson SD, et al. Intraperitoneal chemotherapy for stage III ovarian cancer using the Gynecologic Oncology Group
protocol 172 intraperitoneal regimen: Effect of supportive care using aprepitant and pegfilgrastim on treatment completion rate. Clin Ovarian Cancer.
2008;1:68-71.
14. Ozols RF, Bookman MA, Young RC. Intraperitoneal chemotherapy for
ovarian cancer. N Engl J Med. 2006;354:1641-1643.
15. Armstrong DK, Brady MF. Intraperitoneal therapy for ovarian cancer:
a treatment ready for prime time. J Clin Oncol. 2006;24:4531-4533.
Dose-Dense Chemotherapy and Neoadjuvant

Chemotherapy for Ovarian Cancer
By Keiichi Fujiwara, MD, PhD, Noriyuki Katsumata, MD, PhD, and
Takashi Onda, MD, PhD
Overview: Two of the innovative chemotherapeutic approaches to ovarian cancer treatment, dose-dense chemotherapy and neoadjuvant chemotherapy, will be discussed
herein. The primary concept of dose-dense chemotherapy is to
administer the same cumulative dose of chemotherapy over a
shorter period. Increased dose density is achieved by reducing the interval between each dose of chemotherapy. The
Japanese Gynecologic Oncology Group (JGOG) first demonstrated the survival advantage of dose-dense weekly administration of paclitaxel in 2009. However, there are unanswered
questions, such as the question of dose-dense carboplatin
versus less dose-intensive regimens. Clear cell or mucinous
carcinomas seem to need other strategies, such as targeted
agents. The aim of neoadjuvant chemotherapy is to reduce
VARIAN CANCER is the most lethal disease among

gynecologic malignancies because this disease remains
most commonly diagnosed in advanced stages.1
Treatment of ovarian cancer has been investigated for
more than 30 years, since cisplatin was introduced in the
1970s. The clinical outcome of ovarian cancer was improved
because of the development of new anticancer agents and
advancements in surgical technique, equipment, and anesthesia. Consequently, current standard therapy for advanced ovarian cancer became a combination of maximum
surgical effort to remove bulky abdominal disease followed
by chemotherapy with paclitaxel plus carboplatin. However,
this has not changed since 1999, despite great efforts to find
new therapeutic strategy.
The first approach to improve survival was by finding
other effective drugs or treatment modalities. The most
thrilling area at this time is the development of new anticancer drugs, especially targeted agents. It is unfortunate,
however, that most pharmaceutical companies have not paid
great attention to gynecologic cancer. Therefore, we had to
wait until 2010 before the first targeted agent, bevacizumab,
showed positive results in ovarian cancer chemotherapy.2,3
In the meantime, investigators have conducted academic
trials to find a way to improve the prognosis of patients with
ovarian cancer. Those trials include intraperitoneal (IP)
chemotherapy, maintenance chemotherapy, and dose-dense
weekly chemotherapy.
On the basis of the results of these large-scale randomized
trials, the 4th Ovarian Cancer Consensus Conference in
20104 yielded the following statement: It was agreed unanimously that the basic minimum comparator in a phase III
trial of advanced ovarian carcinoma must contain a taxane
and a platinum agent given for 6 cycles. Acceptable alternatives must be supported by at least 1 clinical trial demonstrating noninferiority or superiority to a standard taxane/
platinum regimen. Acceptable alternatives at present
include IP delivery to patients with small-volume residual
disease, weekly paclitaxel in combination with carboplatin
every 3 weeks, bevacizumab given concurrently with paclitaxel/carboplatin followed by bevacizumab maintenance,
and 12 months of monthly paclitaxel maintenance given to
tumor volume or spread before main treatment. This could

then make the main procedures easier or less invasive, just
like breast-conserving surgery after neoadjuvant chemotherapy. In advanced ovarian cancer, standard procedure is maximum primary debulking surgery followed by chemotherapy.
Recently, a prospective randomized trial demonstrated that
neoadjuvant chemotherapy followed by interval debulking
surgery was not inferior to the standard procedure. However,
there are several questions that remain unanswered, such as
the suitable number of chemotherapy cycles before interval
debulking surgery. Some of those questions regarding dosedense chemotherapy or neoadjuvant chemotherapy may be
resolved by ongoing or future prospective trials.
patients who achieve a clinical complete response with 6

cycles of standard paclitaxel/carboplatin. Dose-dense paclitaxel was the one regimen that dramatically improved the
survival of patients with ovarian cancer.
A second approach is to attempt to reduce the patients
tumor burden, if the clinical outcome is the same regardless
of the intensity or aggressiveness of the treatment. These
less-invasive treatments will improve the patients quality of
life compared with more intensive therapies. Neoadjuvant
chemotherapy is one of these approaches, although it is
controversial. In the 4th Ovarian Consensus Conference
statement,4 it was concluded that Delayed primary surgery
following neoadjuvant chemotherapy is an option for selected patients with stage IIIC and IV ovarian cancer as
included in EORTC 55971,5 although this was the only
issue among the consensus conference in which the total
consensus was not reached.
Dose-Dense Chemotherapy
The basic concept of dose-dense therapy is to administer

the same cumulative dose of chemotherapy over a shorter
period. Increased dose density is achieved by reducing the
interval between each dose of chemotherapy. The theoretical
basis for this dose-dense chemotherapy strategy is derived
from the Gompertzian model, which is based on NortonSimons hypothesis.6 In the Gompertzian model, smaller
tumors grow faster and so tumor regrowth between treatment cycles is more rapid when cell kill is greatest. The
Norton-Simon model suggests that increasing the dose density of chemotherapy will increase efficacy by minimizing
From the Department of Gynecologic Oncology, Saitama Medical School International

Medical Center, Saitama, Japan; Department of Medical Oncology, Nippon Medical
School, Musashikosugi Hospital, Kawasaki-City, Japan; Department of Gynecology, Kitasato University School of Medicine, Sagamihara-City, Kanagawa, Japan.
Address reprint requests to Keiichi Fujiwara, MD, PhD, Department of Gynecologic
Oncology, Saitama Medical School, International Medical Center,1397-1 Yamane, HidakaCity, Saitama, 350-1298, Japan; email: fujiwara@saitama-med.ac.jp.
1092-9118/10/1-10
349
FUJIWARA, KATSUMATA, AND ONDA
the opportunity for regrowth of tumor cells between cycles of

chemotherapy.
Clinically, the benefit of dose-dense chemotherapy was
first proven in breast cancer. Two randomized phase III
trials on breast cancer showed improved survival benefit by
administering paclitaxel with weekly compared with every3-weeks administration.7,8
In ovarian cancer, the Japanese Gynecologic Oncology
Group (JGOG) first demonstrated the survival advantage
of dose-dense weekly administration of paclitaxel.9 In
JGOG3016, 637 patients were randomly assigned to receive
six cycles of either paclitaxel (180 mg/m2; 3-hour intravenous [IV] infusion) plus carboplatin (area under the curve
[AUC] 6 mg/mL/min), administered on day 1 of a 21-day
cycle (conventional regimen; n 320), or dose-dense paclitaxel (80 mg/m2; 1-hour IV infusion) administered on days 1,
8, and 15 plus carboplatin administered on day 1 of a 21-day
cycle (dose-dense regimen; n 317). The primary end point
was progression-free survival, and secondary end points
were overall survival and toxicity. A total of 631 patients
were eligible (dose-dense regimen, n 312; conventional
regimen, n 319). Median progression-free survival was
significantly longer in the dose-dense treatment group (28.0
months; 95% CI, 22.3 to 35.4 months) than in the conventional treatment group (17.2 months; 95% CI, 15.7 to 21.1
months; hazard ratio [HR] 0.71; 95% CI, 0.58 to 0.88; p
0.0015). Overall survival at 3 years was higher in the
dose-dense regimen group (72.1%) than in the conventional
treatment group (65.1%; HR 0.75: 95% CI, 0.57 to 0.98;
p 0.03). Early discontinuation of the treatment was more
frequent in dose-dense group (n 165) than in the conventional group (n 117) primarily because of the toxicities
(n 113 vs. n 69). The most common adverse event was
neutropenia (dose-dense regimen, 286 [92%] of 312 patients;
conventional regimen, 276 [88%] of 314 patients), but not
statistically different. The frequency of grade 3 and 4 anemia was significantly higher in the dose-dense group (n
214 [69%]) than in the conventional treatment group (n
137 [44%]; p 0.0001). The frequencies of other toxicities
including peripheral neuropathy were similar between
groups. On the basis of these results, it was concluded that
dose-dense weekly paclitaxel plus carboplatin improved sur-
KEY POINTS
350
One trial of dose-dense chemotherapy (Japanese Gynecologic Oncology Group [JGOG]-3016) has shown
significant improvement in progression-free survival
and overall survival in patients with advanced ovarian cancer. Three confirmatory studies are ongoing
worldwide.
One trial of neoadjuvant chemotherapy (European
[EORTC]) has shown that there was no substantial
difference in OS in patients with stage III/IV ovarian
cancer. Two confirmatory studies are ongoing worldwide.
The role of dose-dense chemotherapy and neoadjuvant chemotherapy will be clarified with solid highlevel evidence in the near future.
vival compared with the conventional triweekly administration of paclitaxel with the cost of modest increase in
toxicities.
The result of this trial has markedly influenced the
designs of clinical trials, which were planned at the time the
JGOG3016 result was presented. The GOG252 trial was
scheduled to compare IV administration of paclitaxel (day 1)
plus intravenous carboplatin (day 1; arm 1) versus IV
paclitaxel (day 1) plus IP carboplatin (day 1; arm 2) versus
modified GOG172 trial winner regimen, which was IV paclitaxel (day 1) plus IP cisplatin (day 2) plus IP paclitaxel
(day 8; arm 3) (see http://clinicaltrials.gov/ct2/show/
NCT01167712?termGOG0262&rank1 for trial information). All three arms incorporated concurrent and
maintenance bevacizumab. Only arm 3 had administration
of paclitaxel on day 8. This was greatly criticized because
there might be a possibility that day-8 administration of
paclitaxel contributed to the improvement of overall survival, not to the result of IP chemotherapy. Therefore, as
Bookman described in the Commentary,10 it should be
proven that the net contribution of weekly paclitaxel to the
overall survival advantage associated with intraperitoneal
therapy will hopefully be addressed in future studies. In
fact, the GOG decided to incorporate weekly dose-dense
administration of paclitaxel into the two carboplatin arms.
Another scheduled IP trial was OV-20, a National Cancer
Institute of Canada/Gynecologic Cancer Intergroup (GCIG)
trial. The trial design of OV-20 was similar to that of
GOG252 trial, except that one of two IP arms would be
chosen by randomized phase II fashion, and they did not
combine bevacizumab. This trial design was amended to
incorporate day-8 administration of paclitaxel (not dosedense weekly) for IV and IP carboplatin arms (arm 1 and
arm 2).
Another great movement among gynecologic oncology clinical trials was to conduct confirmatory trials of dose-dense
chemotherapy. In addition to the confirmation of the same
dose-dense weekly regimen of paclitaxel, these trials try to
answer the following questions. The first is whether weekly
administration of carboplatin also contributes to improve
survival. The second question is whether simple division of
total dose of paclitaxel will demonstrate similar efficacy with
less toxicity.
At this time, three randomized trials are ongoing. The
GOG262 trial applied exactly the same trial arms conventional triweekly regimen of paclitaxel plus carboplatin
although it only included stage III/IV patients. Use of
bevacizumab is patient choice (Fig. 1). This trial was closed
in early 2012. The second trial is MITO-7 study, led by the
Multicenter Italian Trials in Ovarian Cancer Group (Fig. 2).
This study compares the efficacy and toxicity of weekly
administration of carboplatin in addition to weekly administration of paclitaxel. The dose of paclitaxel in the experimental arm is 60 mg/m2 instead of 80 mg/m2 in the
JGOG3016 study. The dose of carboplatin in the experimental arm is AUC 2 administered every week. This trial was
opened in 2008, and accrual of 800 patients was completed
recently. The third trial is an ambitious study conducted by
Integrated Community Oncology Network (ICON) and European Network of Gynaecological Oncological Trial Groups
(ENGOT; Fig. 3). ICON8-ENGOT OV-13 is a three-arm trial
with conventional triweekly administration of paclitaxel
plus carboplatin as a comparator (arm 1), and two experi-
CHEMOTHERAPY FOR OVARIAN CANCER
Fig. 1. Study design of the GOG0262

trial.
Abbreviations: AUC, area under the
curve; PFS, progression-free survival.
mental arms of dose-dense weekly paclitaxel plus triweekly

carboplatin (arm 2) and dose-dense weekly administration of
both paclitaxel and carboplatin (arm 3). They allowed the
use of neoadjuvant chemotherapy as an investigators
choice. This trial has just opened in 2011.
Another important unresolved question is the fact that
JGOG3016 failed to show the survival benefit in clear cell
and mucinous cancers. A meta-analysis of breast cancer
trials indicated that the dose-dense strategy only contributed to prolonging survival in the hormone receptornegative patients.11 The patient population who truly benefit
from the dose-dense strategy must be clarified by future
studies.
In conclusion, dose-dense administration of paclitaxel
showed a substantial improvement in the survival of patients with ovarian cancer. Further studies ongoing world-
wide and meta-analysis will prove the optimal use of dosedense strategy in ovarian cancer.
Neoadjuvant Chemotherapy
The general concept of neoadjuvant chemotherapy is to

administer chemotherapy before main treatment such as
surgery or radiation therapy. The purpose of this strategy is
to reduce the tumor size or extent of cancer spread before
applying the radical main treatment, thus making the
procedure easier or less invasive. It also provides the chance
to know whether the chemotherapy is effective, which is not
possible when the tumor is completely removed.
Neoadjuvant chemotherapy has been studied in several
types of cancer, such as breast, prostate, cervix, colorectal,
lung, and esophageal cancers. Among them, breast, prostate, and cervical cancers have been investigated more
Fig. 2. Study design of MITO-7 trial.

curve; PFS, progression-free survival.
351
Fig. 3. Study design of ICON8ENGOT OV-13 trial.

curve; DPS, delayed primary surgery;
EOC, epithelial ovarian cancer; FTC,
fallopian tube cancer; IPS, immediate
primary surgery; PFS, progression-free
survival; PPC, primary peritoneal cancer.
than other tumors in which consensus conference or metaanalysis have already been conducted.12-14 In the cervical
cancer, neoadjuvant chemotherapy followed by radical hysterectomy improved survival compared with radiation alone.
But neoadjuvant chemotherapy followed by radiation therapy negatively affected survival compared with radiation
alone, and the benefit of neoadjuvant chemotherapy followed
by radical hysterectomy has not been concluded in comparison with radical hysterectomy. It has not been compared
with current standard, chemoradiotherapy. In prostate cancer, neoadjuvant hormone therapy was proven to be effective
only when radiation therapy was applied afterward, but not
beneficial if surgery was conducted after antiandrogenic
therapy.13 In breast cancer, neoadjuvant therapy was first
used, in the 1980s, typically for patients with inoperable
locally advanced or inflammatory breast cancer, and the
breast-conserving surgery rate dramatically increased. The
next step for the neoadjuvant therapy was to use it as an in
vivo test for chemosensitivity by assessing pathologic complete response. Currently, by using pathologic response and
other biomarkers as intermediate end points, results from
trials of new regimens and therapies that use neoadjuvant
therapy are aimed to proceed and anticipate the results from
larger adjuvant trials.12
In ovarian cancer, primary debulking surgery followed
by adjuvant chemotherapy is a gold standard procedure.
Although some investigators reported their favorable experience of neoadjuvant chemotherapy followed by interval
debulking surgery, meta-analysis suggested that neoadjuvant chemotherapy was associated with poorer outcome.12,15,16 However, there had been no randomized trial
that prospectively demonstrated that primary debulking
surgery is better than neoadjuvant chemotherapy followed
by less-invasive interval debulking surgery. EORTC55971 is
the first prospective randomized study of advanced (stage
IIIC or IV) ovarian carcinoma, fallopian tube carcinoma, or
primary peritoneal carcinoma to compare overall survival
between patients who received standard primary debulking
surgery followed by chemotherapy and those who received
352
neoadjuvant chemotherapy plus interval debulking surgery.5 The majority of patients who entered this trial (n
670) had extensive stage IIIC or IV disease at the treatment.
The largest residual tumor 1 cm or less in diameter was
achieved in 41.6% of patients after primary debulking and in
80.6% of patients after interval debulking. The HR for death
in the neoadjuvant chemotherapy group compared with the
primary debulking surgery group was 0.98 (90% CI, 0.84 to
1.13; p 0.01 for noninferiority), and the HR for progressive
disease was 1.01 (90% CI, 0.89 to 1.15). Complete resection
of all macroscopic disease (at primary or interval surgery)
was the strongest independent variable in predicting overall
survival. Postoperative rates of adverse events and mortality were higher after primary debulking than after interval
debulking.
This study raised considerable controversies13,17-19; thus,
additional phase III trial(s) are necessary to clarify the
benefit of the neoadjuvant strategy. Fortunately, two prospective randomized trials have already completed accrual,
one from the United Kingdom and another from Japan.
The United Kingdom trial CHORUS (Chemotherapy or
Upfront Surgery) is a randomized trial to determine the
impact of timing of surgery and chemotherapy in patients
with newly diagnosed stage III/IV ovarian, primary peritoneal, or fallopian tube carcinoma. Study design is similar to
that of EORTC55971. The patients were randomly assigned
to receive either immediate primary debulking surgery followed by six cycles of chemotherapy, or neoadjuvant chemotherapy for three cycles followed by interval debulking
surgery, and then an additional three cycles of chemotherapy. For patients assigned to receive neoadjuvant chemotherapy, however, histologic or cytologic confirmation of
target diseases was necessary before starting treatment.
The target accrual was 550 and accrual has already been
accomplished. The data will be combined with EORTC55971
to reliably exclude a 5% to 6% difference in 3-year overall
survival. This trial accomplished enrollment and is waiting
for analysis.
JCOG conducted a randomized trial (JCOG0602)20 in
CHEMOTHERAPY FOR OVARIAN CANCER
which patients with stage III/IV ovarian, tubal, or primary

peritoneal cancer, were allocated to receive either primary
debulking surgery followed by eight cycles of chemotherapy
or to receive neoadjuvant chemotherapy with four cycles of
chemotherapy followed by interval debulking surgery plus
an additional four cycles of chemotherapy. This study is
designed as a noninferiority trial with 300 patients in total.
This trial completed accrual in 2011.
The ultimate goal of neoadjuvant chemotherapy is to
minimize the invasiveness of surgical treatment without
jeopardizing its efficacy. The most successful example is
breast-conserving surgery after neoadjuvant chemotherapy.
In the EORTC55791 trial, the survival was same between
primary debulking surgery group and neoadjuvant chemotherapy group, but it was not fully analyzed how they could
minimize the invasiveness of surgery between primary and
interval surgery.
The remaining questions on neoadjuvant chemotherapy
include: 1) accurate selection of candidate patients, 2) duration of chemotherapy before interval debulking surgery, 3)
selection and duration of adjuvant therapy after interval
debulking surgery, and finally 4) whether the failure to
achieve optimal cytoreductive surgery is to the result of the
surgeons skill or biology of the cancer.
For the first question, JCOG investigators conducted a
prospective feasibility study on the selection of the patients
who would be suitable for neoadjuvant chemotherapy
(JCOG0206).21 They assessed the accuracy of clinical diagnosis on the basis of imaging tests, cytology from ascites,
pleural effusion or tumor, and tumor markers (cancer antigen [CA]-125 200 U/mL and Carcinoembryonic antigen
[CEA] 20 ng/mL). The diagnosis was confirmed by diagnostic laparoscopy and these results showed that patients
can be correctly diagnosed as having ovarian, fallopian tube,
or primary peritoneal carcinoma with greater than 90%
accuracy by clinical diagnoses on the basis of findings
including cytology, according to Bayesian statistical methods.
Other questions, however, have not been investigated
prospectively. We can hypothesize that the greater the
number of cycles of neoadjuvant chemotherapy, the less
invasive the interval surgery would be because the disseminated tumor could be eliminated by the chemotherapy.
However, meta-analysis of retrospective studies reported
that increasing the number of chemotherapy cycles before
interval surgery would negatively affect survival.16,22 It is
hoped that these important questions will be answered
prospectively in the near future.
Conclusion
The important concept of dose-dense chemotherapy and

neoadjuvant chemotherapy in the treatment of ovarian
cancer has been tested in phase III trials5,9 and is a great
influence on clinical practice.
Author
Keiichi Fujiwara
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Amgen;
Boehringer
Ingelheim;
GlaxoSmithKline;
Zeria Pharma
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Bristol-Myers
Squibb; Chugai
Pharma;
Janssen
Oncology;
Nihonkayaku;
Sanofi; Taiho
Pharmaceutical
Noriyuki Katsumata*
Takashi Onda*
REFERENCES
1. Stuart GC, Kitchener H, Bacon M, et al. 2010 Gynecologic Cancer
InterGroup (GCIG) consensus statement on clinical trials in ovarian cancer:
report from the Fourth Ovarian Cancer Consensus Conference. Int J Gynecol
Cancer. 2011;21:750-755.
2. Burger RA, Brady MF, Bookman MA, et al. Incorporation of bevacizumab in the primary treatment of ovarian cancer. N Engl J Med. 2011;365:
2473-2483.
3. Perren TJ, Swart AM, Pfisterer J, et al. A phase 3 trial of bevacizumab
in ovarian cancer. N Engl J Med. 2011;365:2484-2496.
4. Thigpen T, duBois A, McAlpine J, et al. First-line therapy in ovarian
cancer trials. Int J Gynecol Cancer. 2011;21:756-1762.
5. Vergote I, Trope CG, Amant F, et al. Neoadjuvant chemotherapy or
primary surgery in stage IIIC or IV ovarian cancer. N Engl J Med. 2010;363:
943-953.
6. Norton L. Theoretical concepts and the emerging role of taxanes in
adjuvant therapy. Oncologist. 2001; 3:30-35 (suppl).
7. Seidman AD, Berry D, Cirrincione C, et al. Randomized phase III trial of
weekly compared with every-3-weeks paclitaxel for metastatic breast cancer,
with trastuzumab for all HER-2 overexpressors and random assignment to
trastuzumab or not in HER-2 nonoverexpressors: final results of Cancer and
Leukemia Group B protocol 9840. J Clin Oncol. 2008;26:1642-1649.
8. Sparano JA, Wang M, Martino S, et al. Weekly paclitaxel in the adjuvant
treatment of breast cancer. N Engl J Med. 2008;358:1663-1671.
9. Katsumata N, Yasuda M, Takahashi F, et al. Dose-dense paclitaxel once

a week in combination with carboplatin every 3 weeks for advanced ovarian
cancer: a phase 3, open-label, randomised controlled trial. Lancet. 2009;374:
1331-1338.
10. Bookman MA. Dose-dense chemotherapy in advanced ovarian cancer.
Lancet. 2009;374:1303-1305.
11. Bonilla L, Ben-Aharon I, Vidal L, et al. Dose-dense chemotherapy in
nonmetastatic breast cancer: a systematic review and meta-analysis of
randomized controlled trials. J Natl Cancer Inst. 2010;102:1845-1854.
12. Kaufmann M, von Minckwitz G, Mamounas EP, et al. Recommendations from an International Consensus Conference on the current status and
future of neoadjuvant systemic therapy in primary breast Cancer. Ann Surg
Oncol. Epub 2011 Dec 23.
13. Shelley MD, Kumar S, Wilt T, et al. A systematic review and metaanalysis of randomised trials of neo-adjuvant hormone therapy for localised
and locally advanced prostate carcinoma. Cancer Treat Rev. 2009;35:9-17.
14. Neoadjuvant chemotherapy for locally advanced cervical cancer: a
systematic review and meta-analysis of individual patient data from 21
randomised trials. Eur J Cancer. 2003;39:2470-2486.
15. Onda T, Yoshikawa H. Neoadjuvant chemotherapy for advanced ovarian cancer: overview of outcomes and unanswered questions. Expert Rev
Anticancer Ther. 2011;11:1053-1067.
16. Bristow RE, Chi DS. Platinum-based neoadjuvant chemotherapy and
353

interval surgical cytoreduction for advanced ovarian cancer: a meta-analysis.
Gynecol Oncol. 2006;103:1070-1076.
17. Chi DS, Bristow RE, Armstrong DK, et al. Is the easier way ever the
better way? J Clin Oncol. 2011;29:4073-4075.
18. Chi DS, Musa F, Dao F, et al. An analysis of patients with bulky
advanced stage ovarian, tubal, and peritoneal carcinoma treated with primary debulking surgery (PDS) during an identical time period as the
randomized EORTC-NCIC trial of PDS vs neoadjuvant chemotherapy
(NACT). Gynecol Oncol. 2012;124:10-14.
19. Vergote I, Trope CG, Amant F, et al. Neoadjuvant chemotherapy is the
better treatment option in some patients with stage IIIc to IV ovarian cancer.
J Clin Oncol. 2011;29:4076-4078.
354
20. Onda T, Matsumoto K, Shibata T, et al. Phase III trial of upfront

debulking surgery versus neoadjuvant chemotherapy for stage III/IV ovarian,
tubal and peritoneal cancers: Japan Clinical Oncology Group Study
JCOG0602. Jpn J Clin Oncol. 2008;38:74-77.
21. Onda T, Kobayashi H, Nakanishi T, et al. Feasibility study of neoadjuvant chemotherapy followed by interval debulking surgery for stage III/IV
ovarian, tubal, and peritoneal cancers: Japan Clinical Oncology Group Study
JCOG0206. Gynecol Oncol. 2009;113:57-62.
22. Bristow RE, Eisenhauer EL, Santillan A, et al. Delaying the primary
surgical effort for advanced ovarian cancer: A systematic review of neoadjuvant chemotherapy and interval cytoreduction. Gynecol Oncol. 2007;104:480490.
UTERINE SARCOMA: CHALLENGING CASES FOR

THE INTERDISCIPLINARY TEAM
CHAIR
Martee L. Hensley, MD
New York, NY
SPEAKERS
Anuja Jhingran, MD
Houston, TX
John P. Curtin, MD
New York University School of Medicine
New York, NY
Uterine Sarcomas: Histology and Its

Implications on Therapy
By Martee L. Hensley, MD
Overview: Uterine sarcomas are rare cancers, they comprise

only 5% of all uterine malignancies. There are about 2,000
cases of uterine sarcoma diagnosed annually in the United
States. Uterine sarcomas may be categorized as either
favorable-risk, low-grade malignancies with a relatively good
prognosis or as poor-risk, high-grade cancers that carry a
high risk for tumor recurrence and disease progression.
Expert histologic review is critical for appropriate diagnosis
and management. Uterine sarcoma histologies considered to
carry a more favorable prognosis include low-grade endometrial stromal sarcomas and adenosarcomas. The high-grade
sarcomas include high-grade leiomyosarcomas, high-grade
undifferentiated endometrial sarcomas, and adenosarcomas
with sarcomatous overgrowth.
The favorable histology, low-grade uterine sarcomas may be

cured with surgical resection of uterus-limited disease. These
tumors are often hormone-sensitive, and treatment with
hormonal therapies may be efficacious for patients with advanced, unresectable disease. High-grade uterine leiomyosarcomas and undifferentiated endometrial sarcomas carry a
high risk for recurrence, even after complete resection of
uterus-limited disease. No adjuvant intervention has been
shown to improve survival outcomes. Advanced, metastatic
disease is generally treated with systemic cytotoxic therapies,
which may result in objective response but is not curative.
Selected patients with isolated metastatic disease and a long
disease-free interval may benefit from metastatectomy.
Low-Grade Endometrial Stromal Sarcomas
Responses to cytotoxic chemotherapy for advanced disease

would be expected to be low in ESS, due to its indolent
growth rate. Since endometrial stromal sarcomas frequently
express ER and PR, objective responses of advanced disease
to hormonal interventions, such as treatment with aromatase inhibitors, have been documented.11,12
NDOMETRIAL STROMAL sarcomas (ESS) are, by

definition, low-grade malignancies. Histologically they
have bland appearance, with few mitotic figures. Immunohistochemistry (IHC) stains for desmin, CD10, estrogen
receptor (ER), and progesterone receptor (PR) are typically
positive. Smooth muscle markers (h-caldesmon, smooth
muscle actin) are generally negative in ESS.1,2 A chromosomal translocation, (t(7;17)(p15;q21), which fuses two zinc
finger genes, JAZF1/JJAZ, has been described in the majority of ESS and may be useful for distinguishing ESS from
high-grade, undifferentiated endometrial sarcoma (HGUS)
and from leiomyosarcoma (LMS).3,4
Fifteen percent to 30% of patients with ESS may have
evidence of metastatic disease at the time of diagnosis, with
lung being the most common site for metastatic disease.
However, reflecting the low-grade, favorable behavior of this
tumor, five-year survival rates are 60% to 90% across all
stages of disease.5,6 There are no randomized trials assessing the influence of bilateral salpingo-oophorectomy (BSO)
on recurrence and survival in ESS. Some retrospective
studies have shown higher recurrence rates among patients
with retained ovaries.7 Surveillance, Epidemiology and End
Results (SEER) retrospective data did not show worse overall survival for women who did not undergo BSO, but this
study did not address recurrence rates, and the number of
patients with retained ovaries was very small.8 Lymph node
involvement has been reported to be found in zero to onethird of patients, and whether routine lymph node dissection
of normal-appearing nodes in ESS is necessary remains
controversial.8,9
For patients with uterus-limited, completely resected
ESS, there are no data to support routine adjuvant therapy.
Retrospective SEER data showed poorer overall survival
among patients who received adjuvant pelvic radiation
(80.1%) than among those who had surgery alone (90.7%).10
There is no role for adjuvant cytotoxic therapy in ESS, and
adjuvant hormonal treatment has not been prospectively
studied. It is reasonable to avoid estrogen replacement
therapy in patients with as diagnosis of ESS, although there
are no prospective randomized trials addressing this issue.
356
Adenosarcomas
Uterine adenosarcomas are low-grade malignancies that

arise most commonly in the uterine fundus. Histologically
they are characterized by a mixed histologic appearance that
contains benign-appearing glandular epithelial components
and low-grade endometrial stromal sarcoma.13 Experienced
histologic review is required to exclude the presence of
sarcomatous overgrowth (see below), the presence of which
portends a poor prognosis.
In the absence of sarcomatous overgrowth, adenosarcomas have a favorable prognosis, with 5-year survival rates
exceeding 90%.14 Adenosarcomas are rarer than endometrial stromal sarcomas, thus data regarding treatment are
very limited. Since the malignant portion of adenosarcomas
resembles endometrial stromal sarcoma, management recommendations are sometimes extrapolated from ESS data.
Like ESS, adenosarcomas commonly express ER, PR, and
CD1015; and, as with ESS, it is reasonable to perform
bilateral oophorectomy, and to avoid hormone replacement
therapy. Table 1 provides a summary of histologic features,
prognosis, and management issues for the favorable risk,
low-grade uterine sarcomas.
From the Memorial Sloan-Kettering Cancer Center, Weill Cornell Medical College, New
York, NY.
Address reprint requests to Author and Session Chair contact information: Martee L.
Hensley, MD, Associate Attending, Gynecologic Medical Oncology, Memorial SloanKettering Cancer Center, Associate Professor of Medicine, Weill Cornell Medical College,
300 E. 66th Street, Suite 1355, New York, NY 10065; email: hensleym@mskcc.org.
1092-9118/10/1-10
HISTOLOGIC FEATURES AND MANAGEMENT OF LOW- AND HIGH-GRADE UTERINE SARCOMAS

Table 1. Histologic Features and Management Summary for Favorable Histology, Low-Grade Uterine Sarcomas
Histologic Features
Endometrial Stromal
Sarcoma
Low mitotic rate, IHC positive

for desmin, CD10, ER, PR
JAZF1/JJAZ translocation
Surgical Issues
Prognosis
Hysterectomy BSO, resection

of suspicious lymph nodes;
lymph node dissection of
normal-appearing nodes
remains controversial
25%30% of patients with uteruslimited disease may recur.

5-year survival 60%90% across
all disease stages
Systemic Management Issues
Observation after resection of

uterus-limited disease
Limited data available

regarding adjuvant radiation
or hormonal treatment
Advanced, measurable disease
may respond to hormone
blockade
Adenosarcoma
Benign glandular epithelium

malignant stromal tissue
that resembles ESS, IHC
positive for ER, PR, CD10
Hysterectomy BSO; resection

of suspicious lymph nodes;
lymph node dissection of
normal-appearing nodes
remains controversial
5-year overall survival

approximately 90%
Expert histologic review to

exclude presence of
sarcomatous overgrowth
Limited data available

regarding adjuvant radiation
or hormonal treatment
Management strategies are
sometime extrapolated from
ESS
Abbreviations: IHC, immunohistochemistry; ER, estrogen receptor; PR, progesterone receptor; ESS, endometrial stromal sarcoma; BSO, bilateral salpingooophorectomy.
Histologic Features and Management of Poor-Risk,

High-Grade Uterine Sarcomas
Leiomyosarcomas
Leiomyosarcomas are generally considered high-grade

cancers. The grading of LMS remains controversial, however, and careful review of the histology with an expert
pathologist is recommended. Clinically, there are some leiomyosarcomas that may exhibit a more indolent disease
course, although it is not clear yet which histologic features
can definitively classify this subgroup.16 Most LMS is char-
KEY POINTS
Low-grade endometrial stromal sarcomas and adenosarcomas are hormone-sensitive tumors with a favorable prognosis.
High-grade uterine leiomyosarcomas have a high risk
for recurrence after resection of uterus-limited diseaseapproximately 50% at 3 years.
No adjuvant treatment has been shown to improve
outcomes following resection of uterus-limited highgrade sarcomathe standard approach following resection remains observation.
Cytotoxic treatments with efficacy in high-grade uterine leiomyosarcoma include doxorubicin with or without ifosfamide, ifosfamide, gemcitabine, and fixed
dose-rate gemcitabine plus docetaxel.
Data regarding management of high-grade, undifferentiated endometrial sarcomas are very limited.
Treatment of advanced disease is currently extrapolated from experience with leiomyosarcoma and other
soft tissue sarcomas.
acterized by the histologic presence of cytologic atypia,

coagulative necrosis, and a moderate to high mitotic rate.17
Smooth muscle markers such as h-caldesmon and smooth
muscle actin are typically positive by immunohistochemistry; p16, p53, and Ki-67 have been reported as useful for
distinguishing LMS from benign smooth muscle tumors.18
Estimates for the risk of recurrence after complete resection of uterus-limited (the International Federation of Gynecology and Obstetrics [FIGO] stage I-limited to the
uterine fundus or stage II-uterine fundus and cervix) uterine
leiomyosarcoma are variable, ranging from 60% to 70% at 2
years19,20 in retrospective studies, to approximately 50% at
3 years in a prospective study.21 The aggressive nature of
this cancer is reflected in the poor overall survival rates. In
one large study the 5-year survival rate was 51% among
patients with FIGO stage I LMS, and only 25% among those
with FIGO stage II LMS.14 FIGO and the American Joint
Committee on Cancer (AJCC) staging systems have been
shown to perform poorly in terms of providing accurate
estimates of overall survival.22 A nomogram that includes
age, tumor size, histologic grade, mitotic index, extrauterine
spread, and distant metastases provides better survival
estimates.23
No form of adjuvant treatment has yet been shown to
improve overall survival or progression-free survival. Retrospective studies suggested that patients who received pelvic
radiation may have fewer local recurrences, however, there
was no improvement in overall survival.24,25 Furthermore, a
prospective randomized trial of adjuvant pelvic radiation
compared to observation did not show benefit for patients
with uterus-limited LMS in terms of local recurrence or
survival.21 One prospective trial of adjuvant doxorubicin for
patients with a variety of uterine sarcoma histologies did not
show benefit (recurrence rate 41% among patients assigned
to doxorubicin compared with 53% among patients assigned
to observation).26 Adjuvant chemotherapy using fixed dose-
357
MARTEE L. HENSLEY
Table 2. Chemotherapy Agents with Activity in Advanced Uterine Leiomyosarcoma
Reference
Sutton, 1992
Omura, 1983
Look, 2003
Sutton, 1996
Hensley, 2008
Hensley, 2008
Talbot, 2003
Ferris, 2010
Anderson, 2005
Monk, 2012
Study Design
Prospective
Prospective
Prospective
Prospective
Prospective
Prospective
Prospective
phase II
phase III
phase II, second line
phase II, first line
phase II, second line
phase II, soft tissue sarcomas
Retrospective, small series

Prospective phase II, first line, uterine LMS
rate gemcitabine plus docetaxel, followed by doxorubicin

was tested in a single-arm phase II trial for patients with
uterus-limited, high-grade LMS. Although 78% of patients
were disease free at 2 years, only 52% remained disease free
at 3 years.27 An international, prospective phase III trial
comparing adjuvant chemotherapy to observation for
women with uterus-limited, high-grade LMS is under development.
Median survival after a diagnosis of metastatic or recurrent leiomyosarcoma is less than 12 months. Patients with
isolated, resectable disease with a long disease-free interval
may be candidates for metastatectomy. Reported outcomes
represent highly selected patients.28,29 Treatment of multisite, unresectable disease is with systemic cytotoxic therapy.
Response rates for single agents in LMS include ifosfamide
(17%),30 doxorubicin (25%),31 gemcitabine (20%),32 temozolomide (approximately 4% to 20%, based on limited
data)33,34,35,36 and trabectedin (8% among patients with
prior treatment, 10% in first-line treatment of uterine LMS,
up to 17% in some subsets of patients).37,38,39 Commonly
used combination chemotherapy regimens with activity include fixed dose-rate gemcitabine plus docetaxel (response
rate 36% as first-line therapy,40 27% as second-line therapy
in uterine LMS41) and doxorubicin plus ifosfamide (response
rate 30%).42 Table 2 provides a summary of agents with
activity in uterine LMS.
The role of targeted therapies in uterine LMS remains to
be elucidated. Retrospective data representing a cohort of
selected patients with small volume, ER- and/or PR-positive
uterine LMS showed response to treatment with aromatase
inhibition in 9% of patients.43 Data supporting the use of
vascular endothelial targeted agents are limited. Use of
such agents is discouraged outside of a clinical trial. A
prospective phase II study of doxorubicin plus bevacizumab
for patients with previously untreated, metastatic soft tissue
sarcoma (40% of whom had uterine LMS) resulted in a
lower-than-expected response rate of only 12%, and a high
rate of cardiac dysfunction (35% with grade 2 or worse
cardiac toxicity).44 The multitargeted kinase inhibitor
sunitinib failed to achieve objective responses or disease
stabilization as second- or third-line therapy in uterine
LMS.45 Sorafenib was similarly disappointing in LMS of
uterine or nonuterine origin.46 Whether the vasculartargeted agent bevacizumab can augment the response rate
and time to disease progression in advanced uterine LMS
when used in combination with fixed dose-rate gemcitabine
plus docetaxel is being investigated in a phase III, placebocontrolled trial.47
358
Agent
Response Rate
Ifosfamide
Doxorubicin
Gemcitabine
Doxorubicin ifosfamide
Fixed dose-rate gemcitabine docetaxel
17%
25%
20%
30%
36%
27%
4%20%
Temozolomide
Trabectedin
10%
High-Grade, Undifferentiated Endometrial Sarcomas
High-grade, undifferentiated endometrial sarcomas were

previously called high-grade ESS in order to distinguish
them from low-grade ESS. However, since these tumors lack
histologic features that resemble endometrial stroma, the
preferred term is now high-grade, undifferentiated endometrial sarcoma (HGUS). These tumors have a pleomorphic,
undifferentiated appearance, do not express epithelial or
smooth muscle markers, and are typically ER and PR
negative by immunohistochemistry. A recently reported
t(10;17) genomic rearrangement which yields an oncoprotein
is present in the majority of HGUS, a finding that may have
diagnostic and perhaps therapeutic potential.48 Since lymph
node involvement at time of diagnosis is common, lymph
node dissection is often performed, however, it is not known
whether this leads to improved outcomes given the overall
aggressive behavior of this cancer.
High-grade undifferentiated endometrial sarcomas have a
poor prognosis, with 5 year survival rates of approximately
25%, and a high risk for disease recurrence after resection of
uterus-limited disease.49,50 While retrospective data suggest
that adjuvant pelvic radiation decreases local recurrence
rates, the number of patients in such retrospective reports is
small.51,52 There are no randomized trials assessing the role
of pelvic radiation, and no data showing improvements in
progression-free or overall survival.
There have been no prospective trials of systemic therapy
specifically for high grade undifferentiated endometrial sarcomas. Retrospective data, and subsets of HGUS patients
enrolled in soft tissue sarcoma studies suggest that HGUS
may respond to doxorubicin or ifosfamide-based regimens.53,54 Efforts should be made to enroll these patients on
clinical trials testing new agents for sarcomas.
Adenosarcomas with Sarcomatous Overgrowth
Uterine adenosarcomas with sarcomatous overgrowth

(ASSO) represent a small subset of adenosarcomas. In
contradistinction from adenosarcomas, in which the malignant portion of the tumor has the bland appearance of low
grade ESS, the sarcomatous portion of ASSO has cellular
atypia, a higher mitotic rate, higher ki-67 staining, and may
have the appearance of nonuterine, high-grade sarcomas
such as rhabdomyosarcoma or chondrosarcoma or others.
The sarcomatous overgrowth portion of these tumors is
generally CD10 and ER-negative, although PR may be
positive.
ASSO are more likely than adenosarcomas without sarcomatous overgrowth to have evidence of metastatic disease at

Table 3. Histologic Features and Management Summary for Poor-Risk, High-Grade Uterine Sarcomas
Histologic Features
Leiomyosarcoma
Cytologic atypia, coagulative

necrosis, high mitotic rate;
Surgery for Early-stage Disease
Prognosis
Systemic Management Issues
Hysterectomy BSO; lymph

node dissection not necessary
for normal-appearing lymph
nodes
For uterus-limited LMS:

approximately 50%
recurrence by 3 year;
5-year overall survival
25%50%; median
survival from diagnosis
of metastatic disease is
1 year, but range may
be wide
For uterus-limited disease:

neither pelvic radiation nor
adjuvant chemotherapy has
been shown to improve PFS or
OS
For advanced, measurable
disease: cytotoxic therapy may
achieve objective responses in
approximately 30% of patients
IHC positive for smooth muscle

markers; may be positive
for ER, PR
High-grade, undifferentiated
endometrial sarcoma
Pleomorphic, undifferentiated
cells; IHC negative for
smooth muscle markers,
negative for ER, PR; t(10;
17) genomic rearrangement
(YWHAE/FAM22)
Hysterectomy BSO; may

include lymph node dissection
(limited data)
Limited prospective data;

risk for recurrence
greater than 50%;
survival with metastatic
disease is poor
No prospective data regarding

adjuvant therapy for completely
resected disease
systemic treatment specifically
for HGUS; enrollment in
clinical trials for soft tissue
sarcoma encouraged
Adenosarcoma with
sarcomatous overgrowth
Sarcomatous portion has

atypia, heterologous
elements resembling
rhabdomyosarcoma,
chondrosarcoma, or others;
CD10 negative; ER negative
Hysterectomy BSO; may

include lymph node dissection
(limited data)
Approximately 30% have

metastatic disease at
diagnosis; median
survival approximately
13 months

adjuvant therapy for completely
resected disease
No prospective data for
treatment of advanced ASSO;
enrollment on clinical trials for
soft tissue sarcoma encouraged
Abbreviations: IHC, immunohistochemistry; ER, estrogen receptor; PR, progesterone receptor; BSO, bilateral salpingo-oophorectomy; PFS, progression-free survival;
OS, overall survival; ASSO, adenosarcoma with sarcomatous overgrowth.
the time of diagnosis. The disease course is dictated by the

high-grade sarcoma portion of the tumor. Survival rates for
patients with metastatic disease are poor: median survival
was 13 months in one study, and the 5-year overall survival
rate was 43% in another.14,55
There are no prospective data to guide management
decisions for patients with ASSO. Patients are at risk for
both local and distant recurrences, but it is not known
whether any adjuvant strategy can diminish that risk. For
patients with recurrent and/or measurable metastatic disease, systemic treatment approaches are reasonable, although the choice of agents to use must be extrapolated from
soft tissue sarcoma data. Enrollment in clinical trials for soft
tissue sarcoma should be encouraged. Table 3 provides a
summary of histologic features, prognosis, and management
issues for poor-risk, high-grade uterine sarcomas.
Carcinosarcomas
Carcinosarcomas are uterine tumors containing both a

malignant epithelial component (i.e., adenocarcinoma) and a
malignant mesenchymal component (which may resemble
uterine tissue, in which case it is called homologous; or
non-uterine tissue, in which case it is called heterologous).
These tumors are sometimes considered a subtype of uterine

sarcomas, and sometimes considered to be high grade endometrial cancers. The constraints of this publication preclude
a complete discussion of these malignancies in this manuscript. For further information, see other publications.56
Critical Research Questions
Although uterine sarcomas are rare, adequately powered,

well-designed studies are needed in these diseases. An
international, randomized phase III trial is under development to address whether adjuvant chemotherapy can improve survival outcomes among women with completely
resected uterus-limited LMS. Large collaborative trials are
needed to define the role of hormonal treatment for ESS,
both as adjuvant therapy and for advanced disease. New
agents are needed for advanced leiomyosarcoma, and prospective data are needed to define whether current agents
have activity in high grade undifferentiated endometrial
sarcomas. Correlative studies are needed to identify molecular drivers of these tumors, and clinical trials will be
needed to determine whether these drivers may serve as
effective targets for treatment.
Author
Martee L. Hensley
Employment or
Leadership
Positions
Sanofi (I)
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
359
MARTEE L. HENSLEY
REFERENCES
1. McCluggage WG, Sumathi VP, Maxwell P. CD10 is a sensitive and
diagnostically useful immunohistochemical marker of normal endometrial
stroma and of endometrial stromal neoplasms. Histopathology. 2001;39:273-278.
2. Nucci MR, OConnell JT, Huettner PC, et al. h-Caldesmon expression
effectively distinguishes endometrial stromal tumors from uterine smooth
muscle tumors. Am J Surg Pathol. 2001;25:455-463.
3. Koontz JI, Soreng AL, Nucci M, et al. Frequent fusion of the JAZF1 and
JJAZ1 genes in endometrial stromal tumors. Proc Natl Acad Sci U S A.
2001;98:6348-6353.
4. Micci F, Walter CU, Teixeira MR, Panagopoulos I, Bjerkehagen B,
Saeter G, Heim S. Cytogenetic and molecular genetic analyses of endometrial
stromal sarcoma: nonrandom involvement of chromosome arms 6p and 7p and
confirmation of JAZF1/JJAZ1 gene fusion in t(7;17). Cancer Genet Cytogenet.
2003;144:119-124.
5. Bodner K, Bodner-Adler B, Obermair A, Windbichler G, Petru E,
Mayerhofer S, Czerwenka K, Leodolter S, Kainz C, Mayerhofer K. Prognostic
parameters in endometrial stromal sarcoma: a clinicopathologic study in 31
patients. Gynecol Oncol. 2001;81:160-165.
6. Gadducci A, Sartori E, Landoni F, et al. Endometrial stromal sarcoma:
analysis of treatment failures and survival. Gynecol Oncol. 1996;63:247-253.
7. Berchuck A, Rubin SC, Hoskins WJ, et al. Treatment of endometrial
stromal tumors. Gynecol Oncol. 1990;36:60-65.
8. Shah JP, Bryant CS, Kumar S, et al. Lymphadenectomy and ovarian
preservation in low-grade endometrial stromal sarcoma. Obstet Gynecol.
2008;112:1102-1108.
9. Dos Santos LA, Garg K, Diaz JP, et al. Incidence of lymph node and
adnexal metastasis in endometrial stromal sarcoma. Gynecol Oncol. 2011;
121:319-322.
10. Barney B, Tward JD, Skidmore T, et al. Does radiotherapy or lymphadenectomy improve survival in endometrial stromal sarcoma? Int J Gynecol
Cancer. 2009;19:1232-1238.
11. Maluf FC, Sabbatini P, Schwartz L, et al. Endometrial stromal sarcoma: objective response to letrozole. Gynecol Oncol. 2001;82:384-388.
12. Chu MC, Mor G, Lim C, et al. Low-grade endometrial stromal sarcoma:
hormonal aspects. Gynecol Oncol. 2003;90:170-176.
13. Verschraegen CF, Vasuratna A, Edwards C, et al. Clinicopathologic
analysis of mullerian adenosarcoma: The M. D. Anderson Cancer Center
experience. Oncol Rep. 1998;5:939-944.
14. Abeler VM, Ryne O, Thoresen S, et al. Uterine sarcomas in Norway. A
histopathological and prognostic survey of a total population from 1970 to
2000 including 419 patients. Histopathology. 2009;54:355-364.
15. Soslow RA, Ali A, Oliva E. Mullerian adenosarcomas: an immunophenotypic analysis of 35 cases. Am J Surg Pathol. 2008;32:1013-1021.
16. Veras E, Zivanovic O, Jacks L, et al. Low-grade leiomyosarcoma and
late-recurring smooth muscle tumors of the uterus: a heterogenous collection
of frequently misdiagnosed tumors associated with an overall favorable
prognosis relative to conventional uterine leiomyosarcomas. Am J Surg
Pathol. 2011;35:1626-1637.
17. Bell SW, Kempson RL, Hendrickson MR. Problematic uterine smooth
muscle neoplasms. A clinicopathologic study of 213 cases. Am J Surg Pathol.
1994;18:535-558.
18. Chen L, Yang B. Immunohistochemical analysis of p16, p53, and Ki-67
expression in uterine smooth muscle tumors. Int J Gynecol Pathol. 2008;27:
326-332.
19. Major FJ, Blessing JA, Silverberg SG, et al. Prognostic factors in
early-stage uterine sarcoma. A Gynecologic Oncology Group study. Cancer.
1993;71:1702-1709.
20. Nordal RR, Kristense GB, Kaern J, et al. The prognostic significance of
stage, tumor size, cellular atypia and DNA ploidy in uterine leiomyosarcoma.
Acta Oncol. 1995;34:797-802.
21. Reed NS, Mangioni C, Malmstrom H, et al. Phase III randomised study
to evaluate the role of adjuvant pelvic radiotherapy in the treatment of
uterine sarcomas stages I and II: an European Organisation for Research and
Treatment of Cancer Gynaecological Cancer Group Study (protocol 55874).
Eur J Cancer. 2008;44:808-818
22. Zivanovic O, Leitao MM, Iasonos A, et al. Stage-specific outcomes of
patients with uterine leiomyosarcoma: a comparison of the international
Federation of gynecology and obstetrics and american joint committee on
cancer staging systems. J Clin Oncol. 2009;27:2066-2072.
23. Zivanovic O, Jacks LM, Iasonos A, et al. A nomogram to predict
postresection 5-year overall survival for patients with uterine leiomyosarcoma. Cancer. 2012;118:660-669.
24. Giuntoli RL, Metzinger DS, DiMarco CS, et al. Retrospective review of
208 patients with leiomyosarcoma of the uterus: Prognostic indicators,
360
surgical management, and adjuvant therapy. Gynecol Oncol. 2003;89:460469.

25. Hornback NB, Omura G, Major FJ. Observations on the use of adjuvant
radiation therapy in patients with stage I and II uterine sarcoma. Int J
26. Omura GA, Blessing JA, Major F, et al. A randomized clinical trial of
adjuvant adriamycin in uterine sarcomas: a Gynecologic Oncology Group
Study. J Clin Oncol. 1985;3:1240-1245.
27. Hensley ML, Wathen K, Maki RG, et al. 3-year follow-up of SARC 005:
adjuvant treatment of high risk primary uterine leiomyosarcoma with gemcitabine/docetaxel (GT) followed by doxorubicin (D). Connective Tissue Oncology Society, Abstract ID: 1130375, 2011.
28. Leitao MM, Brennan MF, Hensley M, et al. Surgical resection of
pulmonary and extrapulmonary recurrences of uterine leiomyosarcoma. Gynecol Oncol. 2002;87:287-294.
29. Clavero JM, Deschamps C, Cassivi SD, et al. Gynecologic cancers:
factors affecting survival after pulmonary metastasectomy. Ann Thorac Surg.
2006;81:2004-2007.
30. Sutton GP, Blessing JA, Barrett RJ, et al. Phase II trial of ifosfamide
and mesna in leiomyosarcoma of the uterus: a Gynecologic Oncology Group
study. Am J Obstet Gynecol. 1992;166:556-559.
31. Omura GA, Major FJ, Blessing JA, et al. A randomized study of
adriamycin with and without dimethyl triazenoimidazole carboxamide in
advanced uterine sarcomas. Cancer. 1983;52:626-632.
32. Look KY, Sandler A, Blessing JA, et al. Phase II trial of gemcitabine as
second-line chemotherapy of uterine leiomyosarcoma: a Gynecologic Oncology
Group (GOG) study. Gynecol Oncol. 2004;92:644-647.
33. Talbot SM, Keohan ML, Hesdorffer M, et al. A phase II trial of
temozolomide in patients with unresectable or metastatic soft tissue sarcoma.
Cancer. 2003;98:1942-1946.
34. Boyar M, Hesdorffer M, Keohan M, et al. Phase II study of temozolomide and thalidomide in patients with unresectable or metastatic leiomyosarcoma. J Clin Oncol. 2005;23 (abstract 9029).
35. Ferriss JS, Atkins KA, Lachance JA, et al. Temozolomide in advanced
and recurrent uterine leiomyosarcoma and correlation with o6methylguanine DNA methyltransferase expression: A case series. Int J
Gynecol Cancer. 2010;20:120-125.
36. Anderson S, Aghajanian C. Temozolomide in uterine leiomyosarcomas.
Gynecol Oncol. 2005;98:99-103.
37. Garcia-Carbonero R, Supko JG, Maki RG, et al. Ecteinascidin-743
(ET-743) for chemotherapy-naive patients with advanced soft tissue sarcomas: multicenter phase II and pharmacokinetic study. J Clin Oncol. 2004;24:
5484-5492.
38. Delaloge S, Yovine A, Taamma A, et al. Ecteinascidin-743: a marinederived compound in advanced, pretreated sarcoma patients-preliminary
evidence of activity. J Clin Oncol. 2001;19:1248-1255.
39. Monk BJ, Blessing JA, Street DG, et al. A phase II evaluation of
trabectedin in the treatment of advanced, persistent, or recurrent uterine
leiomyosarcoma: a gynecologic oncology group study. Gynecol Oncol. 2012;
124:48-52.
40. Hensley ML, Blessing JA, Mannel R, et al. Fixed-dose rate gemcitabine
plus docetaxel as first-line therapy for metastatic uterine leiomyosarcoma: a
Gynecologic Oncology Group phase II trial. Gynecol Oncol. 2008;109:329-334.
41. Hensley ML, Blessing JA, DeGeest K, et al. Fixed-dose rate gemcitabine plus docetaxel as second-line therapy for metastatic uterine leiomyosarcoma: a Gynecologic Oncology Group phase II study. Gynecol Oncol.
2008;109:323-328.
42. Sutton G, Blessing JA, Malfetano JH. Ifosfamide and doxorubicin in the
treatment of advanced leiomyosarcomas of the uterus: a Gynecologic Oncology Group study. Gynecol Oncol. 1996;62:226-229.
43. OCearbhaill R, Zhou Q, Iasonos A, et al. Treatment of Advanced Uterine
Leiomyosarcoma with Aromatase Inhibitors. Gynecol Oncol. 2010;116:424-429.
44. DAdamo DR, Anderson SE, Albritton K, et al. Phase II study of
doxorubicin and bevacizumab for patients with metastatic soft-tissue sarcomas. J Clin Oncol. 2005;23:7135-7142.
45. Hensley ML, Sill MW, Scribner DR Jr, et al. Sunitinib malate in the
treatment of recurrent or persistent uterine leiomyosarcoma: a Gynecologic
Oncology Group phase II study. Gynecol Oncol. 2009;115:460-465.
46. Maki RG, DAdamo DR, Keohan ML, et al. Phase II study of sorafenib
in patients with metastatic or recurrent sarcomas. J Clin Oncol. 2009;27:
3133-3140.
47. Hensley ML. Principal Investigator. (GOG 0250, clinical trial identifier
NCT01012297).

48. Lee CH, Ou WB, Marino-Enriquez A, et al. 14-3-3 fusion oncogenes in
high-grade endometrial stromal sarcoma. Proc Natl Acad Sci U S A. 2012 Jan
5. [Epub ahead of print]
49. Nordal RR, Kristensen GB, Kaern J, et al. The prognostic significance
of surgery, tumor size, malignancy grade, menopausal status, and DNA ploidy
in endometrial stromal sarcoma. Gynecol Oncol. 1996;62:254-259.
50. Gadducci A, Sartori E, Landoni F, et al. Endometrial stromal sarcoma:
analysis of treatment failures and survival. Gynecol Oncol. 1996;63:247-253.
51. Berchuck A, Rubin SC, Hoskins WJ, et al. Treatment of endometrial
stromal tumors. Gynecol Onco. 1990;36:60-65.
52. Piver MS, Rutledge FN, Copeland L, et al. Uterine endolymphastic
stromal myosis: A collaborative study. Obstet Gynecol 1995;64:173.
53. Sutton G, Blessing JA, Park R, et al. Ifosfamide treatment of recurrent

or metastatic endometrial stromal sarcomas previously unexposed to chemotherapy: a study of the Gynecologic Oncology Group. Obstet Gynecol. 1996;
87:747-750.
54. Yamawaki T, Shimizu Y, Hasumi K. Treatment of stage IV highgrade endometrial stromal sarcoma with ifosfamide, adriamycin, and cisplatin. Gynecol Oncol. 1997;64:265-269.
55. Krivak TC, Seidman JD, McBroom JW, et al. Uterine adenosarcoma
with sarcomatous overgrowth versus uterine carcinosarcoma: comparison of
treatment and survival. Gynecol Oncol. 2001;83:89-94.
56. Acharya S, Hensley ML, Montag AC, et al. Rare Uterine Cancers.
Lancet Oncology. 2005;6:961-971.
361
Surgical Options for Recurrent

Uterine Sarcomas
By Sharmilee B. Korets, MD, and John P. Curtin, MD
Overview: Leiomyosarcoma, the most frequent pure uterine

sarcoma, is an aggressive tumor with a tendency toward early
relapse. Survival for patients with recurrent disease is poor.
In contrast, endometrial stromal sarcoma, the second most
common uterine sarcoma, is a more indolent malignancy with
a tendency toward recurrence after a long latency period. The
relative infrequency of both diseases makes the study and
standardization of treatment for recurrent disease challenging. Treatment of recurrence with cytotoxic chemotherapy,
radiation therapy, or hormone therapy produces modest to
poor response rates. Surgical resection is one treatment
modality offering the potential for cure and perhaps a more
durable response than is seen with medical management.
Although initial studies focused on pulmonary metastasec-
tomy in recurrent soft tissue sarcoma, an increasingly large

body of data specifically evaluating outcomes after both
thoracic and extrathoracic metastasectomy in patients with
recurrent uterine sarcoma is now available. Though no prospective trials have been conducted, retrospective comparisons of chemotherapy or radiation therapy with surgery for
recurrent uterine sarcoma suggest improvement in diseasespecific survival for the surgery group. Clearly defined factors
are associated with better prognosis after surgical resection
of recurrence, including a prolonged disease-free interval
and complete resection of disease. In properly selected
women, surgery and even repeated metastasectomy for
recurrent disease may improve survival and should be considered.
Case History: An otherwise healthy, middle-age woman was

diagnosed with leiomyosarcoma at the time of a myomectomy
for symptomatic fibroids in 2000. Six weeks later, she underwent a total abdominal hysterectomy with bilateral salpingooophorectomy for leiomyosarcoma and was diagnosed with
stage IB disease. Surgery was followed by three cycles of
adjuvant chemotherapy. She had an initial disease-free interval of 7.5 years. However, in fall 2008, she was diagnosed
with a large left upper lobe thoracic metastasis and a
synchronous left acetabular lesion. She underwent a videoassisted thoracoscopic left upper lobectomy and mediastinal
lymph node dissection for a 5 cm mass. All pulmonary
disease was completely resected and lymph nodes were negative for disease. Three months later, she had intermittent
abdominal discomfort and vague bowel symptoms and was
noted to have a jejunal mass that was suspected to be
recurrent disease. The mass was completely removed by
small bowel resection with reanastomosis. At the time of
surgery, there was no evidence of residual intra-abdominal
disease. She had resection of the acetabular lesion in spring
2009. At that time, she was thought to be free of disease.
However within 1 month, she a calcaneal metastasis was
found, and after a course of radiation therapy, systemic
chemotherapy was initiated. She was treated with multiple
chemotherapy regimens over the course of the next 16 months.
She then chose to pursue palliative treatment and died
approximately 10 years after the initial diagnosis and 30
months after the initial surgical resection for recurrence.
This case illustrates several key points in our review of
surgical management of recurrent uterine sarcoma and provides illustrative radiographic images of resectable metastases in uterine sarcoma (Fig. 1).
leiomyosarcoma, which accounts for the majority of uterine

sarcomas, followed by endometrial stromal sarcoma (ESS,
previously called low-grade endometrial stromal sarcoma),
undifferentiated endometrial sarcoma (previously called
high-grade endometrial stromal sarcoma), and adenosarcoma.
Even early-stage uterine sarcomas demonstrate aggressive clinical behavior and confer a poor prognosis. Leiomyosarcomas have a propensity toward hematogenous spread
and early recurrence. Five-year survival rates range from
30% to 48%, and relapse rates approach 60%, with 42% of
relapsed disease occurring outside the pelvis.2 Most extrapelvic relapses occur in the lung. As with other soft tissue
sarcomas, leiomyosarcomas are relatively chemoresistant,3
making treatment of recurrent disease challenging. In contrast, ESS is more indolent, with a tendency toward local or
distant relapse after a long latency period. Five-year survival rates are between 80% and 100%,1 the median time to
recurrence in women with stage I disease is 65 months, and
the rate of relapse ranges from 36% to 56%.4 Recurrences
are primarily pelvic, intra-abdominal, or pulmonary;
however, intravascular, cardiac, and central nervous
system metastases have also been reported. Because of
their slow growth, repeated resection may be warranted,
and even secondary and tertiary cytoreductions likely improve prognosis.4 Adenosarcomas are similarly indolent,
with excellent survival when disease is at an early stage at
the time of diagnosis and there is a long latency period
before relapse.
The relative rarity of uterine sarcomas makes clinical
investigation difficult and prospective randomized trials
nearly impossible, especially for evaluating the management of recurrent disease. Most studies are retrospective,
and various primary disease sites and histologic subtypes
ALIGNANT MESENCHYMAL tumors of the uterus

are rare, accounting for less than 3% of all uterine
malignancies. The annual incidence of uterine sarcoma
approaches two per 100,000 women.1 Recent changes in
terminology and classification now exclude carcinosarcomas
from this group, as the biology and clinical behavior of those
tumors point toward an epithelial origin. This review focuses on surgical management of recurrence in the most
common pure uterine sarcomas, listed in order of incidence:
362
From the Division of Gynecologic Oncology, New York University School of Medicine, New
York, NY.
Address reprint requests to John P. Curtin, MD, NYU Clinical Cancer Center, 160 E. 34th
St., 4th Floor, New York, NY 10016; email: john.curtin@med.nyu.edu.
1092-9118/10/1-10
SURGICAL OPTIONS FOR UTERINE SARCOMAS
have been combined in order to attain sufficient power.

Thus, optimal management strategies for recurrent disease
and definition of favorable clinicopathologic characteristics
for secondary cytoreduction or metastasectomy are not well
established. The purpose of this chapter is to better characterize candidates for resection of recurrent disease and to
discuss factors associated with better prognosis.
Treatment Options for Recurrent Disease
Survival for patients with recurrent disease is poor. Systemic therapy is often recommended; in ESS, hormone
therapy is frequently used, whereas in leiomyosarcoma,
cytotoxic chemotherapy may be the treatment of choice. In
uterine leiomyosarcoma, doxorubicin is the most active
single agent, with a response rate of 25%.5 Combination
therapies containing doxorubicin and ifosfamide and gemcitabine and docetaxel also have demonstrated activity in
metastatic or recurrent disease.6,7
External-beam radiation therapy has been used in the
adjuvant setting for uterine sarcoma, and most studies have
demonstrated an improved local control rate without a
substantial effect on overall survival. Radiation therapy
may have a role in the palliative treatment of distant
metastases.1 However, of all possible treatments for recurrent disease, including cytotoxic chemotherapy, hormone
therapy, radiation therapy, or surgery, only surgery is
associated with a high cure rate and prolonged survival.8
Surgical Resection of Metastatic Disease
In a single-institution retrospective cohort study of 33

patients with recurrent or metastatic uterine leiomyosarcoma, survival was improved for patients who had surgical resection of metastases compared with patients who
did not have surgery, with a median overall survival 45
compared with 31 months, respectively.9 In a larger study
(128 patients), survival after secondary cytoreduction for
recurrent uterine leiomyosarcoma was compared with survival after chemotherapy and/or radiation therapy for recur-
KEY POINTS
Recurrent uterine sarcoma is an aggressive disease in

which nonsurgical treatment modalities offer limited
benefit.
Due to the relative rarity of uterine sarcoma, few
prospective data on optimal management of recurrent disease are available.
Retrospective data demonstrate a survival benefit for
well-selected women in whom surgery is performed
for the management of recurrent disease.
Although initial data suggested that surgery for metastastic disease should be limited to pulmonary
metastasectomy only, additional studies have shown
that extrathoracic metastasectomy offers a similar
survival advantage.
Prolonged disease-free interval and complete resection of tumor correlate with improved outcome for
patients undergoing surgery for metastatic uterine
sarcoma.
rent disease. No distinction was made between abdominal,

pelvic, or thoracic procedures for resection. Disease-specific
survival as well as overall survival were substantially improved for the surgery group compared with the medical
management group, with a mean overall survival of 2.0
compared with 1.1 years from the time of recurrence.10
Due to the rarity of uterine sarcoma, much of the data
regarding surgical management of recurrence come from
studies in which heterogeneous histologic subtypes and
primary disease sites are pooled. Because the lung is a
preferred site of recurrence for soft tissue sarcoma, a large
body of literature focuses on indications for pulmonary metastasectomy and factors affecting survival after resection.
Pulmonary Metastasectomy
Indications for pulmonary metastasectomy in recurrent

uterine sarcoma can be extrapolated from studies of the
procedure for patients with all types of soft tissue sarcomas.
In the surgical and thoracic oncology literature, indications
for pulmonary metastasectomy include medical suitability
for surgery, sufficient pulmonary reserve to tolerate loss of
lung capacity, control of disease at the primary site, no
evidence of extrapulmonary disease, and no better therapy
available.11 Pulmonary metastasectomy with curative intent is widely accepted for well-selected patients with sarcomatous lung metastases.3,12
Thoracic procedures for metastasectomy range from thoracoscopic wedge resection to open pneumonectomy or bilobectomy via median sternotomy. Wedge resection is
performed in most women undergoing surgery for sarcomatous lung metastases.11,13,14 In a series from Brigham and
Womens hospital, approximately 75% of patients who had
pulmonary metastasectomy for recurrent leiomyosarcoma
were treated with wedge resection. In that study, it was also
demonstrated that disease-specific and overall survival were
similar for patients who underwent video-assisted thoracic
surgery (VATS) compared with thoracotomy or sternotomy.
Because recurrent lung metastases will develop in many
women, procedures that preserve the ability to tolerate
repeated resections, such as VATS and wedge resection, are
preferred.13
In multiple series, 34 to 48% of women who underwent
initial pulmonary metastasectomy developed recurrent lung
metastasis and had repeat resection.13-15 Sixteen to thirtyseven percent of those patients had a tertiary resection.14,15
Repeated metastasectomies correlate with improved survival, likely because tumor biology is more favorable in
patients who survive long enough for more than one metastasectomy.16
Extrathoracic Metastasectomy
More recently, outcomes after extrathoracic metastasectomy have been evaluated in studies of recurrent sarcoma.13,16,17 For patients undergoing resection of recurrent
uterine leiomyosarcoma, survival associated with thoracic
metastasectomy is similar to that associated with nonthoracic procedures for recurrent disease, with a median
disease-specific survival of nearly 4 years from the time of
first metastasectomy.17 A long-term survival benefit may be
conferred by metastasectomy for patients with extrapulmonary disease in the case of complete resection. These data
suggest that the traditional criteria for pulmonary metasta-
363
reports and small series have documented extensive resections for recurrent ESS with inferior vena cava and intracardiac extension.21,22 Preoperatively, imaging studies such
as CT, magnetic resonance imaging, and transesophageal
ultrasound or echocardiography can delineate the extent of
disease. Depending on the findings of these studies, the
surgery may be done by laparotomy, thoracotomy, or combined sternolaparotomy. In one series of 19 patients, cardiopulmonary bypass was required during seven procedures.
When reconstruction of major vascular structures such as
the inferior vena cava was necessary, xenopericardium and
graft replacements were used. In this series, a radical
resection resulting in complete tumor removal was possible
in 10 patients, some of whom required concurrent surgical
procedures for synchronous metastases, including pulmonary metastasectomy or pelvic extenteration. The median
survival was 2 years, with a range of 0.3 to 4.5 years.21
Because of the overall excellent prognosis in completely
resected ESS and the likelihood of imminent heart failure or
pulmonary tumor embolism in women with intracaval or
intracardiac extension, surgical excision is considered appropriate for well-selected patients.22,23
Resection of Recurrent Adenosarcoma
Fig. 1. PET CT of synchronous lung and acetabular metastatic
lesions, managed surgically after an initial 7.5-year disease-free
interval.
Abbreviation: PET CT, positron emission tomography/computerized
tomography.
sectomy, listed earlier, should be expanded to encourage

consideration of pulmonary metastasectomy for patients
who have either synchronous or prior resectable metastases
outside of the chest.16 This point is illustrated by our case
history. Figure 1 is a positron emission tomography/computerized tomography (CT) demonstrating the two sites of
recurrent disease at the time of the initial metastasectomy.
Liver Metastasectomy
Resection is frequently performed for liver metastases

resulting from colorectal cancer. However, the role of surgery is less clear in the case of sarcomatous liver metastases.
Due to the paucity of cases, few studies specifically address
liver metastasectomy in uterine sarcoma. However, several
small studies of heterogeneous patient populations with
metastatic leiomyosarcoma have demonstrated that in appropriately chosen patients, liver resection for metastatic
disease can prolong survival.18,19 In one small series of 66
patients who underwent resection, resection with radiofrequency ablation, or radiofrequency ablation alone, the median overall survival after the procedure was 47 months.
Longer survival was associated with metastases 3 cm or
smaller and with resection alone compared with radiofrequency ablation with or without surgical resection.19 A
study of 11 patients demonstrated a median survival of 39
months after resection and improved survival associated
with complete resection of metastatic disease.18
Inferior Vena Cava Resection or Intracardiac
Metastasectomy for Recurrent ESS
Vascular extension is a common characteristic of ESS.

Inferior vena cava tumor thrombus likely begins as tumor
growth within the uterine or ovarian veins.20 Multiple case
364
Few data are available on resection for the treatment of

recurrent adenosarcoma. In one small study of 23 women,24
17 had resection of recurrence in the vagina, pelvis, or
abdomen. Of these women, eight had a durable and possibly
curative response, with disease-free intervals of 5 to 12
years.
Survival after Surgery for Recurrent Disease
Survival after resection for recurrence of uterine sarcoma

varies among studies. In studies that include patients with
all types of soft tissue sarcoma,15,25 the median overall
survival as well as disease-specific survival after metastasectomy is poorer than in studies of more homogenous
populations of patients with gynecologic sarcomas.14,17,26
This finding may reflect more aggressive tumor biology in
patients with nongynecologic sarcomatous metastases.
In studies of patients with metastatic gynecologic sarcoma, the median survival after resection of metastatic
disease ranges from 24 to 31 months, similar to the survival
for the patient described in our case history.9,10 In one study,
disease-free survival after surgical resection of first recurrence was nearly 4 years.17 Five and 10-year survival ranges
from 38% to 47% and 34% to 35%, respectively.12,14,26
Studies addressing outcomes after surgical resection of recurrent uterine sarcoma are summarized in Table 1. Given
these favorable survival data, it is clear that in the appropriately selected patient, surgical resection of metastatic
disease may prolong disease-free and overall survival.
Characteristics Correlating with Improved Outcome
after Metastasectomy
Characteristics associated with improved survival after

resection of recurrent disease have been retrospectively
assessed in multiple studies. Factors that have been assessed for correlation with improved outcome after metastasectomy include complete response after upfront
treatment, disease-free interval between primary diagnosis
and surgery for first recurrence, presence of residual disease
SURGICAL OPTIONS FOR UTERINE SARCOMAS

Table 1. Outcomes of Surgery for Recurrent Uterine Sarcoma
Total No. of Pts.
(No. with Uterine
Sarcoma)
Study
Anderson et al (2001)
11
Anraku et al (2004)12
19 (12)
133 (11)
Bernstein-Molho et al (2010)9
33 (33)
Clavero et al (2006)26
70 (41)
Giuntoli et al (2007)10
128 (128)
Histology
Recurrence Site
LMS, ESS, nonsarcomatous uterine

malignancies
LMS, cervical cancer, uterine
choriocarcinoma, uterine
adenocarcinoma
LMS
LMS, ESS, adenocarcinoma,

other sarcoma, SCCA,
choriocarcinoma
LMS
Leitao et al (2002)17
41 (41)
LMS
Levenback et al (1992)14
45 (45)
LMS, ESS, carcinosarcoma (7%)
Lung
Outcome
Median survival, 25 mo for patients

with recurrent uterine LMS
5-yr survival, 37.9% for patients with
recurrent LMS
Lung
Lung, pelvis, abdomen, bone,

retroperitoneum, brain,
liver, adrenal glands
Lung
Pelvis, abdomen, thoracic cavity
Pelvis, lung, abdomen, bone,

multiple synchronous sites
Lung
Median PFS, 7.9 mo; OS, 45 mo
5-yr survival, 46.8%; 10-yr survival,

34.3% in all histologic subgroups
Median survival, 2.0 yr vs. 1.1 yr in
metastasectomy vs. chemotherapy
and/or RT group
Median DSS, 3.9 yr; 2-yr survival, 71%
5-yr survival 43%; 10-yr survival, 35%
Abbreviations: LMS, leiomyosarcoma; ESS, endometrial stromal sarcoma; PFS, progression-free survival; OS, overall survival; SCCA, squamous cell carcinoma; RT,
radiation therapy; DSS, disease-specific survival.
after metastasectomy, number of metastases, lesion size,

bilaterality of pulmonary lesions, and tumor doubling time.
In one study, data on patients who underwent thoracic
procedures for recurrent disease were compared with data
on patients undergoing extrathoracic surgery, and no survival difference was noted between the two groups.17
A disease-free interval longer than 6 to 12 months between initial diagnosis and recurrence may indicate less
aggressive tumor biology and correlates with improved survival after metastasectomy.4,10-13,16,17,26 One study demonstrated a 5-year survival of 60% compared with 37% with a
disease-free interval longer than 12 months. Another study
resulted in a marked difference of 5.1 compared with 1.5
years mean survival after surgery for recurrence in women
with a disease-free interval of more than 12 months.17
Another study showed a demonstrable survival improvement with each month increase in disease-free interval.16
Other factors that have been associated with survival
after surgery for recurrent disease include lesion size, history of prior metastasectomy, number of metastases, and
bilaterality of lesions.13,14 The utility of these prognostic
indicators may be as markers for resectability of the recurrent tumor.10 Presence of residual disease after resection for
recurrence has been shown by multiple investigators to be
associated with a considerably poorer prognosis and shorter
disease-free survival.10,15 One study demonstrated a
marked difference in median survival (3.9 vs. 0.7 years after
surgery) according to the completeness of resection for
recurrence.17
In a study of patients undergoing repeat pulmonary me-
tastasectomy for all soft tissue sarcoma, multivariate analysis showed that incomplete resection, resection of more
than two nodules, resection of nodules larger than 2 cm, and
high grade of primary tumor were meaningful factors associated with poorer survival. The 5-year survival rate was
100% for patients who had none of these poor prognostic
factors and was 10% for patients who had three factors.15
Of note, all but one of these four characteristics would be
known preoperatively, allowing for careful assessment of
the potential benefit of surgical resection in the case of
recurrent disease.
In two studies in which the use of adjuvant therapy after
surgery for recurrent uterine sarcoma was evaluated, postoperative chemotherapy or radiation therapy did not improve outcomes.14,17
Conclusion
Women with recurrent uterine sarcoma have an overall

poor prognosis. Surgical resection of metastases, in a wellselected patient, offers the possibility of cure and improved
disease-specific survival compared with other treatment
modalities. Prolonged disease-free interval and resectability
of tumor have been identified as favorable prognostic characteristics in multiple studies. Many characteristics correlating with optimal tumor resection, and therefore improved
outcome, are evaluable preoperatively. Thus, with appropriate preoperative assessment and risk stratification, patients
with recurrent uterine sarcoma who are medically stable
may be considered candidates for surgical resection for the
management of recurrent disease.
Author
John P. Curtin*
Sharmilee B. Korets*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
365
REFERENCES
1. Gadducci A, Cosio S, Romanini A, et al. The management of patients
with uterine sarcoma: a debated clinical challenge. Crit Rev Oncol Hematol.
2008;65:129-142.
2. Barakat RR, Markman M, Randall M. Principles and Practice of Gynecologic Oncology. 5th ed. Philadelphia: Wolters Kluwer Health/Lippincott
Williams & Wilkins; 2009.
3. Seddon BM, Davda R. Uterine sarcomas-recent progress and future
challenges. Eur J Radiol. 2011;78:30-40.
4. Amant F, Coosemans A, Debiec-Rychter M, et al. Clinical management
of uterine sarcomas. Lancet Oncol. 2009;10:1188-1198.
5. Omura GA, Major FJ, Blessing JA, et al. A randomized study of
adriamycin with and without dimethyl triazenoimidazole carboxamide in
advanced uterine sarcomas. Cancer. 1983;52:626-632.
6. Sutton G, Blessing JA, Malfetano JH. Ifosfamide and doxorubicin in the
treatment of advanced leiomyosarcomas of the uterus: a Gynecologic Oncology Group study. Gynecol Oncol. 1996;62:226-229.
7. Hensley ML, Maki R, Venkatraman E, et al. Gemcitabine and docetaxel
in patients with unresectable leiomyosarcoma: results of a phase II trial.
J Clin Oncol. 2002;20:2824-2831.
8. Nam JH. Surgical treatment of uterine sarcoma. Best Pract Res Clin
Obstet Gynaecol. 2011;25:751-760.
9. Bernstein-Molho R, Grisaro D, Soyfer V, et al. Metastatic uterine
leiomyosarcomas: a single-institution experience. Int J Gynecol Cancer.
2010;20:255-260.
10. Giuntoli RL 2nd, Garrett-Mayer E, Bristow RE, et al. Secondary
cytoreduction in the management of recurrent uterine leiomyosarcoma.
Gynecol Oncol. 2007;106:82-88.
11. Anderson TM, McMahon JJ, Nwogu CE, et al. Pulmonary resection in
metastatic uterine and cervical malignancies. Gynecol Oncol. 2001;83:472476.
12. Anraku M, Yokoi K, Nakagawa K, et al. Pulmonary metastases from
uterine malignancies: results of surgical resection in 133 patients. J Thorac
Cardiovasc Surg. 2004;127:1107-1112.
13. Burt BM, Ocejo S, Mery CM, et al. Repeated and aggressive pulmonary
resections for leiomyosarcoma metastases extends survival. Ann Thorac
Surg. 2011;92:1202-1207.
14. Levenback C, Rubin SC, McCormack PM, et al. Resection of pulmonary
metastases from uterine sarcomas. Gynecol Oncol. 1992;45:202-205.
366
15. Weiser MR, Downey RJ, Leung DH, et al. Repeat resection of pulmonary metastases in patients with soft-tissue sarcoma. J Am Coll Surg.
2000;191:184-190, discussion 190-181.
16. Blackmon SH, Shah N, Roth JA, et al. Resection of pulmonary and
extrapulmonary sarcomatous metastases is associated with long-term survival. Ann Thorac Surg. 2009;88:877-884, discussion 884-875.
17. Leitao MM, Brennan MF, Hensley M, et al. Surgical resection of
pulmonary and extrapulmonary recurrences of uterine leiomyosarcoma.
Gynecol Oncol. 2002;87:287-294.
18. Chen H, Pruitt A, Nicol TL, et al. Complete hepatic resection of
metastases from leiomyosarcoma prolongs survival. J Gastrointest Surg.
1998;2:151-155.
19. Pawlik TM, Vauthey JN, Abdalla EK, et al. Results of a single-center
experience with resection and ablation for sarcoma metastatic to the liver.
Arch Surg. 2006;141:537-543, discussion 543-534.
20. Montag TW, Manart FD. Endolymphatic stromal myosis: surgical and
hormonal therapy for extensive venous recurrence. Gynecol Oncol. 1989;33:
255-260.
21. Renzulli P, Weimann R, Barras JP, et al. Low-grade endometrial
stromal sarcoma with inferior vena cava tumor thrombus and intracardiac
extension: radical resection may improve recurrence free survival. Surg
Oncol. 2009;18:57-64.
22. Veroux P, Veroux M, Nicosia A, et al. Thrombectomy of the inferior
vena cava from recurrent low-grade endometrial stromal sarcoma: case report
and review of the literature. J Surg Oncol. 2000;74:45-48.
23. Yokoyama Y, Ono Y, Sakamoto T, et al. Asymptomatic intracardiac
metastasis from a low-grade endometrial stromal sarcoma with successful
surgical resection. Gynecol Oncol. 2004;92:999-1001.
24. Clement PB, Scully RE. Mullerian adenosarcoma of the uterus: a
clinicopathologic analysis of 100 cases with a review of the literature. Hum
Pathol. 1990;21:363-381.
25. Smith R, Pak Y, Kraybill W, et al. Factors associated with actual
long-term survival following soft tissue sarcoma pulmonary metastasectomy.
Eur J Surg Oncol. 2009;35:356-361.
26. Clavero JM, Deschamps C, Cassivi SD, et al. Gynecologic cancers:
factors affecting survival after pulmonary metastasectomy. Ann Thorac Surg.
2006;81:2004-2007.
PATIENTS WITH HPV-POSITIVE OROPHARYNX

TUMORS: COMMUNICATING DIAGNOSIS TO
FAMILY AND TREATMENT OPTIONS
CHAIR
David Brizel, MD
Durham, NC
SPEAKERS
William H. Westra, MD
Johns Hopkins School of Medicine
Baltimore, MD
Maura L. Gillison, MD, PhD
Columbus, OH
Management of Human PapillomavirusInduced

Oropharynx Cancer
By David Brizel, MD
Overview: Oropharynx cancer (OPC) constitutes the most

common location for squamous-cell head and neck cancer,
and most OPC is caused by the human papilloma virus (HPV).
Early-stage (American Joint Committee on Cancer [AJCC]
stage I and II) disease should be treated with single modality
surgery or radiotherapy whenever possible. More advanced
presentations generally require combined-modality therapy
with various combinations of surgery, radiotherapy, and che-
HE VAST majority ( 90%) of squamous-cell head and

neck cancers (HNCs) that are induced by the human
papillomavirus (HPV) originate within the oropharynx. This
anatomic region comprises the tonsils, soft palate, and base
of tongue. Understanding the general management principles of oropharynx cancer (OPC) is central to comprehending the development of new therapeutic options for HPVpositive OPC. The first step in this process is to recognize
that OPC can be clinically divided into early-stage and
advanced-stage presentations. Early-stage presentations
typically include T1N0 and T2N0 presentations (stage I and
II), whereas advanced-stage disease is more heterogeneous,
with presentations ranging from T1N1 to T4N3 (stage III to
IVB).
The fundamental management tenet for early-stage disease is to adopt a strategy that maximizes the likelihood
that the disease can be treated with single-modality therapy either surgery or radiotherapy (RT) alone. These two
modalities have equivalent efficacy in the early-stage setting. The rationale for the use of a single modality is to
minimize morbidity, which invariably increases in step with
the use of multiple therapeutic modalities. Surgical management consists of resection of the primary site, most often
with an ipsilateral neck dissection, whereas definitive RT
treats the same primary tumor site and ipsilateral lymph
nodes as surgery. The advent of minimally invasive surgical
techniques including transoral robotic resection and transoral laser excision and the increasing adoption of intensitymodulated RT (IMRT) have all led to substantial reductions
in the long-term functional morbidity of treating early-stage
OPC. The National Comprehensive Cancer Center Network
(NCCN) Head and Neck Cancer Guidelines provide additional detail regarding the choices for radiation fractionation schemes (www.nccn.org). Postoperative irradiation for
early-stage disease should be used only on the basis of
adverse histopathology including perineural invasion, positive margins, the presence of multiply involved lymph nodes,
or the presence of extracapsular nodal extension. Conversely, when RT is used as the primary modality, surgery
should be used only for the presence of persistent disease or
for the salvage of recurrent disease. Primary RT is usually
preferred for tumors originating in the base of tongue or
extending from the tonsil onto the soft palate close to the
midline because of the higher risk of occult contralateral
lymph node involvement, which necessitates treatment of
both sides of the neck. These regions can usually be treated
with less long-term functional morbidity by using RT as
opposed to bilateral neck dissection.
368
motherapy or molecularly targeted therapy. All of these approaches expose patients to a substantial risk of serious
long-term functional morbidity. HPV-induced OPC has a very
favorable prognosis compared with its HPV-negative counterpart irrespective of the treatment platform that is used.
Current clinical trials are investigating the concept of therapeutic deintensification with the dual objectives of decreasing
toxicity and maintaining efficacy.
T1N1 and T2N1 presentations are classified as stage III

and technically advanced disease, but the management
principles for stage I/II presentations are still generally
applicable, namely treatment with primary surgery or RT as
single modality with an understanding, however, that the
larger burden of disease increases the probability that the
addition of adjuvant neck dissection or postoperative irradiation will be necessary. Traditionally, primary nonsurgical
management has assumed the most prominent role for the
management of more extensive stage III and IV presentations, with surgery being used more in an adjuvant setting
(neck dissection or salvage role for residual/recurrent disease at the primary site). Three different strategies can
be considered as standards of care for management of
advanced-stage disease: RT alone using modified fractionation, RT plus concurrent epidermal growth factor receptor
(EGFR) blockade, and RT and concurrent chemotherapy.
Randomized trials have confirmed improvements in locoregional control with the use of both hyperfractionation
and accelerated fractionation by means of improvement in
locoregional disease control.1-3 A recent meta-analysis from
Bourhis and colleagues evaluated updated individual patients data from 6,515 patients in 15 randomized trials of
patients with stage III/IV oropharynx and larynx carcinoma.4 It showed an 8.2% absolute difference in 5-year
survival for patients who received altered fractionation RT
compared with conventional once-daily RT (36.7% vs. 28.5%,
respectively).
EGFR is overexpressed in the majority of patients with
head and neck squamous-cell carcinoma. Bonner and colleagues conducted a prospective randomized trial that
tested the value of adding EGFR blockade to a course of
RT.5,6 Four hundred twenty-four patients were randomly
assigned to receive either RT alone or RT plus weekly
cetuximab, a chimeric monoclonal antibody to the EGFR
receptor. The majority of these patients had oropharynx
primary tumors, and 75% of them received treatment with
accelerated or hyperfractionated irradiation. The RT/cetuximab patients had significant improvements in median survival (49 months vs. 29 months; hazard ratio [HR] 0.73;
p 0.02), and 5-year overall survival (46% vs. 36%). Subset
From the Duke University Cancer Institute, Durham, NC.

Address reprint requests to David Brizel, MD, Duke University Cancer Institute, Box 3085
DUMC, Durham, NC 27710; email: david.brizel@duke.edu.
1092-9118/10/1-10
HPV-INDUCED OROPHARYNX CANCER
analysis demonstrated the most pronounced therapeutic

benefit in patients with oropharynx primaries, N presentations, and treatment with accelerated fractionation.
Concurrent chemoradiation represents the largest experience in the nonsurgical management of advanced HNC.
Pignon and colleagues performed an extensive metaanalysis of individual patient data from 17,346 patients
enrolled on 93 randomized controlled trials comparing RT
alone against RT and chemotherapy.7,8 These trials were
published from 1965 to 2000, and it is important to recognize
that this era was before the use of taxane-based chemotherapy in HNC. Their analysis demonstrated a 21% reduction
in the risk of recurrence or death as a result of the addition of concurrent chemotherapy to RT. This risk reduction
corresponded to a 6.5% absolute improvement in 5-year
survival (33.7 vs. 27.2%). The majority of this improvement
was attributable to a 13% absolute improvement in locoregional disease control. A more modest 2.9% absolute
improvement in distant failure was attributable to concurrent therapy. The analysis also showed that the use of
concurrent cisplatin was most important. Induction chemotherapy did not provide an improvement in locoregional
disease control.
The role of induction chemotherapy continues to be investigated. Several phase III trials have confirmed that the
addition of a taxane to a platinum and flurorouracil (FU)
based induction regimen yields superior survival compared
with induction with platinum and FU only. Most of these
trials administered substandard locoregional therapy, however.9-11 Whether taxane-based induction followed by stateof-the-art concurrent chemoradiation is superior to concurrent chemoradiation alone, however, remains unknown.
Toxicity considerations must be incorporated into the
treatment decision-making process for HNC in general and
OPC in particular. The standard investigational treatment
paradigm has been one of therapeutic intensification via the
combination of multiple modalities with increasing disease
stage and tumor burden. Concurrent chemoradiation programs do improve overall efficacy, but at the price of a
considerable increase in toxicity. The incidence of grade 3
confluent mucositis approximately doubles with the addi-
KEY POINTS
The incidence of oropharynx cancer is increasing even

as the incidence of head and neck cancer (HNC) in
other locations is decreasing.
This increase is attributable to the human papillomavirus (HPV), which causes the majority of cases of
oropharynx cancer.
The prognosis of HPV-associated HNC is much more
favorable than that of HPV-negative HNC.
The favorable prognostic effect of HPV positivity is
independent of treatment platform, although a history of cigarette smoking appears to negate some of
the benefit.
Trials under development for HPV-positive HNC will
explore the concept of therapeutic de-escalation designed to reduce treatment-associated morbidity with
preservation of thereapeutic efficacy.
tion of concurrent chemotherapy to RT-alone regimens. The

percentage of patients experiencing serious late toxicities
including cervical fibrosis, dental disease, and neuropathy
increases by approximately 50%.12 The addition of neck
dissection after concurrent chemoradiation also substantially increases the risk of late toxicity after concurrent
chemoradiation.13 Recent developments in the use of postchemoradiation computed tomography and positron emission tomography scanning, however, have substantially
improved the ability to separate those patients who do
require adjuvant neck dissection from those who do not.14
Bentzen and Trotti analyzed the Radiation Therapy Oncology Group (RTOG) portfolio of randomized trials in HNC
that have either compared different RT schedules against
one another or compared RT and chemotherapy against RT
alone.15 They showed that both induction chemotherapy and
concurrent chemotherapy approaches increased the overall
burden of severe toxicity on the patient by five-fold compared with conventionally fractionated irradiation. Moreover, their data clearly demonstrated that a single patient
can have more than one severe toxicity event, something
that is not routinely captured in conventional toxicity scoring systems.
The toxicity of a given treatment strategy can also be
judged by the probability that a patient will be able to
complete the entire prescribed course of treatment. Sequential chemoradiation regimens (induction chemotherapy followed by concurrent chemoradiotherapy) sound a cautionary
note in this regard. Compliance rates with an entire course
of treatment in the published clinical trials range from
only 50% to 73%.10,16 Sequential therapy regimens that use
state-of-the-art concurrent chemoradiation schemes appear
to be particularly problematic from a compliance standpoint.9,17
A recent evaluation of tissue obtained through the Surveillance, Epidemiology and End Results (SEER) registries
of patients with OPC receiving treatment from 1984 to 2004
shows an increasing incidence of OPC even as the overall
incidence of HNC is declining.18 OPC now constitutes the
most prevalent subtype of HNC within the United States.
Moreover, the majority (approximately 70%) of these OPCs
are HPV positive. Current estimations are that by the year
2020 the annual number of HPV-positive OPCs will exceed
the annual number of HPV-positive uterine cervix carcinomas. Additional projections are that there will be more than
15,000 cases of HPV-positive OPC per year by 2030.
The SEER analysis included patients who received various types of treatment. The overall 5- and 10-year survivals
were approximately two times better for those patients with
HPV-positive disease. HPV-positive disease appears to convey this more favorable prognosis in a platform-independent
fashion. RTOG-0129 was a phase III trial that compared
standard fractionation RT against accelerated fractionation
RT, with both arms receiving concurrent cisplatin chemotherapy.19 The accelerated fractionation arm received two
cycles, and the standard fractionation arm received three
cycles. There was no difference in overall outcome. Therefore, the patients from the two arms were pooled, and
outcomes were reanalyzed as a function of HPV positivity as
determined by p16 immunohistochemistry. The probability
of long-term survival was more than three times greater for
HPV-positive patients than for their HPV-negative counterparts (HR 0.29; p 0.001). A history of cigarette smoking
369
DAVID BRIZEL
counteracted the favorable effect of HPV positivity, however.

HPV-positive patients who received sequential chemoradiation on the Tax-324 trial also had an approximate three-fold
improvement in overall survival at 5 years compared with
the HPV-negative patients.20 Patients who are treated
with RT alone also have a two- to threefold more favorable
prognosis than similarly staged patients with HPV-negative
disease.21 A single-arm trial of transoral laser excision of
OPC has also demonstrated a more favorable prognosis for
HPV-positive patients.22
The vastly different prognosis of OPC according to HPV
status is now guiding the development of new therapeutic
strategies. Long-term survival for HPV-negative patients
stubbornly remains in the 30% to 40% range, which is
clearly suboptimal. Therapeutic intensification with the
objective of improving efficacy is still appropriate for this
group of patients.
The one-size-fits-all intensification approach, with its associated acute and long-term morbidity, is no longer appropriate for HPV-positive patients because long-term survival
for these patients is in the 70% to 85% range. Most investigational efforts for this group are now aimed toward therapeutic de-escalation with dual objectives of reducing toxicity
and maintaining overall therapeutic efficacy. The favorable
effect of HPV positivity is equal for multiple treatment
platforms from a relative standpoint, but it is unknown
which, if any, of these platforms is superior to the others in
absolute terms.
Several rational strategies exist for treatment de-escalation. These include the use of RT alone instead of concurrent chemoradiation, the use of RT and EGFR inhibition
instead of concurrent chemoradiation, the use of induction
chemotherapy followed by reduced dose (chemo)RT for favorable responses, and primary surgery instead of RT or
chemoradiation.
The Eastern Cooperative Oncology Group recently completed enrollment onto a phase II trial (ECOG 1308; www.
clinicaltrials.gov; NCT01084083) in which 83 patients with
stage III-IVB HPV-positive OPC received three cycles of
induction chemotherapy consisting of paclitaxel, cisplatin,
and cetuximab. Those patients who had a clinical and
radiographic complete response (CR) then received reduceddose IMRT to 54 Gy via conventional fractionation and
weekly cetxuimab. Those who had less than CR received
accelerated fractionation IMRT (69.3 Gy/33 fractions). Results of this study are pending.
RTOG is currently conducting a phase III trial in patients
with HPV-positive (determined by p16 immunohistochemis-
try) locally advanced OPC (RTOG 1016; www.clinicaltrials.

gov; NCT01302834). Randomization is stratified by low
and high T stage, low and high N stage, and smoking history
( or 10 pack years). Both arms of this trial will use
accelerated fractionation IMRT (70 Gy in 6 weeks). The
control arm will receive two cycles of cisplatin, and the
experimental arm will receive weekly cetuximab. This trial
will enroll 700 patients overall.
The National Cancer Institute (NCI) Head and Neck
Cancer Steering Committee conducted a clinical trials planning meeting in November 2011 to formulate a concept that
would be developed into a clinical trial designed to assess
the role of transoral resection as a primary modality in the
management of HPV-positive OPC. This concept is in the
final stages of development, and the clinical trial will be
conducted through the NCI Cooperative Group mechanism.
Most likely, a risk-based approach will be tested. Specifically, patients with I-IIB tonsil cancer will undergo transoral resection with lymph node dissection. Those patients
who are at low risk for recurrence on the basis of surgical
pathology will be observed. Those patients who are at high
risk for locoregional recurrence (positive margins and/or
positive extracapsular extension [ECE]) will receive standard postoperative irradiation and concurrent cisplatin chemotherapy. Those patients with intermediate risk features
such as positive nodes without ECE, perineural invasion, or
angiolymphatic invasion, or high pathologic T stage will
undergo a phase II random assignment between lowercompared with higher-dose postoperative irradiation. Prospective and standardized assessments of quality of life and
patient outcomes will be an important part of the evaluation
of the transoral resection strategy in addition to the standard end points of locoregional disease control, progressionfree survival, and overall survival. Furthermore, this
strategy will allow for tissue banking to be performed for all
patients for future evaluation of new biologic targets in the
treatment of this disease.
A note of caution is in order. Therapeutic deintensification
for HPV-positive OPC carries the very real risk of a reduction in efficacy. Conversely, intensification potentially exposes HPV-negative patients to the risk of even more
toxicity without a commensurate improvement in efficacy.
Both de-escalation and escalation strategies should be undertaken only within the context of rigorously designed
clinical trials that will be able to detect adverse outcomes
should they exist. Outside of a clinical trial, both HPVpositive and HPV-negative disease should be managed according to best practices that are based on existing evidence.
Author
David Brizel
Employment or
Leadership
Positions
NCI (L)
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Bristol-Myers
Squibb; Siemens
Molecular
Oncology
REFERENCES
1. Horiot JC, Le Fur R, NGuyen T, et al. Hyperfractionation versus
conventional fractionation in oropharyngeal carcinoma: final analysis of a
randomized trial of the EORTC cooperative group of radiotherapy. Radiother
Oncol. 1992;25:231-241.
2. Fu KK, Pajak TF, Trotti A, et al. A Radiation Therapy Oncology Group
370
(RTOG) phase III randomized study to compare hyperfractionation and two

variants of accelerated fractionation to standard fractionation radiotherapy
for head and neck squamous cell carcinomas: first report of RTOG 9003. Int J
3. Overgaard J, Hansen HS, Specht L, et al. Five compared with six
HPV-INDUCED OROPHARYNX CANCER

fractions per week of conventional radiotherapy of squamous-cell carcinoma
of head and neck: DAHANCA 6 and 7 randomised controlled trial. Lancet.
2003;362:933-940.
4. Bourhis J, Overgaard J, Audry H, et al. Hyperfractionated or accelerated
radiotherapy in head and neck cancer: a meta-analysis. Lancet. 2006;368:843854.
5. Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for
squamous-cell carcinoma of the head and neck. N Engl J Med. 2006;354:567578.
6. Bonner JA, Harari PM, Giralt J, et al. Radiotherapy plus cetuximab for
locoregionally advanced head and neck cancer: 5-year survival data from a
phase 3 randomised trial, and relation between cetuximab-induced rash and
survival. Lancet Oncol. 2010;11:21-28.
7. Pignon JP, Bourhis J, Domenge C, Designe L. Chemotherapy added to
locoregional treatment for head and neck squamous-cell carcinoma: Three
meta-analyses of updated individual data: MACH-NC Collaborative Group
Meta-Analysis of Chemotherapy on Head and Neck Cancer. Lancet. 2000;355:
949-955.
8. Pignon JP, le Matre A, Maillard E, Bourhis J. Meta-analysis of chemotherapy in head and neck cancer (MACH-NC): an update on 93 randomised
trials and 17,346 patients. Radiother Oncol. 2009;92:4-14.
9. Hitt R, Lopez-Pousa A, Martnez-Trufero J, et al. Phase III study
comparing cisplatin plus fluorouracil to paclitaxel, cisplatin, and fluorouracil
induction chemotherapy followed by chemoradiotherapy in locally advanced
head and neck cancer. J Clin Oncol. 2005;23:8636-8645.
10. Posner MR, Hershock DM, Blajman CR, et al. Cisplatin and fluorouracil alone or with docetaxel in head and neck cancer. New Engl J Med..
2007;357:1705-1715.
11. Vermorken JB, Remenar E, van Herpen C, et al. Standard cisplatin/
infusional 5-fluorouracil (PF) vs docetaxel (T) plus PF (TPF) as neoadjuvant
chemotherapy for nonresectable locally advanced squamous cell carcinoma of
the head andneck (LA-SCCHN): a phase III trial of the EORTC Head and
Neck Cancer Group (EORTC #24971). J Clin Oncol. 2004;22:490s (suppl;
abstr 5508).
12. Denis F, Garaud P, Bardet E, et al. Late toxicity results of the GORTEC
94-01 randomized trial comparing radiotherapy with concomitant radiochemotherapy for advanced-stage oropharynx carcinoma: comparison of LENT/
SOMA, RTOG/EORTC, and NCI-CTC scoring systems. Int J Radiat Oncol

Biol Phys. 2003;55:93-98.
13. Machtay M, Moughan J, Trotti A, et al. Factors associated with severe
late toxicity after concurrent chemoradiation for locally advanced head and
neck cancer: an RTOG analysis. J Clin Oncol. 2008;26:3582-3589.
14. Porceddu SV, Pryor DI, Burmeister E, et al. Results of a prospective
study of positron emission tomography-directed management of residual
nodal abnormalities in node-positive head and neck cancer after definitive
radiotherapy with or without systemic therapy. Head Neck. 2011;33:16751682.
15. Bentzen SM, Trotti A. Evaluation of early and late toxicities in
chemoradiation trials. J Clin Oncol. 2007;25:4096-4103.
16. Vermorken JB, Remenar E, van Herpen C, et al. Cisplatin, fluorouracil,
and docetaxel in unresectable head and neck cancer. New Engl J Med.
2007;357:1695-1704.
17. Adelstein DJ, Moon J, Hanna E, et al. Docetaxel, cisplatin, and
fluorouracil induction chemotherapy followed by accelerated fractionation/
concomitant boost radiation and concurrent cisplatin in patients with advanced squamous cell head and neck cancer: A Southwest Oncology Group
phase II trial (S0216). Head Neck. 2010;32:221-228.
18. Chaturvedi AK, Engels EA, Pfeiffer RM, et al. Human papillomavirus
and rising oropharyngeal cancer incidence in the United States. J Clin Oncol.
2011;29:4294-4301.
19. Ang KK, Zhang RH, Wheeler DI. A phase III trial (RTOG 0129) of two
radiation-cisplatin regimens for head and neck carcinomas (HNC): impact of
radiation and cisplatin intensity on outcome. J Clin Oncol. 2010;28:15s(suppl;
abstr 5507).
20. Posner MR, Lorch JH, Goloubeva O, et al. Survival and human
papillomavirus in oropharynx cancer in TAX 324: a subset analysis from an
international phase III trial. Ann Oncol. 2011;22:1071-1077.
21. Lassen P, Eriksen JG, Hamilton-Dutoit S, et al. Effect of HPVassociated p16INK4A expression on response to radiotherapy and survival in
squamous cell carcinoma of the head and neck. J Clin Oncol 2009;27:19921998.
22. Rich JT, Milov S, Lewis JS Jr.et al. Transoral laser microsurgery (TLM)
/ adjuvant therapy for advanced stage oropharyngeal cancer: outcomes
and prognostic factors. Laryngoscope. 2009;119:1709-1719.
371
TRANSLATING BIOLOGIC DISCOVERIES INTO

CLINICAL TRIALS
CHAIR
Lisa F. Licitra, MD
Fondazione IRCCS Istituto Nazionale dei Tumori
Milan, Italy
SPEAKERS
Christine H. Chung, MD
Baltimore, MD
Lillian L. Siu, MD
Toronto, ON, Canada
Important Early Advances in Squamous Cell

Carcinoma of the Head and Neck
By Eric Bissada, MD, DMD, Irene Brana, MD, and Lillian L. Siu, MD
Overview: Therapeutic advances in squamous cell carcinomas of the head and neck (SCCHN) are attained by improvement in locoregional and/or distant disease control, as well as
by reduction in treatment-related morbidity especially longterm complications affecting normal organ functions and
quality of life. New technological innovations in surgical
management, such as transoral robotic surgery (TORS), and in
radiotherapy (RT), such as proton and carbon ion therapy,
bring promises of equal or superior efficacy outcomes coupled
with the potential to minimize normal tissue toxicity. Scientific
HE DELIVERY of precision medicine is relevant in the

management of both early and advanced stages of
SCCHN, not only to improve functional outcomes and disease control, but also to reduce the burden of acute and
long-term treatment-related morbidity. Technical precision
may be enhanced surgically with the emergence of TORS,
accompanied by the benefit of reduced manipulation of
surrounding normal structures. The use of energy particles
such as protons or carbon ions offers precise RT dose
distributions, thus limiting undesirable scatter to nearby
normal organs. In the systemic treatment of SCCHN, precision medicine is based on the principles that druggable
molecular drivers of sensitivity or resistance exist in cancers, and that matching of pharmaceutical agents to specific
oncogenic aberrations may improve therapeutic outcome.
The evaluation of these multidisciplinary advances in SCCH
through the conduct of properly designed clinical trials is
necessary to support their application in practice.
Novel Advances in Surgical Management of SCCHN

The Emergence of TORS and Its Potential Benefits in SCCHN
The use of robotic surgery is gaining popularity and is

being applied to several fields including urology, orthopaedic
surgery and cardiac surgery. The latest use of this technology is in otolaryngology in which some surgeons have found
it helpful for the ablation of difficult-to-access tumors of the
upper aerodigestive tract, particularly oropharyngeal cancers. TORS is a topic of great interest in head and neck
surgery, and some surgeons believe its popularity will only
increase in the coming years. In TORS, the surgeon sits at a
remote console and controls micromanipulators that in turn
move the arms of a robot placed at the patients bedside. A
highly magnified three-dimensional view of the surgical field
is procured with precise, scaled, and filtered motions to the
operating arms. The proponents of TORS state several
benefits over the open approach. The avoidance of a mandibulotomy and its associated morbidity is clearly the main
advantage this technology offers. Decreased manipulation
and dissection of healthy tissue, improved cosmetic outcome,
decreased need for tracheotomies, early return to oral intake, and shortened hospital stay are but a few other
potential suggested benefits.1,2
At present, there are no clear and convincing data to
suggest the superiority of either primary surgery or RT in
terms of disease control and survival benefit for early-stage
oropharyngeal cancer.3 Functional outcomes and predicted
insights in the systemic treatment of SCCHN, such as novel

approaches to overcome epidermal growth factor receptor
(EGFR) resistance, may enable more effective molecular
targeting in SCCHN beyond the current armamentarium of
available agents. An overarching theme of these early multidisciplinary advances is to enable the delivery of precisionbased therapeutic regimens from both the technical and
scientific perspectives. Rigorous clinical trial evaluations are
necessary to help define their roles in practice.
quality of life after treatment have therefore largely been

the determining factors in the choice of therapy offered to
such patients. RT with or without concurrent chemotherapy
is currently the most frequently advocated primary treatment modality for all stages of oropharyngeal cancer,
whereas surgery is usually reserved for local and/or regional
failures. This stems from the belief that the surgical morbidity can be more substantial than the anticipated morbidity from nonsurgical approaches. Speech and swallowing can
be substantially compromised by an open surgical approach
to the oropharynx. This coupled with the fact that patients
may ultimately require adjuvant RT or concurrent chemoradiotherapy (CRT) has led most head and neck oncologists
to agree that primary RT should serve as the definitive
treatment for these patients.
In contrast to advanced-stage oropharyngeal cancers,
speech and eating quality-of-life outcomes have not shown
any advantage to RT with or without chemotherapy over
surgery followed by adjuvant RT in early-stage T1/T2 disease.4 With the advent of TORS, surgeons are given a new
and potentially less morbid way of approaching the oropharynx directly, leading many to reconsider surgery as the
primary treatment in selected cases of oropharyngeal cancers. Leaving the surrounding structures untouched has led
many surgeons to believe that TORS is less morbid when
compared with conventional open surgical approaches, and
to nonsurgical treatment regimens. Proponents of primary
surgery argue that adjuvant treatment should be based on
operative and pathologic findings tailoring such therapy to
the patients needs. Surgically treated early stage T1/T2,
N0/N1 carcinomas of the oropharynx avoided the need for
adjuvant RT in up to 38% of cases without sacrificing disease
control.5-7 Primary surgery to such tumors enabled a reduction of the median RT dose to target volumes and prevented
the use of concurrent chemotherapy in many cases.5,7 TORS
may also decrease target volumes to undissected tissues that
From the Department of OtolaryngologyHead and Neck Surgery, Division of Medical

Oncology and Hematology, Princess Margaret Hospital, University Health Network, Toronto, Ontario, Canada.
Address reprint requests to Lillian L. Siu, MD, Division of Medical Oncology and
Hematology, Drug Development Program, Princess Margaret Hospital, 610 University
Avenue, Suite 5-718, Toronto, ON, M5G 2M9, Canada; email: lillian.siu@uhn.on.ca.
1092-9118/10/1-10
373
BISSADA, BRANA, AND SIU
would otherwise have been included with more traditional

open approaches because of potential tumor seeding.7
Treatment De-Intensification in Favorable-Risk Patient Groups
De-intensified adjuvant treatment has the potential benefit of lessening the reported long-term adverse effects of
primary CRT. The demographic shift of patients with human papillomavirus (HPV)-positive oropharyngeal cancers
may present a population that may benefit from the avoidance of the long-term sequelae inherent to RT and CRT.
These are important considerations, particularly with respect to ongoing efforts to decrease long-term adverse effects
in potentially more favorable groups such as in HPV-positive
patients who are diagnosed younger and who will live
longer. TORS is one of several potential de-intensification
strategies that needs to be studied. The ideal indications and
applications of its use have not yet been completely elucidated. Patient selection and adequate preoperative assessment are essential. This would ideally exclude those
patients who would clearly benefit from adjuvant CRT, such
as those in whom margin clearance would be difficult or
whose disease shows evidence of extracapsular nodal extension and invasion on preoperative imaging. Long-term results of randomized trials comparing primary surgery with
RT are needed to clarify oncologic outcomes and quality-oflife issues in early-stage orophayrngeal cancer. This would
help support the belief that dose de-escalation or avoidance
of CRT results in reduced late toxicity and improved quality
of life.
Novel Advances in Radiotherapy in SCCHN
The ability to optimize the therapeutic ratio by delivering

sufficient radiation dose to tumor tissues while sparing
adjacent critical organs is highly relevant in the treatment
of SCCHN, and it relies on the dose distribution of the
radiation energy. Particle therapy using protons or carbon
ion has gained increasing interest in the treatment of
SCCHN, largely as a result of their physical characteristics
to travel only finite distances through tissues and release
most of their energy immediately before they come to rest, in
the so-called Bragg Peak. The prospect of delivering less
radiation to nontargeted healthy tissues is appealing, as is
the feasibility to intensify or escalate doses to tumor areas.
In a recent systematic review and meta-analysis of RT in
various head and neck cancers comparing photons, protons
and carbon ions, which included 86 observational studies
and eight comparative in silico studies, 5-year survival was
markedly higher after carbon ion therapy compared with
conventional photon therapy for mucosal malignant melanomas. Additionally, 5-year local control after proton therapy
was markedly higher for paranasal and sinonasal cancer
compared with intensity-modulated photon therapy. No
other substantial differences were observed. However, the
overall quality and quantity of data are limited, precluding
definitive comparisons of these modalities.8 Randomized
clinical trials in SCCHN evaluating the comparative effectiveness of different RT techniques and particle therapy are
awaited.
Novel Advances in Systemic Management of SCCHN
KEY POINTS
374
Examples of early advances in surgical, radiation,

and medical oncology have emerged in the treatment
of squamous cell carcinomas of the head and neck
(SCCHN), including transoral robotic surgery, particle therapy, and molecular strategies to overcome
epidermal growth factor receptor (EGFR) resistance.
Transoral robotic surgery provides better access to
challenging anatomic sites while avoiding morbidity
associated with open surgery. Its role alone or in
combination with other treatments, especially to enable de-intensification of favorable risk oropharyngeal cancers, warrants clinical evaluation.
Particle therapy such as the use of protons or carbon
ions offers dose distribution that may be intensified
or escalated for tumor tissues while scatter to adjacent critical healthy structures is minimized; definitive evaluations of comparative effectiveness of
particle therapy compared with photon-based radiation therapy are awaited.
Multiple molecular strategies to overcome EGFR resistance are under clinical development, including
targeting other members of the erbB/HER family
besides EGFR, other nodes along the EGFRmitogenactivated protein kinase (MAPK) pathway, or oncogenic
pathways other than EGFR, such as the phosphoinositide 3-kinase (PI3K), insulin-like growth factor 1
receptor (IGF-1R) and MET pathways.
The systemic management of SCCHN continues to represent an unmet medical need because cetuximab, the monoclonal antibody against the epidermal growth factor receptor
(EGFR), remains the only molecularly targeted agents approved by the U.S. Food and Drug Administration (FDA) for
the treatment of SCCHN.
Targeting EGFR Resistance in SCCHN
Multiple mechanisms of resistance to EGFR inhibition

have been proposed in SCCHN, including: 1) presence of the
mutant type III variant of EGFR (EGFRvIII) as the result of
an in-frame deletion mutation of exons 2 to 7 spanning the
extracellular ligand-binding domain and 2) extensive crosstalk among the ErbB/HER family receptors and with other
signaling pathways. EGFRvIII prevents the binding of EGF
and other ligands to the receptor and leads to constitutive
activation of the EGFR tyrosine kinase domain and its
downstream effectors.9 With crosstalk, blockade of a single
receptor such as EGFR can lead to compensatory upregulation of escape mechanisms along the same EGFR receptor
axis or its downstream effectors, or alternatively, via rescue
from another pathway. Figure 1 depicts several strategies
that are currently being evaluated in clinical trials to
overcome EGFR resistance in cancer, including SCCHN.
Targeting Members of the ErbB/HER Family Other Than EGFR
Currently, agents that target multiple members of the

ErbB/HER family other than EGFR are actively being developed in the clinic, with the rationale to overcome de novo
and/or acquired resistance to EGFR inhibition.
Pan-HER inhibitors. Afatinib and dacomitinib are both
orally available, small-molecule, irreversible Pan-HER in-
EARLY ADVANCES IN SCCHN

Targeting receptor
tyrosine kinases e.g.
HER family
receptors
Targeting other
receptor
tyrosine kinases
IRS
PI3K
Fig 1. Targeting epidermal growth
factor resistance in squamous cell carcinoma of the head and neck.
Abbreviations: ERK, extracellular signalregulated kinase; mTOR, mammalian
target of rapamycin; PI3K, phosphoinositide 3-kinase; PTEN, phosphatase and
tensin homolog.
TSC2
AKT
PTEN
Shc
Sos
Grb2
Ras
TSC1
Targeting
multiple
oncogenic
pathways
(horizontal
blockade)
RHEB
mTOR
Raf
MEK
1/2
Targeting
multiple
nodes of the
same
oncogenic
pathway
(vertical
blockade)
ERK
1/2
S6K
4E-BP
Transcription
Translation
Growth, Proliferation, Metastasis, Survival

hibitors with specificities against EGFR, HER2, and HER4
tyrosine kinases. In addition to targeting ErbB/HER family
receptor crosstalk, both of these agents have reported activity against EGFRvIII in preclinical models. In a randomized,
open-label phase II study of afatinib compared with cetxuimab in 124 platinum-refractory patients with recurrent or
metastatic SCCHN, the objective response rate (ORR) favored afatinib compared with cetuximab (22% vs. 13%,
respectively), as did the median progression-free survival
(PFS; 16 weeks compared with 10 weeks, respectively). The
most common adverse events related to afatinib included
diarrhea and skin toxicity.10 Afatinib is currently being
evaluated in two phase III trials, including one in the
recurrent or metastatic setting against methotrexate after
progression with platinum-based chemotherapy (NCT01345682), and the other in the locally advanced setting
against placebo as maintenance therapy after completion of
definitive CRT; this latter trial (NCT01345669) is limited to
only those patients whose tumors are HPV negative. Dacomitinib has undergone single-arm, open-label phase II
evaluation as a first-line therapy in 69 patients with recurrent or metastatic SCCHN and reported an ORR of 11%,
median PFS of 2.8 months, and median overall survival (OS)
of 7.6 months. Similar to afatinib, diarrhea and skin toxicity
were the most frequent treatment-related adverse events.11
HER3-targeting agents. HER3 has emerged to become a
key target in the therapeutic strategy to overcome EGFR
resistance. Despite its catalytically inactive tyrosine kinase
domain, HER3 is capable of allosterically activating the
kinase domain of its heterodimerized partners, preferably
HER2.12 In addition, phosphorylation of the C-terminal tail
of HER3 leads to direct recruitment of the p85 subunit of
phosphoinositide 3-kinase (PI3K), rendering it a potent
activator of the PI3K-AKT survival pathway.13,14 Several
HER3 targeting agents are undergoing or completing phase
I development, including MEHD7945A (Genentech, Inc.,
South San Francisco, CA), a humanized, dual-specific, immunoglobulin (Ig) G1 antibody capable of binding to both
EGFR and HER3 (NCT01207323)15; MM-121/SAR256212

(NCT00734305; Merrimack Pharmaceuticals Inc., Cambridge, MA and Sanofi Oncology, Paris, France) and U31287/AMG 888 (NCT00730470; U3 Pharma, Munich,
Germany and Amgen, Inc., Thousand Oaks, CA), both fully
human monoclonal antibodies directed against HER3;
among others. In a recent report using a mouse monoclonal
antibody (RTJ.2, Santa Cruz Biotechnology, Santa Cruz,
CA; dilution 1:750) against HER3, immunohistochemical
evaluation of 387 primary SCCHN specimens showed positive membranous and cytoplasmic HER3 expression in 34
(8.8%) and 300 (77.5%) cases, respectively. Membranous
HER3 overexpression was significantly associated with
worse OS (p 0.027) and was an independent prognostic
factor in multivariate analysis.16 In addition, Wilson and
colleagues demonstrated that a substantial proportion of
primary SCCHN samples and cell lines have increased
mRNA expression of the ligand heregulin, which in turn can
lead to autocrine activation of HER3.17 As such, heregulindriven tumors with coexpression and activation of HER3
may represent a target SCCHN patient population for
HER3-directed therapies.
Targeting EGFRMitogen-Activated Protein Kinase and
Other Oncogenic Pathways Concurrently
In many advanced solid tumors, even among molecularly

profiled patients whose tumors harbor activating mutations,
the activity of selective pathway inhibitors administered as
monotherapy appears limited, or in cases of initial response,
the duration of response is typically short lived. Therapeutic
resistance may occur as a result of incomplete pathway
blockade with a single inhibitor, crosstalk and/or upregulation of compensatory escape signaling pathways. A logical
approach to evade such mechanisms of resistance is by using
targeted combinations with two agents either interrogating
the same oncogenic signaling pathway (vertical blockade) or
via concurrent modulation of two separate signaling pathways (horizontal blockade). Targeting other receptor ty-
375
BISSADA, BRANA, AND SIU
rosine kinases that can lead to activation of both the

mitogen-activated protein kinase (MAPK) and PI3K pathways, such as the insulin-like growth factor 1 receptor
(IGF-1R) and MET pathways, represents an attractive strategy to reverse EGFR resistance and is actively being investigated in the clinical setting.18,19 In addition, phase I trials
combining inhibitors of the MAPK pathway (such as RAF or
MEK) and PI3K pathway (such as PI3K, AKT, or mammalian target of rapamycin [mTOR]) are near completion and
would be of interest in SCCHN, especially among patients
whose tumors harbor molecular aberrations in these key
signaling cascades.
Conclusion
Emerging innovations in surgical, radiation, and medical

oncology aim to improve technical precision and molecular
targeting in disease management for patients with SCCHN.
Clinical trial evaluations are needed to translate these early
advances into practice. Ultimately, the goal of these
precision-based strategies is to optimize disease control with
increased technical and scientific targeting while sparing
healthy tissue toxicity and late effects, such that translation
to tangible clinical benefits in patients with SCCHN is
achieved.
Author
Eric Bissada*
Irene Brana*
Lillian L. Siu
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Amgen
EntreMed (I)
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Abraxis
BioScience;
Bristol-Myers
Squibb;
Genentech;
GlaxoSmithKline;
Merck;
Millennium;
Novartis; Pfizer;
Regeneron;
Roche
REFERENCES
1. Iseli TA, Kulbersh BD, Iseli CE, et al. Functional outcomes after
transoral robotic surgery for head and neck cancer. Otolaryngol Head Neck
Surg. 2009;141:166-171.
2. Genden EM, Desai S, Sung CK. Transoral robotic surgery for the
management of head and neck cancer: a preliminary experience. Head Neck.
2009;31:283-289.
3. Cosmidis A, Rame JP, Dassonville O, et al. T1-T2 NO oropharyngeal
cancers treated with surgery alone: a GETTEC study. Eur Arch Otorhinolaryngol. 2004;261:276-281.
4. Allal AS, Nicoucar K, Mach N, et al. Quality of life in patients with
oropharynx carcinomas: assessment after accelerated radiotherapy with or
without chemotherapy versus radical surgery and postoperative radiotherapy. Head Neck 2003;25:833-839.
5. Moncrieff M, Sandilla J, Clark J, et al. Outcomes of primary surgical
treatment of T1 and T2 carcinomas of the oropharynx. Laryngoscope. 2009;
119:307-311.
6. Weinstein GS, OMalley BW Jr, Snyder W, et al. Transoral robotic
surgery: radical tonsillectomy. Arch Otolaryngol Head Neck Surg. 2007;133:
1220-1226.
7. Weinstein GS, Quon H, OMalley BW Jr, et al. Selective neck dissection
and deintensified postoperative radiation and chemotherapy for oropharyngeal cancer: a subset analysis of the University of Pennsylvania transoral
robotic surgery trial. Laryngoscope 2010;120:1749-1755.
8. Ramaekers BL, Pijls-Johannesma M, Joore MA, et al. Systematic review
and meta-analysis of radiotherapy in various head and neck cancers: comparing photons, carbon-ions and protons. Cancer Treat Rev. 2011;37:185-201.
9. Sok JC, Coppelli FM, Thomas SM, et al. Mutant epidermal growth factor
receptor (EGFRvIII) contributes to head and neck cancer growth and resistance to EGFR targeting. Clin Cancer Res. 2006;12:5064-5073.
10. Seiwert TY, Fayette J, Del Campo JM, et al. Updated results of a
randomized, open-label phase II study exploring BIBW 2992 versus cetux-
376
imab in patients with platinum-refractory metastatic/recurrent head and

neck cancer (SCCHN). Ann Oncol 2010;21:viii316 (abstr 1010PD).
11. Siu LL, Hotte SJ, Laurie SA, et al. Phase II trial of the irreversible oral
pan-human EGF receptor (HER) inhibitor PF-00299804 (PF) as first-line
treatment in recurrent and/or metastatic (RM) squamous cell carcinoma of
the head and neck (SCCHN). J Clin Oncol. 2011;29:45S (suppl; abstr 5561).
12. Jura N, Endres NF, Engel K, et al. Mechanism for activation of the EGF
receptor catalytic domain by the juxtamembrane segment. Cell. 2009;137:
1293-1307.
13. Soltoff SP, Carraway KL III, Prigent SA, et al. ErbB3 is involved in
activation of phosphatidylinositol 3-kinase by epidermal growth factor. Mol
Cell Biol. 1994;14:3550-3558.
14. Prigent SA, Gullick WJ. Identification of c-erbB-3 binding sites for
phosphatidylinositol 3-kinase and SHC using an EGF receptor/c-erbB-3
chimera. EMBO J. 1994;13:2831-2841.
15. Schaefer G, Haber L, Crocker LM, et al. A two-in-one antibody against
HER3 and EGFR has superior inhibitory activity compared with monospecific
antibodies. Cancer Cell. 2011;20:472-486.
16. Takikita M, Xie R, Chung JY, et al. Membranous expression of Her3 is
associated with a decreased survival in head and neck squamous cell
carcinoma. J Transl Med. 2011;9:126.
17. Wilson TR, Lee DY, Berry L, et al. Neuregulin-1-mediated autocrine
signaling underlies sensitivity to HER2 kinase inhibitors in a subset of
human cancers. Cancer Cell. 2011;20:158-172.
18. Wheeler DL, Huang S, Kruser TJ, et al. Mechanisms of acquired
resistance to cetuximab: role of HER (ErbB) family members. Oncogene.
2008;27:3944-3956.
19. Jameson MJ, Beckler AD, Taniguchi LE, et al. Activation of the
insulin-like growth factor-1 receptor induces resistance to epidermal growth
factor receptor antagonism in head and neck squamous carcinoma cells. Mol
Cancer Ther. 2011;10:2124-2134.
Application of Genomic and Proteomic

Technologies in Biomarker Discovery
By Elana J. Fertig, PhD, Robbert Slebos, PhD, and Christine H. Chung, MD
Overview: Sequencing of the human genome was completed

in 2001. Building on the technology and experience of wholeexome sequencing, numerous cancer genomes have been
sequenced, including head and neck squamous cell carcinoma
(HNSCC) in 2011. Although DNA sequencing data reveals a
complex genome with numerous mutations, the biologic interaction and clinical significance of the overall genetic aberrations are largely unknown. Comprehensive analyses of the
tumors using genomics and proteomics beyond sequencing
HE SEQUENCING of the human genome was completed in 2001 and the whole exome sequencing of head
and neck squamous cell carcinoma (HNSCC) was completed
in 2011.1-4 Although these DNA sequencing data reveal a
complex genome with numerous mutations, the biologic
interaction and clinical significance of the overall genetic
aberrations are largely unknown. With rapid development of
genomic and proteomic technologies, comprehensive analyses of the tumors reflective of individual genetic context
are feasible, and these data are expected to increase novel
biomarkers that may have a significant potential in the
clinical management of HNSCC. We will summarize the
current genomic and proteomic technologies and approaches
to data analyses. Although the experimental platforms vary,
the analysis techniques for these experiments for cancer
remain platform independent; therefore, we will review
experimental platforms individually but describe analysis
techniques largely without regard for the technology. Furthermore, we will discuss the general biomarker-discovery
paradigm using the technology and present a few examples
of the biomarker discovery and development with clinical
implications and their limitations.
Genomic Technology
In recent years, genomic technologies have expanded from

measuring global gene expression to measuring diverse
genomic features, including exon-level gene expression,
DNA binding, single nucleotide polymorphisms (SNPs), and
DNA methylation patterns (Table 1). Regardless of the
biomaterial they quantify (e.g., RNA, DNA or DNA methylation), the genomic platforms are generally based on either
microarray- or sequencing-based technologies. The first set
of microarray technologies focused on the transcriptome,
typically measuring differential expression between case
and control samples for a multitude of genes simultaneously.5 Current microarrays consist of sequence-specific
probes designed to hybridize with complementary portions
of the fluorescently labeled sample RNA or single-stranded
DNA under analysis. The intensity of fluorescence for each
probe increases with the amount of sample bound to that
probe, providing a quantitative measure of DNA or RNA
abundance, as previously described.6
However, microarray-based technologies for genomic measurements are being rapidly replaced by next-generation
sequencing (NGS).7 The NGS technologies simultaneously
sequence short, single-strand DNA fragments bound to a
solid surface. The sequencers then obtain fluorescence sig-
data can potentially accelerate the rate and number of biomarker discoveries to improve biology-driven classification of
tumors for prognosis and patient selection for a specific
therapy. In this review, we will summarize the current genomic
and proteomic technologies, general biomarker-discovery
paradigms using the technology and published data in
HNSCCincluding potential clinical applications and limitations.
nals for each of the sequence fragments based on complimentary binding of fluorescently tagged bases to each base
in the fragment sequentially.8 The fluorescent signals are
converted to sequences for each fragment and then aligned
to a full-length reference genome to match a sequencing
signal to genomic location.9 Counting the number of sequencing reads that align to a given genomic location is
analogous to microarray intensities for a probe with a
specific sequence. Similar to microarrays, sequencing technologies have subsequently evolved from measuring DNA
sequence to also measuring gene expression, DNA methylation and binding (Table 1). Sequencing technologies can
further identify variation between samples by identifying
genomic locations whereas reads that do not perfectly match
the reference genome may indicate individual genetic variation or a sequencing error.
Proteomic Technology
Proteomics refers to high-throughput studies of proteins

for detection, identification, quantification, and characterization. Global analysis of proteins has long been elusive,
but recent advances in mass-spectrometry (MS) have
made comprehensive and unbiased analysis of proteins in
complex mixtures possible. These advances include improved separation techniques, such as high performance
liquid chromatography (HPLC), 2-dimensional difference gel
electrophoresis (2D-DIGE), new labeling techniques (multicolor fluorophores, stable isotopes), and new detection and
identification strategies based on MS. Strategies to detect
whole proteins include tissue microarrays (TMAs), protein
microarrays, 2D-DIGE- and MS-based approaches.6,10
Tissue microarrays (TMAs) allow the detection of a single
protein in hundreds of separate tissue samples by wellestablished immunohistochemistry (IHC) using antibodies
specific to the protein of interest. The advantage of TMAs is
the preservation of histologic information, but the quality of
detection is dependent on biochemical characteristics of the
antibody, such as specificity and sensitivity. The IHC has a
From the Department of Oncology, Johns Hopkins University School of Medicine,

Baltimore, MD; Department of Cancer Biology, Vanderbilt University School of Medicine,
Nashville, TN.
Address reprint requests to Christine H. Chung, MD, Department of Oncology, Johns
Hopkins University School of Medicine, 1650 Orleans Street, CRB-1 Room 344, Baltimore,
MD 21231; email: cchung11@jhmi.edu.
1092-9118/10/1-10
377
FERTIG, SLEBOS, AND CHUNG

Table 1. Summary of Genomics Measurement Technologies and
Normalization Techniques.
Genomic Measurement
Technology
DNA sequence
Sequencing
Gene expression microarray
Gene expression
RNA-seq
ChIP-chip
ChIP-seq
Methylation array
Bisulfite sequencing
Binding assays (MBDSeq, MeDiP)
SNP-chip
DNA binding
Methylation
SNP
Sequencing
Platforms
NGS
Affymetrix GeneChip
Agilent microarray
Illumina BeadChip
NGS
Roche NibleGen
NGS
Illumina BeadChip
NGS
NGS or microarray
Affymetrix SNP array
Illumina SNP array
NGS
Abbreviations: NGS, next generation sequencing (performed sequencers, including Illumina Solexa, Applied Biosystems SOLiD and Roche 454); ChIP,
chromatin immunoprecipitation; MBDSeq, methyl-CpG binding domain protein
sequencing; MeDiP, methylated DNA immunoprecipitation; SNP, single nucleotide polymorphism.
low dynamic range for quantification, and high-quality antibodies may not be available for all targets of interest.
Reverse-phase protein microarrays (RPPM) are similar to
TMAs in that the protein mixture of interest is spotted onto
a glass slide and used for hybridization to an antibody. The
spots used for RPPM can be smaller than for TMA, and,
thus, a larger number of specimens can be surveyed simultaneously. Again, the quality of the RPPM is dependent on
the availability and biochemical properties of the antibody.
Two-dimensional protein separation using 2D-DIGE is
achieved by isoelectric focusing and electrophoresis in a
polyacrylamide gel. Newly developed multicolor fluorophores allow separate labeling of different protein mixtures,
which are then combined and analyzed together by 2DDIGE. However, reproducibility and sensitivity of the 2D
gels can be problematic, and it may miss some of the
smallest and largest proteins in the analysis.
KEY POINTS
378
Although DNA sequencing data reveal a complex

genome with numerous mutations, the biologic and
clinical significance of genetic aberrations are largely
unknown.
Microarray-based technologies for gene expression
analysis are being rapidly replaced by next generationsequencing technology.
New advances in proteomic technologies have made
unbiased and quantitative analysis of thousands of
proteins possible, leading to novel insights into tumor
biology and clinical features.
Although they are powerful discovery tools, each
finding must be vigorously validated before clinical
application.
New genetic and proteomic analysis technology allows comprehensive analyses of pathways rather
than individual genes or proteins, and they are promising tools to identify biomarkers and potential therapeutic targets.
Mass spectrometry-based applications have revolutionized protein detection and quantification, and the advances
are now starting to be used for clinical applications. All MS
instruments work by first ionizing biomolecules into a gasphase, separating these ions by their mass/charge properties
and detecting them after separation in the instrument. One
of the first MS applications uses was Matrix-Assisted Laser
Desorption Ionization (MALDI), where low molecularweight proteins are ionized from a solid matrix and separated by mass and charge to yield a detection profile unique
for a given mixture of proteins. MALDI-MS has the advantage that it requires minimal sample processing and that
it can analyze a large number of samples in a short time.
However, it is limited to the smallest proteins, does not
provide protein identification, and is very prone to experimental variation.
Since MS instruments are most sensitive and accurate
with small molecules, detection strategies that include digestion of proteins to peptides have been highly successful in
recent years. Here, a complex mixture of proteins is first
digested into peptides by trypsin, which cleaves proteins at
lysine (K) and arginine (R) residues. These peptides are then
ionized using a solid matrix (MALDI) or, more commonly, by
electrospray, where a solution of peptides is forced through
a thin needle to form a continuous spray that creates peptide
ions for analysis. Additional biochemical separation can be
applied before MS analysis using isoelectric focusing- or
strong cation exchange columns for a multidimensional
peptide separation. After separation and ionization, peptides may be analyzed and detected directly by MS, creating
a fingerprint of the original proteins; however, this only
works for relatively simple protein mixtures. For true identification purposes, peptide ions collide with an inert gas to
form fragment ions that are analyzed in a second MS phase
(tandem-MS or MS/MS) to create fragment MS spectra that
are unique for each peptide. Modern instruments can collide
and analyze thousands of peptides per minute, and these
analyses can be performed on a large scale and typically
yield thousands of protein identifications with quantitative
information. Identification is achieved by creating a list of
spectra predicted from genome sequencing information to
include all possible human peptides that could potentially
match the obtained MS spectra. These theoretical spectra
are then matched to the observed MS spectra through
database-search algorithms (Sequest, X!Tandem, Myrimatch). Knowledge of the full sequences of all proteins is
then used to reassemble the identified peptides into a list of
proteins, similar to the process used to create DNA sequencing alignments from known genomic data. The number of
times spectra were observed can function as a measure of
quantity for each protein.
When only a limited number of proteins are of interest, a
more targeted approach is possible in which the MS instrument is restricted to a narrow range of peptide masses,
which can then be measured with high sensitivity and over
a large dynamic range (selected or multiple reaction monitoring [SRM or MRM]). These assays can be multiplexed to
create highly specific assays that are also capable of measuring specific protein isoforms or modified protein forms.
Proteins can be detected even in lysates generated from
formalin-fixed paraffin-embedded tissues, since trypsin digestion can free up sufficient amounts of peptides for SRM or
MRM detection.
GENOMIC AND PROTEOMIC TECHNOLOGIES AND BIOMARKERS
Fig 1. Analysis steps. Genomics and proteomics analyses from all measurement platforms should follow steps (a)(f) to obtain clinically
relevant genomic signatures. (a) Sample experimental design accounting for technical artifacts in batches (e.g., processing date, lab technician,
etc). (b) Sample raw output from Affymetrix microarray. (c) Comparing the distribution of genomic data in each sample before and after
normalization, at which point measurements for each sample should be on the same scale. (d) Dendrogram for clustering of all data colored by
batch to identify artifacts in the data. (e) Heatmap of genes that the analysis identified as distinguishing tumor and normal samples (green
represents low measurement values and red high). (f) State predicted based on gene signature from the analysis in (e) for a new test set of tumor
and normal samples.
Approaches to Genomic and Proteomic Data Analyses
All genomic and proteomic analyses generally follow the

stages outlined in Figure 1. We will discuss these stages
using genomic analyses since they are better developed for
genomic data, although the various approaches apply to
proteomic datasets as well. First, careful experimental design is required to ensure that the measurements collected
can quantify the hypothesized genomic state or differences
in tumors and have large enough number of samples to
provide sufficient statistical power. Moreover, technical elements of the experimental design, including processing date,
lab technician, and processing center, can introduce significant artifacts known as batch effects into the data.11 These
batch effects are present in all genomics data, including
sequencing technologies. Therefore, the experimental design
should ensure that samples for each of the biologic conditions are processed in each of the technical conditions (Fig.
1A). If such representation is impossible because of the size
or scope of the experiment, the experiments should adopt a
randomized block design. Once collected, the raw data (Fig.
1B) must be preprocessed to convert the raw output of the
platforms to identifiable genomic or proteomic measurements that are comparable across the experiments (Fig. 1C),
including statistical removal of or control for batch effects
(Fig. 1D). Only after that can computational algorithms be
applied to infer gene-expression signatures pertinent to
cancer (Fig. 1E).6 Finally, as with any statistical analyses,

genomic analyses require careful cross-validation to ensure
the relevance of the inferred signatures to future datasets
collected from different biologic samples in similar experimental conditions (Fig. 1F). Further, directed experimentation is required to prove the functional mechanisms
suggested by the inferred signatures.
In general, there are three main analysis approaches:
unsupervised, supervised, and class-prediction algorithms.
Unsupervised (or class-discovery) algorithms seek gene- or
protein-expression signatures that distinguish samples
without regard for the biologic conditions in the experimental design. Hierarchical clustering has become perhaps the
most widely adopted algorithm for unsupervised genomic
analysis.12 These algorithms identify distances between sets
of genes or samples. Visualization of these clustering across
samples in dendrograms (Fig. 2A) can compare biologic
samples measured under different technical conditions,
making them particularly useful for identifying batch effects
and success of algorithms removing those effects (Fig. 1D).
Once removed, the clustering will classify samples to ideally
distinguish patients with different clinical outcomes. Likewise, genes that clustered together can be hypothesized to
function similarly. Similar utility can be gained from another form of clustering, k-means clustering. This algorithm
divides genes or samples from the data into a predetermined
379
Fig 2. Visualization of unsupervised analysis

algorithms. (a) A dendrogram provides a visualization of hierarchical clustering algorithms. The
length of vertical lines on the y-axis indicates
distances between samples or groups of samples
separated by a horizontal line. (b) K-means clustering assigns samples to clusters shown in the
three ovals colored in red (triangle samples), blue
(plus samples), or purple (circle samples). The xand y-axes are two dimensions representing the
majority of the variability in the samples, typically
determined with PCA. (c) Matrix factorization algorithms divide the measurement data into two
matrices. Columns in the Amplitude matrix represent genes with similar genomic properties
across the samples in the corresponding rows of
the Pattern matrix.
number of sets separated by a specified distance measure

(Fig. 2B). However, k-means clustering requires knowledge
of the number of sets a priori. Furthermore, all clustering
algorithms assign genes only to a single set, making it
impossible to account for the gene reuse that occurs in all
biochemical systems and notably in the processes enabled in
cancer.
In addition, matrix factorization algorithms can simultaneously infer a fixed number of relationships between genes
that are active to varying degrees in subsets of the samples.
For example, principal component analysis (PCA) has been
widely applied to the data to infer genomic signatures that
explain the majority of the variance in the data.13 However,
because of the underlying statistical assumptions, each
signature inferred from PCA will mix genomic responses
from all active processes, unable to disentangle the genomic
response of each active biochemical process. Non-negative
matrix factorization (NMF)14 and Bayesian Decomposition
(BD)15 were developed simultaneously to learn the parts of
systems, equivalent to signatures for biochemical processes.
These algorithms yield continuous, non-negative relationships between genes and samples that are explicitly allowed
to contain overlap across patterns (Fig. 2C). They have
subsequently been adapted in open-source software: the
CRAN package NMF for NMF16 and the Bioconductor package, CoGAPS, as an alternative to BD.17 Both algorithms
are computationally intensive, and their accuracy depends
on the number of patterns, analogous to selecting the number of clusters in k-means clustering.
Supervised genomic algorithms seek relationships in the
genomic data that distinguish the biologic conditions in the
experimental design. In their simplest form, the set of class
380
comparison algorithms typically compare measurements of

each gene measured in the data. Frequently, these analyses
are applied to find genes with significant differential expression in cancer samples, as compared to normal samples. For
continuous genomic measurements, such as gene expression
and methylation, class-comparison algorithms typically use
a t test to quantify gene-by-gene differences between the
experimental groups. Because the sample sizes are often
small, the p values obtained from these test statistics are
most accurately inferred from statistical models such as
permutation tests (e.g., SAM18) or through empirical Bayes
moderated t-statistics (e.g., LIMMA19). Because the empirical Bayes approach can also infer genomic differences
between multiple classes (e.g., tumor stage) or continuous
variables (e.g., drug-resistance measures), it has widely
replaced permutation tests in analyses. Discrete measurements of sequence alterations require more complex methods, such as hidden Markov models used to detect copy
number variation.20 Coordinated changes in these statistics
among groups of genes can further implicate activity in
specific biologic processes, such as signaling activity.
Regardless of the class discovery algorithm, raw, gene-bygene p values will identify huge numbers of false discoveries.
By definition, a p value of 5% means that the statistical test
will fail to reject the null hypothesis 5% of the time by
chance alone. Therefore, if a genomics data set measured
1000 genes, 50 genes will falsely be inferred as statistically
different between the groups. Simply dividing the desired
p-value threshold by the number of statistical tests can
reduce the number of false positives (the per-comparison
error rate). However, this correction is often too stringent
for genomics and proteomics data, resulting in fewif any
GENOMIC AND PROTEOMIC TECHNOLOGIES AND BIOMARKERS
genes being detected. As a result, more complex but permissive algorithms, such as the Bonferroni correction,
Benjamini-Hotchberg adjustment, or q-value have been developed to correct for the multiple-testing problem.21
Whereas unsupervised or class-discovery algorithms usually aim to uncover the biologic mechanisms underlying
differences in samples, class-prediction algorithms seek
genomic signatures that can accurately classify future samples. Numerous algorithms have been adapted for class
discovery in genomics from the field of machine-learning,
including neural networks and support-vector machines
among others.22 All of these techniques provide genomic
signatures that correlate with the desired phenotypic outcome. As a result, the genomic signatures are often complex
and difficult to interpret clinically. Validation of class discovery algorithms rests in the performance of these signatures in new data sets, rather than validation of the inferred
biologic mechanisms. The parameters of these algorithms
will be tuned to maximize true-positives (sensitivity) and
minimize false-negatives (specificity). For cancer diagnostics, class-discovery algorithms are often biased toward false
positives over false-negatives, erring on the side of additional patient care. However, this bias may result in excessive treatments.
Ideally, one set of data (training set) will be used to
develop the genomic signature with class-discovery algorithms. The resulting genomic signature will undergo final
validation in this independent test set and cannot undergo
any further correction. Often, these data sets have too few
samples to feasibly divide the data into test and training
sets that are large enough for accurate statistics. In these
cases, cross-validation techniques can test the ability of the
experiment and class discovery algorithm to predict the
phenotypes. However, there is no substitute for a truly
independent data set for testing the inferred signature
before clinical application. Such validation techniques are
also required for clinical application of signatures identified
with unsupervised or class-comparison algorithms.
Examples of Biomarker Discovery Using Genomic and
Proteomic Technologies
Recent completion of whole-exome sequencing of HNSCC

with NGS technologies revealed that it has a complex
genome with numerous genetic aberrations, such as mutations and gene amplification.3,4 The frequent mutations in
HNSCC are mostly tumor suppressors rather than oncogenes, which are difficult to develop as biomarkers and also
impose therapeutic challenges. Comprehensive analyses of
the tumors using genomic and proteomic technologies beyond the sequencing data can accelerate the rate and number of biomarker discoveries to improve biology-driven
classification of tumors for prognosis and patient selection
for a specific therapy. However, clinical application of these
technologies is still in the early stage with HNSCC. A few
examples are presented here.
Molecular Classification for Prognosis Using Genomics
Although HNSCC is classified as a histopathological

group of squamous cell carcinoma, it is a highly heterogeneous disease, including several anatomic subsites such as
oral cavity, oropharynx, hypopharynx, and larynx. Clinicians routinely observe varying outcomes within patients
with presumably comparable clinical prognostic factors,

such performance status and disease staging. Therefore,
identification of patients at a high-risk for recurrence and
poor survival will potentially improve the selective intensification of treatments. In a study by Chung and colleagues,
60 tumors obtained from patients with HNSCC were characterized using microarrays and found that there were at
least four groups within HNSCC segregated by distinct
expression profiles with different survival outcomes.23 By
examining the expression signatures, molecular characteristics of these groups could be inferred. For example, the
group with a high-risk of poor survival had a signature
reflective of EGFR activation whereas the group with an
intermediate risk had higher expression of genes involved
in epithelial-to-mesenchymal transition (EMT). The clinical
significance of the EMT process was further validated by the
association with a high risk of recurrence in HNSCC.24,25
Prediction of Local Recurrence Based on Proteomic
Profile in HNSCC
Even with the advancement of the intense multimodality

therapies for the management of patients with locally advanced HNSCC, the majority of the treatment failure is with
locoregional recurrence. Identification of the patients at risk
of local recurrence at the time of diagnosis may allow further
intensification of the therapy. Schaaij-Visser and colleagues
examined this clinical problem by applying differential proteomics.26 They compared paired normal, precursor and
tumor samples from eight patients using 2D-DIGE, and
identified differentially expressed proteins. After identifying
four proteins as potential biomarkers (keratin 4, keratin 13,
cornulin, and small proline-rich protein 3), they confirmed
that a lack of keratin 4 and cornulin expression as the
most discriminate proteins to predict local recurrence in
an independent data set. Equivalent studies using mass
spectrometry-based approaches are still limited, although
several abundant proteins have been studied as potential
biomarkers for HNSCC.27 Further research using modern
proteomic analysis tools has great potential to generate
novel biomarkers to predict clinical outcome and can potentially affect future patient care.
Future Directions
In conclusion, genomic and proteomic technologies have a

great potential to contribute to clinical research by generating an unbiased snapshot of the activity and composition of
cancer cells. Integration of these measurements is essential
to infer the underlying and complex mechanisms in cancer
development and maintenance. For example, simultaneous
analyses of gene and protein expression using a same
sample set can provide post-transcriptional regulation that
could not be detected with either genomic or proteomic
measurements alone.28 Once genes and proteins of interest
are identified, targeted assays can be developed for rapid
clinical translation. However, erroneous genomic and proteomic signatures that early on were claimed to be predictive
of patient response to chemotherapy or early cancer diagnosis cast a shadow on genomic and proteomic analyses.
Because of the complexity of these types of data, minor and
unintentional processing errors can render many findings
nonreproducible.29 Despite the steep learning curve, members of the genomics community are adopting open source-
381
based software and tools, such as R for all analyses.

Releasing scripts that formally document all stages of the
genomic and proteomic analyses with raw data when submitting for publication can provide an additional quality
assurance of the analyses at the manuscript-review stage.30
Although disclosure of the analysis scripts and subsequent
review cannot eliminate all mistakes from complex data
analyses, they allow others to reproduce the results and can
further facilitate validation of any genomic signatures before clinical application.

Acknowledgment
The authors appreciate the assistance of Michael F. Ochs,
Sarah Wheelan, and Luigi Marchioni in writing this manuscript.
Author
Elana J. Fertig*
Robbert Slebos*
Christine H. Chung
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Boehringer
Ingelheim;
Bristol-Myers
Squibb
Research
Funding
Expert
Testimony
Other
Remuneration
AstraZeneca;
Bayer; Lilly
REFERENCES
1. McPherson JD, Marra M, Hillier L, et al. A physical map of the human
genome. Nature. 2001;409:934-941.
2. Venter JC, Adams MD, Myers EW, et al. The sequence of the human
genome. Science. 2001;291:1304-1351.
3. Stransky N, Egloff AM, Tward AD, et al. The mutational landscape of
head and neck squamous cell carcinoma. Science. 2011;333:1157-1160.
4. Agrawal N, Frederick MJ, Pickering CR, et al. Exome sequencing of head
and neck squamous cell carcinoma reveals inactivating mutations in
NOTCH1. Science. 2011;333:1154-1157.
5. Brown PO, Botstein D. Exploring the new world of the genome with DNA
microarrays. Nat Genet. 1999;21:33-37.
6. Chung CH, Levy S, Chaurand P, Carbone DP. Genomics and proteomics:
emerging technologies in clinical cancer research. Crit Rev Oncol Hematol.
2007;61:1-25.
7. Ledford H. The death of microarrays? Nature. 2008;455:847.
8. Metzker ML. Sequencing technologiesthe next generation. Nat Rev
Genet. 2010;11:31-46.
9. Trapnell C, Salzberg SL. How to map billions of short reads onto
genomes. Nat Biotechnol. 2009;27:455-457.
10. Yates JR, Ruse CI, Nakorchevsky A. Proteomics by mass spectrometry:
approaches, advances, and applications. Annu Rev Biomed Eng. 2009;11:4979.
11. Leek JT, Scharpf RB, Bravo HC, et al. Tackling the widespread and
critical impact of batch effects in high-throughput data. Nat Rev Genet.
2011;11:733-739.
12. Eisen MB, Spellman PT, Brown PO, et al. Cluster analysis and display
of genome-wide expression patterns. Proc Natl Acad Sci U S A. 1998;95:
14863-14868.
13. Ringner M. What is principal component analysis? Nat Biotechnol.
2008;26:303-304.
14. Lee DD, Seung HS. Learning the parts of objects by non-negative
matrix factorization. Nature. 1999;401:788-791.
15. Ochs MF, Stoyanova RS, Arias-Mendoza F, et al. A new method for
spectral decomposition using a bilinear Bayesian approach. J Magn Reson.
1999;137:161-176.
16. Gaujoux R, Seoighe C. A flexible R package for nonnegative matrix
factorization. BMC Bioinformatics. 2010;11:367.
382
17. Fertig EJ, Ding J, Favorov AV, et al. CoGAPS: an R/C package to
identify patterns and biological process activity in transcriptomic data.
Bioinformatics. 2010;26:2792-2793.
18. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays
applied to the ionizing radiation response. Proc Natl Acad Sci U S A.
2001;98:5116-5121.
19. Smyth GK. Bioinformatics and Computational Biology Solutions Using
R and Bioconductor (ed XIX), Springer, 2005:397-420.
20. Winchester L, Yau C, Ragoussis J. Comparing CNV detection methods
for SNP arrays. Brief Funct Genomic Proteomic. 2009;8:353-366.
21. Noble WS. How does multiple testing correction work? Nat Biotechnol.
2009;27:1135-1137.
22. Ressom HW, Varghese RS, Zhang Z, et al. Classification algorithms for
phenotype prediction in genomics and proteomics. Front Biosci. 2008;13:691708.
23. Chung CH, Parker JS, Karaca G, et al. Molecular classification of head
and neck squamous cell carcinomas using patterns of gene expression. Cancer
Cell. 2004;5:489-500.
24. Chung CH, Parker JS, Ely K, et al. Gene expression profiles identify
epithelial-to-mesenchymal transition and activation of nuclear factor-kappaB
signaling as characteristics of a high-risk head and neck squamous cell
carcinoma. Cancer Res. 2006;66:8210-8218.
25. Pramana J, Van den Brekel MW, van Velthuysen ML, et al. Gene
expression profiling to predict outcome after chemoradiation in head and neck
cancer. Int J Radiat Oncol Biol Phys. 2007;69:1544-1552.
26. Schaaij-Visser TB, Graveland AP, Gauci S, et al. Differential proteomics identifies protein biomarkers that predict local relapse of head and neck
squamous cell carcinomas. Clin Cancer Res. 2009;15:7666-7675.
27. Matta A, Ralhan R, DeSouza LV, et al. Mass spectrometry-based
clinical proteomics: head-and-neck cancer biomarkers and drug-targets discovery. Mass Spectrom Rev. 2010;29:945-961.
28. Halvey PJ, Zhang B, Coffey RJ, et al. Proteomic consequences of a
single gene mutation in a colorectal cancer model. J Proteome Res. 2012;11:
1184-1195.
29. Ioannidis JP, Allison DB, Ball CA, et al. Repeatability of published
microarray gene expression analyses. Nat Genet. 2009;41:149-155.
30. Baggerly K. Disclose all data in publications. Nature. 2010; 467:401.
TREATMENT OF THYROID CANCERS: NEW

INFORMATION FOR MEDICAL ONCOLOGISTS
CHAIR
A. Dimitrios Colevas, MD
Stanford University School of Medicine
Stanford, CA
SPEAKERS
Michael Xing, MD, PhD
Baltimore, MD
Manisha H. Shah, MD
The Ohio State University Comprehensive Cancer Center
Columbus, OH
Keith C. Bible, MD, PhD
Mayo Clinic
Rochester, MN
Evaluation of Patients with Disseminated or

Locoregionally Advanced Thyroid Cancer:
A Primer for Medical Oncologists
By A. Dimitrios Colevas, MD, and Manisha H. Shah, MD
Overview: Historically, patients with thyroid cancers are

managed by endocrinologists, surgeons and radiation oncologists. Due to recent progress in this field with advances in
treatment of thyroid cancer, medical oncologists are now
commonly involved in care of patients with advanced thyroid
RADITIONALLY, TREATMENT of patients with differentiated thyroid carcinomas (DTC), which includes
papillary, follicular, and Hurthle cell carcinomas, has been
highly constrained by the lack of effective treatment of
patients with metastatic or recurrent locoregional disease
after surgical treatments have been exhausted or are not
feasible. Treatments in this setting have been mostly limited
to external-beam radiation treatment focused on the area of
concern or systemic radioactive iodine (RAI) for patients
with rising serum thyroglobulin levels or evidence of iodineavid evaluable metastases. Because there are no convincing
data on extending survival for patients with disease that is
evaluable only by thyroglobulin levels, endocrinologists and
nuclear medicine physicians have struggled with the selection of patients for whom RAI is indicated.1 Although some
advocate treatment with RAI for most patients with thyroglobulin levels higher than 10 ng/mL, others support a more
focused approach of treatment for patients who are thought
to be at high risk for clinically relevant sequellae of progressive disease, as disease volume can be reduced even in
patients with thyroglobulin-positive but imaging-negative
disease.2,3 Therefore, patients referred to the medical oncologist for RAI-refractory thyroid cancer have commonly received RAI in the adjuvant setting and often with an
additional one or two doses of RAI in the metastatic setting,
defined by elevated thyroglobulin levels in the absence of
radiographically evaluable metastatic disease. Historically,
it was not uncommon to encounter patients with relatively
indolent disease who had received cumulative RAI doses
exceeding 600 mCi. More recently, because of an increased
awareness of long-term adverse events from RAI, such as
bone marrow suppression, salivary gland compromise, and
gonadal dysfunction, many patients treated with RAI, even
those with clinically indolent disease, are referred to the
medical oncologist with lower prior cumulative RAI doses, in
the range of 200 400 mCi.
Traditional cytotoxic chemotherapy has played a minimal
role in the treatment of patients with RAI-refractory thyroid
cancer. Despite reports or small retrospective experiential
reviews suggesting promise, especially in the induction
setting for patients with no prior cancer treatment, most
well-documented studies demonstrate little to no response to
conventional chemotherapy in patients with RAI-refractory
disease.4,5 Doxorubicin is often cited as a standard of care on
the basis of a 37% response rate, but that rate was reported
in an oldstudy, and neither the time to progression nor the
methods used to determine responses were reported.6 Subsequent studies in patients with RAI-refractory disease have
yielded a 5% response rate.7 Combination chemotherapy has
384
cancers. In this manuscript, we describe general principles in

management of patients with various types of thyroid cancers
including differentiated, medullary and anaplastic thyroid cancers.
been equally disappointing. In a small study in which the

combination of cisplatin and doxorubicin was compared with
doxorubicin alone, the combination had a numerically lower
response rate than single-agent doxorubicin.8 There are
hints that newer combinations of cytotoxic chemotherapy
may be achieving better results. In a study of 14 patients
treated with gemcitabine plus oxaliplatin, a complete response was achieved in one patient and a partial response
was achieved in seven.9
The medical oncologist caring for patients with advanced
thyroid cancer should be familiar with alternatives to systemic cytotoxic chemotherapy for patients who have received
RAI (Sidebar 1). Patients with locoregionally recurrent disease following thyroidectomy and RAI should be considered
for surgery or external-beam radiation treatment. Occasionally, patients benefit from other locoregional modalities,
such as the use of ethanol injection or radiofrequency ablation for lymph nodes involved with clinically apparent recurrent disease.
In patients with locoregionally recurrent or metastatic
disease, the medical oncologist should also ensure that
aggressive suppression of thyroid-stimulating hormone has
been achieved. Some data suggest that aggressive thyroid
stimulating hormone (TSH) suppression leads to improved
outcomes, although one large randomized controlled study of
TSH suppression compared with replacement in the adjuvant setting after surgery did not demonstrate superiority of
suppression.10,11 There is consensus among experts that, for
patients with known recurrent differentiated thyroid cancer
(as opposed to patients being treated in the adjuvant setting), the target TSH level should be less than 0.1 mU/L.
For patients who have clinically meaningful progression
of disease during submaximal TSH suppression (TSH 0.1
mU/L), the trade-off of side effects from more aggressive
TSH suppression must be weighed against the possibility of
antitumor benefit. The levothyroxine dose necessary to
achieve suppression is typically higher than 1.6 mcg/kg/day,
often cited in the literature as a minimal thyroid suppressive dose for patients with benign thyroid conditions. One
study found that 2.56 mcg/kg/d was the average dose neces-
From the Stanford Cancer Institute, Stanford University, Palo Alto, CA; The Ohio State
University Comprehensive Cancer Center, Columbus, OH.
Address reprint requests to Manisha H. Shah, MD, Ohio State University, 320 W. 10th
Avenue, A438 Starling-Loving Hall, Columbus, OH 43210; email: manisha.shah@osumc.edu.
1092-9118/10/1-10
EVALUATION OF ADVANCED THYROID CANCER
Sidebar 1. Treatment Considerations for Recurrent or Metastatic Differentiated Thyroid Carcinoma
Consideration
Surgery
Specific Issues
Selected patients may benefit from repeated neck dissections or metastatectomies
of limited disease.
External-beam radiation
Often useful to control inoperable relapses in the neck and treatment of foci of
metastatic disease, especially central nervous system or bone lesions
Thyrotropin suppression
Target serum thyroid stimulating hormone level 0.1 mU/L. May require 2.5
mcg/kg/d or more of levothyroxine
Confirmation of optimal
preparation for prior
radioactive iodine
treatments
Was a low iodine diet maintained?

Was the TSH adequately elevated via thyroid hormone withdrawal or recombinant TSH administration?
Did the patient receive iodinated contrast medium in the 3 mos. preceding
treatment with radioactive iodine?
Was a low-iodine diet maintained prior to treatment with radioactive iodine?
Bone lesions
Bisphosphonates or RANK-ligand inhibitors reduce the risk of fracture.
sary to adequately suppress TSH in patients with thyroid

cancer.12
It is essential that the medical oncologist seeing patients
with recurrent or metastatic DTC be familiar with issues
usually addressed by endocrinologists or nuclear medicine
physicians. For example, before considering treatment with
either cytotoxic chemotherapy or molecularly targeted systemic agents, the medical oncologist should determine
whether RAI is an option for the patient. There is no
universally accepted definition of RAI-refractory DTC (Sidebar 2). Some clinicians accept a rise in thyroglobulin level
following administration of RAI as evidence of refractory
KEY POINTS
Spectrum of thyroid follicular-cell derived cancers

ranges from indolent differentiated thyroid cancers
(DTC) to very aggressive undifferentiated thyroid
cancer commonly known as anaplastic thyroid cancer
(ATC).
Medullary thyroid cancer (MTC) originates from
parafollicular c-cells of thyroid and is one of the best
characterized solid tumors.
Surgery plays critical role in management of all types
of thyroid cancers while targeted systemic therapies
with levo-thyroxine for thyroid stimulating hormonesuppression and radio-iodine are restricted to management of differentiated thyroid cancer.
Tyrosine kinase inhibitors are promising novel therapies for patients with DTC and MTC. Vandetanib
was approved by the U.S. Food and Drug Administration in 2011 for treatment of progressive or symptomatic advanced MTC.
Despite multimodality treatment with surgery, radiation and cytotoxic chemotherapy, prognosis remains
dismal with median survival of 6 months in patients
with advanced ATC.
disease. Others cite lack of detectable radioactivity on a

whole-body scan following diagnostic or therapeutic administration of RAI as definition of RAI-refractory disease.
Although progression after RAI treatment as documented by
conventional imaging with computerized tomography (CT),
magnetic resonance imaging, or bone scans is probably the
most conservative method of declaring a patients tumor to
be RAI refractory, conventional imaging comparisons are
often not available for patients who have been followed up
exclusively by serum thyroglobulin levels and whole-body
scans. Because of the inverse relationship between
fluorodeoxyglucose-positron emission tomography (FDGPET) avidity and RAI uptake, some clinicians consider a
positive FDG-PET scan itself as a definition of RAIrefractory disease.13,14
Under certain circumstances, it is important to review
how therapeutic RAI was administered to ensure that it was
given in a setting in which efficacy was maximized. In order
to facilitate maximal RAI uptake by the disease, there must
not be a sink for the RAI. In general, normal thyroid tissue
is more iodine avid than DTC; because of this, the presence
Sidebar 2. Various Definitions of Radioactive Iodine

(RAI) Refractory Differentiated Thyroid Carcinoma
1. Lack of uptake seen on a whole-body scan following diagnostic RAI dosing

2. Lack of uptake seen on a whole-body scan following
therapeutic RAI dosing.
3. Rising thyroglobulin following therapeutic RAI dosing
4. Progression of lesions as documented by conventional imaging (e.g., computerized tomography,
magnetic resonance imaging or bone scan) following
therapeutic RAI dosing.
5. Cumulative dose of 600 mCi of 131-iodine (131I)
6. Fluorodeoxyglucose (FDG)avid lesions on FDGpositron emission tomography.
385
COLEVAS AND SHAH

Table 1. Prognostic Considerations for Patients with Advanced
Differentiated Thyroid Carcinoma
Variable
BRAF mutation in tumor

RET/PTC oncogenic
rearrangement in
tumor
FDG-PET scan
Serum thyroglobulin
VEGF
Histology
Sites of metastasis
Age
Issue
Mutation found in 2969% of papillary thyroid cancer;

associated with a poor prognosis.18
Abnormality present in 530% of papillary thyroid
cancer and 6070% of radiation-induced papillary
thyroid cancer.19
The median survival for atients with FDG-avid disease is
approximately 4 years.20
Thyroglobulin level 10 ng/mL 612 mos. after ablation
is associated with an odds ratio for relapse of 16 vs.
those lower levels.
The doubling time of thyroglobulin strongly predicts
survival.21
High immunostaining for VEGF in tumor is associated
with risk of metastases.22
Follicular variant of papillary thyroid carcinoma is
associated with shorter survival in patients with
metastatic disease.23
Sites other than lung portend a worse prognosis.24,25
Patients 40 yr old have a better prognosis.25,26
Abbreviations: FDG-PET, fluorodeoxyglucose-positron emission tomography;

VEGF, vascular endothelial growth factor.
of residual normal thyroid tissue can diminish the effective

dose taken up by the cancer. Therefore, when residual
normal thyroid tissues cannot be removed surgically, a
two-stage approach to RAI dosing is often used. An initial
smaller dose of RAI is administered with the intent of
ablating normal thyroid tissue, followed by the anticancer
therapeutic dose after the normal thyroid tissue iodine sink
has been eliminated. A phenomenon known as thyroid
stunning, a situation in which residual DTC may take up
less iodine after ablative doses of RAI, has been described,
but the impact of stunning and whether it is of meaningful
therapeutic relevance has been contested by nuclear medicine experts.15 Therefore, when patients have metastatic
DTC that is not thought to be amenable to surgical resection, a surgical debulking of normal and cancerous thyroid
tissue in the neck followed by low or intermediate doses of
RAI in preparation for definitive RAI treatment may be an
approach preferable to systemic treatment with cytotoxic or
molecularly targeted agents.
Because uptake of iodine by both normal thyroid tissue
and DTC is related to serum TSH levels, it is important to
achieve levels of less than 30 mU/L before RAI dosing.
Historically, this level was achieved through withdrawal of
thyroid hormone replacement in patients without an intact
thyroid gland. With the recent availability of recombinant
thyrotropin (rTSH), many clinicians have moved from a
levothyroxine withdrawal strategy to administration of
rTSH administration. Although some data suggest that
there may be a reduction in sensitivity of radioiodine scans
using rTSH, a comparison of the therapeutic efficacy of these
two approaches in the metastatic setting failed to show a
difference in survival.16,17 In addition to rendering the DTC
maximally avid for iodine, it is also important that the
patient not be exposed to high levels of iodine before RAI
dosing. One common occurrence is the use of CT scans with
iodinated contrast medium in patients being evaluated for
cancer before the diagnosis of DTC has been made. The load
of iodine administered as part of a thoracic CT with contrast
medium is on the order of 30 g, which is 200 times the
recommended daily intake of iodine of 150 mcg/day. If there
is a question concerning a patients compliance with a low
iodine diet, urinary iodine levels can be used as a marker of
iodine consumption, but it is not well established to what
extent dietary iodine levels influence the efficacy of RAI in
this setting.
Once it has been determined that the patient has RAI-
Table 2. Principles in Management of Medullary Thyroid Carcinoma

Category
Principles of Management
RET genetic testing

Serum tumor markers
Staging studies
Genetic testing for germline mutation in RET is recommended for all patients with disease, regardless of positive family history
Calcitonin and carcinoembryonic antigen
Computerized tomography of the neck and chest; triple-phase scan of the abdomen-pelvis (or magnetic resonance imaging of
abdomen).
Magnetic resonance imaging for suspected bone metastasis; x-ray for suspected bone metastasis in long bones.
Bone scan and positron emission tomography are not reliable and should not be routinely used for staging medullary thyroid
carcinoma.
Antidiarrheal medication:
Treatment of hormonal symptoms

(flushing/diarrhea)
Treatment of primary tumor
Treatment of locally recurrent disease

Treatment of bone metastasis
Treatment of metastatic disease
386
short-acting octreotide given subcutaneously; if treatment for 12 wks is effective in ameliorating symptoms, can use octreotide
LAR, given intramuscularly.
Treatment of tumor, including debulking surgery, also palliates these symptoms.
Surgery can be curative for early-stage disease.
External-beam radiation therapy is of questionable benefit, even in the setting of high risk for local recurrence.
Resecting bulky primary tumor, even in setting of metastatic disease, is of value for palliation of refractory symptoms.
Surgery for locally recurrent disease is of value in selected patients.
External-beam radiation therapy is of questionable benefit.
Orthopedic or neurosurgery team should evaluate patients with metastatic lytic lesion in the long bones or fpatients with high-risk
spine metastasis.
External-beam radiation therapy is of great palliative value for symptomatic bone metastasis.
Wait-and-watch approach with periodic restaging studies and monitoring is feasible for patients who have low tumor burden,
asymptomatic disease, andno evidence of progression.
Clinical trial enrollment is encouraged for patients with progressive or symptomatic disease or for patients with high tumor burden.
Cytotoxic chemotherapy (doxorubicin- or dacarbazine-based) is associated with a 030% response rate and is of limited value.
Vandetanib was approved by the Food & Drug Administration in 2011 for patients with progressive or symptomatic metastatic
disease.
EVALUATION OF ADVANCED THYROID CANCER
refractory DTC and the options discussed here have been

exhausted, use of targeted agents can be considered. Three
questions to answer are (1.) Does the patient currently have
symptoms from DTC that would be relieved from reduction
of the tumor? (2.) Based on the clinical history and imaging
studies, are there lesions, although asymptomatic now, that
pose an imminent threat? (3.) Is there evidence from the
patient history, imaging studies, or thyroglobulin levels that
the velocity of tumor progression is such that deferred
treatment might result in a rapid decline in performance
status? Some clinical features can be useful guides for
predicting which patients are likely to have more aggressive
disease (Table 1). As is the case with many common solid
tumors such as lung, colorectal, and breast cancers, often
the decision to treat in the metastatic or recurrent setting is
made on the basis of the extent of disease without knowledge
of the progression velocity. DTC, on the other hand, even in
patients with widely disseminated disease or a high thyroglobulin level, can be indolent, and the timing of treatment
initiation can be challenging in an asymptomatic patient. As
analogies, one can think of patients with prostate cancer and
a high prostate-specific antigen level but no symptoms or
measurable disease or patients with low-grade neuroendocrine tumors with high volume but minimally symptomatic
disease; the art of deciding whom and when to treat remains
a challenge.
Once the decision has been made that a patient with DTC
is appropriate for systemic drug treatment, ample data
support the use of several tyrosine kinase inhibitors (TKIs).
Medullary Thyroid Carcinoma
Unlike DTC, medullary thyroid carcinoma (MTC) is a rare

type of thyroid cancer and accounts for approximately 2% to
8% of all thyroid cancer. However, it is one of the best
characterized solid tumors. MTC is derived from parafollicular cells of the thyroid and exhibit well-differentiated neuroendocrine carcinoma histology. Given that MTC does not
originate from follicular cells of the thyroid, it is not iodine
avid and does not secrete thyroglobulin. MTC occurs in the
sporadic (75%) and hereditary (25%) setting. Germline mutation in the RET proto-oncogene is a critical pathogenetic
defect associated with nearly 95% of all cases of the hereditary type of MTC (MEN-2A, MEN-2B, or familial MTC).
Somatic mutation in the RET proto-oncogene is observed in
30% to 40% of tumors of sporadic MTC. Serum calcitonin
and carcinoembryonic antigen (CEA) are reliable tumor
markers with prognostic significance. Patients with MTC may
Sidebar 3. Guidelines for Management of Thyroid

Cancer
The following organizations provide guidelines for

the management of various types of thyroid cancer.
National Comprehensive Cancer Network (NCCN):

www.nccn.org
American Thyroid Association (ATA): http://thyroid
guidelines.com
British Thyroid Association (BTA): www.britishthyroid-association.org/Guidelines
have hormone-associated symptoms of flushing and diarrhea.

MTC is typically a systemic disease and is known to spread in
the setting of small primary tumors. In addition to lymph
nodes in the neck, common sites of metastasis include the
lung, mediastinal lymph nodes, liver, and bones. While cytotoxic chemotherapies are ineffective, systemic therapies with
TKIs are promising. Over last decade, numerous clinical trials
ranging from phase I and phase III trials have been conducted
in patients with advanced MTC. Vandetanib is approved by
FDA in 2011 for patients with progressive advanced MTC. We
summarize the principles of diagnosis and management in
Table 2.
Anaplastic Thyroid Carcinoma
Anaplastic thyroid cancer (ATC) accounts for 2% of all

thyroid cancer. It is one of the most aggressive cancers among
all solid tumors and its progression can be clinically evident in
a matter of days to weeks. ATC is classified only as stage IV
and is divided into stage IVA, IVB, and IVC. Despite multimodality aggressive therapy, the prognosis for patients with
ATC is poor, with a median survival of 6 months. Treatment
options usually include combinations of surgery, cytotoxic
chemotherapy, and radiation therapy. Cytotoxic chemotherapy agents such as taxanes, platinums, or doxorubicin can be
used. Concurrent or sequential chemoradiation treatment is
an option. Supportive care and hospice should be mentioned
to the patient when discussing management of the disease.
In addition to multimodality therapy, other topics that
should be addressed are airway management and nutritional status, with consideration of a percutaneous endoscopic
gastrostomy tube. Despite recent advances in the understanding of the pathogenetics of ATC, novel targeted therapies
tested to date have proved to be ineffective. Patients should be
encouraged to participate in clinical trials.
Author
A. Dimitrios Colevas
Manisha H. Shah
Employment or
Leadership
Positions
Consultant or
Advisory Role
Bayer
Stock
Ownership
Honoraria
Research
Funding
ActoGeniX;
Bayer/Onyx;
Boehringer
Ingelheim;
Exelixis;
GlaxoSmithKline;
Roche
Daiichi Sankyo;
Eisai; Exelixis
Expert
Testimony
Other
Remuneration
387
COLEVAS AND SHAH
REFERENCES
1. Mazzaferri EL. Treating high thyroglobulin with radioiodine: a magic
bullet or a shot in the dark? J Clin Endocrinol Metab. 1995;80:1485-1487.
2. Schlumberger M, Mancusi F, Baudin E, et al. 131I therapy for elevated
thyroglobulin levels. Thyroid. 1997;7:273-276.
3. Pineda JD, Lee T, Ain K, et al. Iodine-131 therapy for thyroid cancer
patients with elevated thyroglobulin and negative diagnostic scan. J Clin
Endocrinol Metab. 1995;80:1488-1492.
4. Besic N, Auersperg M, Gazic B, et al. Neoadjuvant chemotherapy in 29
patients with locally advanced follicular or Hurthle cell thyroid carcinoma: a
phase 2 study. Thyroid. 2012;22:131-137.
5. Bernhardt B. Follicular thyroid carcinoma: response to chemotherapy.
Am J Med Sci. 1981;282:45-46.
6. Gottlieb JA, Hill CS, Jr. Chemotherapy of thyroid cancer with Adriamycin. Experience with 30 patients. N Engl J Med. 1974;290:193-197.
7. Matuszczyk A, Petersenn S, Bockisch A, et al. Chemotherapy with
doxorubicin in progressive medullary and thyroid carcinoma of the follicular
epithelium. Horm Metab Res. 2008;40:210-213.
8. Shimaoka K, Schoenfeld DA, DeWys WD, et al. A randomized trial of
doxorubicin versus doxorubicin plus cisplatin in patients with advanced
thyroid carcinoma. Cancer. 1985;56:2155-2160.
9. Spano JP, Vano Y, Vignot S, et al. GEMOX regimen in the treatment of
metastatic differentiated refractory thyroid carcinoma. Med Oncol. 2011 Sept
25.
10. Jonklaas J, Sarlis NJ, Litofsky D, et al. Outcomes of patients with
differentiated thyroid carcinoma following initial therapy. Thyroid. 2006;16:
1229-1242.
11. Sugitani I, Fujimoto Y. Does postoperative thyrotropin suppression
therapy truly decrease recurrence in papillary thyroid carcinoma? A randomized controlled trial. J Clin Endocrinol Metab. 2010;95:4576-4583.
12. Burmeister LA, Goumaz MO, Mariash CN, et al. Levothyroxine dose
requirements for thyrotropin suppression in the treatment of differentiated
thyroid cancer. J Clin Endocrinol Metab. 1992;75:344-350.
13. Nanni C, Rubello D, Fanti S, et al. Role of 18F-FDG-PET and PET/CT
imaging in thyroid cancer. Biomed Pharmacother. 2006;60:409-413.
14. Mosci C, Iagaru A. PET/CT imaging of thyroid cancer. Clin Nucl Med.
2011;36:e180-e185.
15. McDougall IR, Iagaru A. Thyroid stunning: fact or fiction? Semin Nucl
Med. 2011;41:105-112.
388
16. Ladenson PW, Braverman LE, Mazzaferri EL, et al. Comparison of

administration of recombinant human thyrotropin with withdrawal of thyroid
hormone for radioactive iodine scanning in patients with thyroid carcinoma.
N Engl J Med. 1997;337:888-896.
17. Tala H, Robbins R, Fagin JA, et al. Five-year survival is similar in
thyroid cancer patients with distant metastases prepared for radioactive
iodine therapy with either thyroid hormone withdrawal or recombinant
human TSH. J Clin Endocrinol Metab. 2011;96:2105-2111.
18. Xing M, Westra WH, Tufano RP, et al. BRAF mutation predicts a
poorer clinical prognosis for papillary thyroid cancer. J Clin Endocrinol
Metab. 2005;90:6373-6379.
19. Lanzi C, Cassinelli G, Nicolini V, et al. Targeting RET for thyroid
cancer therapy. Biochem Pharmacol. 2009;77:297-309.
20. Robbins RJ, Wan Q, Grewal RK, et al. Real-time prognosis for metastatic thyroid carcinoma based on 2-[18F]fluoro-2-deoxy-D-glucose-positron
emission tomography scanning. J Clin Endocrinol Metab. 2006;91:498-505.
21. Miyauchi A, Kudo T, Miya A, et al. Prognostic impact of serum
thyroglobulin doubling-time under thyrotropin suppression in patients with
papillary thyroid carcinoma who underwent total thyroidectomy. Thyroid.
2011;21:707-716.
22. Klein M, Vignaud JM, Hennequin V, et al. Increased expression of
the vascular endothelial growth factor is a pejorative prognosis marker in
papillary thyroid carcinoma. J Clin Endocrinol Metab. 2001;86:656-658.
23. Chrisoulidou A, Boudina M, Tzemailas A, et al. Histological subtype is
the most important determinant of survival in metastatic papillary thyroid
cancer. Thyroid Res. 2011;4:12.
24. Ito Y, Higashiyama T, Takamura Y, et al. Clinical outcomes of patients
with papillary thyroid carcinoma after the detection of distant recurrence.
World J Surg. 2010;34:2333-2337.
25. Shoup M, Stojadinovic A, Nissan A, et al. Prognostic indicators of
outcomes in patients with distant metastases from differentiated thyroid
carcinoma. J Am Coll Surg. 2003;197:191-197.
26. Showalter TN, Siegel BA, Moley JF, et al. Prognostic factors in patients
with well-differentiated thyroid cancer presenting with pulmonary metastasis. Cancer Biother Radiopharm. 2008;23:655-659.
27. Wells SA, Jr, Robinson BG, Gagel RF, et al. Vandetanib in patients
with locally advanced or metastatic medullary thyroid cancer: a randomized,
double-blind phase III trial. J Clin Oncol. 2012;30:134-141.
Systemic Therapeutic Approaches to

Advanced Thyroid Cancers
By Michael E. Menefee, MD, Robert C. Smallridge, MD, and Keith C. Bible, MD, PhD,
and the Mayo Clinic Endocrine Malignancies Disease-Oriented Group
Overview: Until only recently, few effective systemic therapies were available to treat patients with metastatic thyroid
cancers. Recent advances in better understanding the pathogenesis and altered signaling pathways especially in medullary and differentiated thyroid cancers (MTCs and DTCs)
have begun to change this situation substantially. Vandetanib,
an orally bioavailable inhibitor of the RET kinase that is
constitutively activated in MTC, has now been approved by the
U.S. Food and Drug Administration (FDA) for use in progressive and symptomatic metastatic MTC; it has been shown to
delay time to progression relative to placebo in a randomized
phase III trial. Further, vascular endothelial growth factor
receptor (VEGF-R) inhibitory agents including sorafenib,
sunitinib, pazopanib, and axitinib that are already approved in
HYROID CANCER has become increasingly more common worldwide, doubling in incidence during the last
decade in the United States to place it among the 10 most
common cancers1 and prompting increasing media and public attention. Although thyroid cancers have historically
been more in the purview of endocrinologists than of medical
oncologists, recent therapeutic innovations have made it
increasingly important that medical oncologists develop
expertise in the management of thyroid cancers, especially
when advanced.2 This expertise not only must include familiarity with the array of available and emerging therapeutics for various histologic subtypes of thyroid cancer, but also
requires a working knowledge of when to apply available
systemic therapies. In this regard, because most patients
even with advanced differentiated and medullary thyroid
cancers (DTCs and MTCs) are subject to slowly progressive
disease compared with most other cancers encountered in
medical oncology practice, optimal management often involves the delayed and conservative application of systemic
therapies. Anaplastic thyroid cancer (ATC), in contrast, is
perhaps the most aggressive of all solid tumors with a
doubling time as brief as 10 daysthereby demanding
immediate action. The following sections review recent progress in developing more effective systemic therapies for
advanced thyroid cancers (Table 1), discussing each of the
three major histotypes in separate sections, and providing
general rationale for the application of available systemic
therapies.
Systemic Therapeutic Approaches to Advanced DTC
DTC arises from the thyroxine-producing follicular cells of

the thyroid gland and occurs in two primary histotypes:
papillary thyroid cancer (PTC), the most incident of all
thyroid cancers that accounts for the majority of all cases
and is the most indolent histotype overall; and follicular
thyroid cancer (FTC), encompassing also the Hurthle cell
(HCC) histologic variant, which tends to be somewhat more
aggressive in behavior. Because DTC arises from follicular
thyroid cells that accumulate iodine as required for thyroid
hormone production, most DTCs at least initially retain the
ability to concentrate iodine facilitated by a functional
sodium iodine symporter (NIS). As a consequence, therapeu-
the United States for use in advanced renal cell carcinoma

have shown high response rates in treating advanced DTCs in
multiple phase II trials, and have become commonly used in
progressive radioiodine-refractory metastatic DTC. Yet additional agents are now in development, with several including
XL184 (cabozantinib) also showing promise in DTC and MTC.
In anaplastic thyroid cancer (ATC), progress has been slower,
with the greatest apparent gains resulting more from the
application of systemic therapies earlier in the disease
course, especially when used in conjunction with initial surgical and radiation therapies. Despite recent progress, additional effective systemic therapeutic approaches remain
sorely needed for treating metastatic MTC, DTC, and ATC.
tic RAI remains the standard of care for initial therapy in

metastatic DTC.
It is critical to realize that only a small minority of all
DTC patients have 1) RAI-refractory metastatic disease,
2) disease not amenable to focal palliative approaches, and
3) rapidly progressive, symptomatic, and imminently threatening disease sufficient to prompt consideration of additional systemic therapies beyond RAI. In decisions related to
systemic therapy it is, therefore, important to carefully weigh
the anticipated risks and benefits of a candidate treatment
compared with the risks imposed by the disease itself.
Because cytotoxic chemotherapy has historically had minimal efficacy in DTC, recent efforts have instead focused
primarily on alternative approaches, summarized Harris
and Bible.2 In part, recent advances in DTC therapy arose
in parallel with the development of vandetanib for use
in treating advanced MTC. Vandetanib is a small-molecule
inhibitor of tyrosine kinasesincluding the RET kinase that
is of pathogenic importance in MTC because of its constitutive activation. Because RET is also upregulated in some
DTCs, interest consequently also developed in parallel in
applying kinase inhibitors in DTCs, resulting in multiple
therapeutic clinical trials of tyrosine kinase inhibitors
(TKIs) also in DTC.
Strikingly, response rates to TKIs in advanced DTC have
ranged from 10% to 50%, with attained responses often quite
durable. Generally, all tested VEGF-R inhibitory TKIs have
activity in DTC, with pazopanib,3 sorafenib,4,5 and axitinib6
seeming to have the highest response rates and most favorable toxicity profiles according to published results. Currently, a randomized phase III trial of sorafenib compared
with placebo is ongoing in RAI-refractory DTC (clinicaltrials.
From the Division of Medical Oncology; Division of Endocrinology, Mayo Clinic Florida,
Jacksonville, FL; Division of Medical Oncology, Mayo Clinic, Rochester, MN.
Address reprint requests to Keith C. Bible, MD, PhD, Mayo Clinic, 200 First Street SW,
Rochester, MN 55901; email: bible.keith@mayo.edu.
1092-9118/10/1-10
389
MENEFEE ET AL
gov identifier NCT00984282), with other TKI trials also

ongoing and planned in RAI-refractory metastatic DTC.
Another antiangiogenic strategy that has been successful,
but less well studied, has been the use of immunomodulatory agents (IMiDs). The effects of the IMiDs in thyroid
cancer may extend beyond their impact on angiogenesis.
Thalidomide,7 and more recently the thalidomide analog
lenalidomide,8 have produced clinical benefit in patients
with DTC, with response rates ranging between 18% and
22% and stable disease between 32% and 44%. The IMiDs
are another therapeutic option in the limited, although
rapidly expanding, armamentarium against DTC.
The paradigm indicating that cytotoxic therapies are
ineffective in DTC has also been changing. For example, the
combination of gemcitabine and oxaliplatin (GEMOX) was
recently reported to produce a 57% response rate, with 7%
complete responses9; these results rival or are superior to
outcomes reported from the use of TKIs in DTC. Hence,
newer cytotoxic agents and combinations such as GEMOX
KEY POINTS
390
Vandetanib, an orally bioavailable inhibitor of the

RET kinase that is constitutively activated in medullary thyroid cancer (MTC), has been approved by the
U.S. Food and Drug Administration for use in treating progressive and symptomatic metastatic MTC; it
has been shown to delay time to progression compared with placebo in a randomized phase III trial.
Vascular endothelial growth factor receptor
(VEGF-R) inhibitory and antiangiogenic agents including sorafenib, sunitinib, axitinib, and pazopanib
have shown high response rates in treating advanced
differentiated thyroid cancers (DTCs) in multiple
phase II trials, and are now commonly used in treating patients with progressive radioiodine (RAI)refractory metastatic DTC.
In the decision to initiate systemic therapies in metastatic and progressive DTC or MTC, considerable
restraint should be exercised, as many patients will
have slowly progressive disease not requiring active
systemic therapy. Alternatives for focal palliation of
areas of threatening disease (especially in the neck
and in bone) should first be prominently considered
before initiation of systemic therapies. Moreover, the
risk imposed by the disease versus by the systemic
approaches under consideration should be carefully
weighed.
Application of systemic therapies earlier in the course
of anaplastic thyroid cancer (ATC) has begun to show
promise, especially when used in conjunction with
initial surgical and radiation therapies.
Although considerable recent progress has been
made in better understanding candidate therapeutic
targets and approaches to treating advanced thyroid
cancers, additional and more effective systemic therapies for these cancers remain sorely needed.
Table 1. Selected Therapeutics for Advanced Thyroid Cancers

Histotype
Differentiated thyroid
cancer
Medullary thyroid
cancer
Anaplastic thyroid
cancer
Agent
Molecular Target(s)
Axitinib
Gemcitabine/oxaliplatin9
Lenalidomide8
Pazopanib3
Thalidomide7
Sorafenib4,5
Sunitinib23
Vandetanib10
Cabozantinib11
Kinases, including VEGF-R

DNA
Incompletely defined
Incompletely defined
Kinases, including RET
Kinases, including RET, MET,
VEGF-R
Microtubules
Microtubules
Topoisomerase II/DNA
Microtubules
Crolibulin17,18
Docetaxel16
Doxorubicin cisplatin19
Paclitaxel15
Abbreviation: VEGF-R, vascular endothelial growth factor receptor.
provide an alternative approach to TKIs to consider in

patients with RAI-refractory DTC.
Although there has been considerable progress in developing therapies for patients with rapidly progressive and
imminently threatening radioactive iodine-insensitive DTC,
most patients with DTC (even those with metastatic disease) are subject to overall indolent disease courses. As a
result, there should be prominent consideration of focal
palliation of locally threatening disease (especially in the
neck and in bone) before initiation of systemic therapies.
Moreover, the risks imposed by the disease versus those
imposed by contemplated systemic therapies must be carefully weighed so as to ensure that the initiation of systemic
therapy is in a particular patients best interests. Almost all
patients receiving treatment with TKIs will experience
adverse effects, with fatigue, diarrhea, induced hypertension, and cutaneous toxicities overall quite common, and
with serious adverse effects also encountered in a minority
of patients. Similarly, the application of cytotoxic chemotherapy is often associated with nausea, cytopenias, and
other adverse effects that have potential to adversely affect
patient quality of life and impose risks to patient welfare.
Overall, judicious application of available systemic therapies is therefore required in advanced DTC.
Systemic Therapeutic Approaches to Advanced MTC
As noted earlier herein, MTC has proven to be an excellent

model system in the development and study of novel therapeutics in thyroid cancer. In particular, MTC is commonly
heritable and characterized by constitutive activation of the
RET kinasetriggered by a germ-line mutation in the case
of heritable MTC, and alternatively by tumor-specific mutation in the case of sporadic MTC. RET activation drives
proliferation in affected parafollicular C-cells within the
thyroid, contributing thereby to MTC pathogenesis and
serving as a specific mutational therapeutic molecular target in MTC.
The RET inhibitor vandetanib has been evaluated in
phase II trials and in a randomized phase III trial compared
with placebo, with results indicating a high response rate
and improved progression-free survival in patients receiving
vandetanib compared with those receiving placebo (as a
result of the cross-over trial design; however, the effects of
vandetanib on overall survival are not reliably assessable).10
As a result, vandetanib was approved by the U.S. Food and
APPROACHES TO ADVANCED THYROID CANCERS
Drug Administration (FDA) for use in progressive and

symptomatic metastatic MTC in 2011.
Other RET-inhibitory small molecules have also shown
promise in MTC, including XL184 (cabozantinib),11 which
has also been subject to a recent randomized phase III trial
in MTC that has yielded apparently promising results,
but that has not yet been published. Furthermore, VEGFinhibitory TKIs also appear to have activity in MTC and are
the subject of further study in advanced MTC.2 It should
also be noted that cytotoxic chemotherapy has shown some
activity in MTC, albeit perhaps more modest in extent than
seen in response to TKI therapy.12-14
As noted above in the case of DTC, potential risks and
benefits of candidate systemic therapeutic approaches
should be carefully considered before their initiation in
patients with MTC because many such patients will have
indolent disease courses most often not requiring systemic
therapies.
Systemic Therapeutic Approaches to Advanced ATC
Cytotoxic chemotherapy remains the most effective, albeit

marginally so, approach to treating advanced ATC. The two
classes of agents with highest activity overall in ATC are
anthracyclines and antimicrotubule inhibitors, with the
greatest amount of data available in the case of paclitaxel. In
particular, the CATCHIT trial demonstrated single-agent
paclitaxel to produce a transient response rate of 53% in
ATC but the requirement for confirmation of response had
to be shortened to 2 weeks to yield this result15; to be sure,
responses to most any systemic therapy in advanced ATC
tend to be very brief. The related taxane docetaxel has
also shown some activity in ATC.16 EPC2407 (Crolibulin;
EpiCept Corporation, Tarrytown, NY) have also shown
some promise in ATC,17,18 with a phase II trial of EPC2407
combined with cisplatin now ongoing (clinicaltrials.gov identifier NCT01240590).
Doxorubicin, approved by the U.S FDA for use in advanced thyroid cancer in the 1970s, has been used as a single
agent in thyroid cancer, as well as combined with cisplatin,
with somewhat better outcomes resulting from the twoagent combination.19
Although results from evaluation of TKIs in ATC have
been disappointing overall, imatinib monotherapy was recent reported to induce responses in several patients with
ATC.20 Further, the combination of the TKI pazopanib with
paclitaxel has produced sufficiently encouraging preclinical

results to be subject to evaluation in a randomized Radiation
Oncology Treatment Group (RTOG) of intensity-modulated
radiotherapy plus paclitaxel with or without pazopanib
(clinicaltrials.gov identifier NCT01236547). Yet additional
agents and combinations are actively being evaluated in
advanced ATC.2
Also of potential promise in ATC is the application of
systemic therapies in conjunction with the initial treatment
of neck-confined disease especially when systemic therapy
is combined with radiation therapy. In applying this approach, several groups have observed unexpectedly favorable outcomes and survival,21,22 suggesting that it may be
fruitful in patients with good performance status seeking an
aggressive approach to their initial therapy.
Conclusion
Historically, progress had been slow in developing more

effective systemic approaches to treating advanced thyroid
cancers. This situation has recently begun to change with
the elaboration of an increasing amount of information
related to the molecular pathways contributing to thyroid
cancer pathogenesis, thereby resulting in the identification
of an ever-increasing array of candidate therapeutic molecular targets that has already begun to lead to therapeutic
advances in thyroid cancers. In particular, the RET kinase
inhibitor vandetanib has been identified as sufficiently
promising to merit its approval by the U.S. FDA for use in
metastatic progressive and symptomatic medullary thyroid
cancer, because it has been shown to delay time to progression and to induce a high rate of clinical response in MTC.
Additionally, VEGF-Rinhibitory multi-targeted kinase inhibitors including sunitinib, sorafenib, axitinib, and pazopanib have demonstrated high rates of durable clinical
responses in DTC, prompting their application as a new
standard of care in treating patients with metastatic
RAI-refractory DTC. Although therapeutic progress in ATC
has been slow, emerging evidence suggests promise in the
early application of cytotoxic chemotherapy in conjunction
with other up-front therapies (e.g., radiation therapy and
surgery) for locoregionally confined ATC. Nevertheless, despite considerable recent progress, further basic advances
and therapeutic progress related to thyroid cancers are still
very much needed, especially for patients with rapidly
progressive metastatic disease.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Michael E. Menefee*
Robert C. Smallridge*
Keith C. Bible*
REFERENCES
J Clin. 2012;62:10-29.
2. Harris PJ, Bible KC. Emerging therapeutics for advanced thyroid
malignancies: rationale and targeted approaches. Expert Opin Investig Drugs.
2011;20:1357-1275.
3. Bible KC, Suman VJ, Molina JR, et al. Efficacy of pazopanib in
progressive, radioiodine-refractory, metastatic differentiated thyroid cancers:

results of a phase 2 consortium study. Lancet Oncol. 2010;11:962-972.
4. Kloos RT, Ringel MD, Knopp MV, et al. Phase II trial of sorafenib in
metastatic thyroid cancer. J Clin Oncol. 2009;27:1675-1684.
5. Gupta-Abramson V, Troxel AB, Nellore A, et al. Phase II trial of
sorafenib in advanced thyroid cancer. J Clin Oncol. 2008;26:4714-4719.
391
MENEFEE ET AL
6. Cohen EE, Rosen LS, Vokes EE, et al. Axitinib is an active treatment
for all histologic subtypes of advanced thyroid cancer: results from a phase II
7. Ain KB, Lee C, Williams KD. Phase II trial of thalidomide for therapy
of radioiodine-unresponsive and rapidly progressive thyroid carcinomas.
Thyroid. 2007;17:663-670.
8. Ain KB, Lee C, Holbrook KM, et al. Phase II study of lenalidomide in
distantly metastatic, rapidly progressive, and radioiodine-unresponsive thyroid
carcinomas: preliminary results. J Clin Oncol. 2008;26:XXX (suppl; abstr 6027).
9. Spano JP, Vano Y, Vignot S, et al. GEMOX regimen in the treatment
of metastatic differentiated refractory thyroid carcinoma. Med Oncol. Epub
2011 Sep 25.
10. Wells SA Jr, Robinson BG, Gagel RF, et al. Vandetanib in Patients
With Locally Advanced or Metastatic Medullary Thyroid Cancer: A Randomized, Double-Blind Phase III Trial. J Clin Oncol. 2012;30:134-141.
11. Kurzrock R, Sherman SI, Ball DW, et al. Activity of XL184 (Cabozantinib), an oral tyrosine kinase inhibitor, in patients with medullary thyroid
12. Nocera M, Baudin E, Pellegriti G, et al. Treatment of advanced
medullary thyroid cancer with an alternating combination of doxorubicinstreptozocin and 5 FU-dacarbazine. Groupe dEtude desTumeurs a` Calcitonine (GETC). Br J Cancer. 2000;83:715-718.
13. Orlandi F, Caraci P, Berruti A, et al. Chemotherapy with dacarbazine
and 5-fluorouracil in advanced medullary thyroid cancer. Ann Oncol. 1994;5:
763-765.
14. Wu LT, Averbuch SD, Ball DW, et al. Treatment of advanced medullary
thyroid carcinoma with a combination of cyclophosphamide, vincristine, and
dacarbazine. Cancer. 1994;73:432-436.
15. Ain KB, Egorin MJ, DeSimone PA. Treatment of anaplastic thyroid
392
carcinoma with paclitaxel: phase 2 trial using ninety-six-hour infusion.

Collaborative Anaplastic Thyroid Cancer Health Intervention Trials (CATCHIT)
Group. Thyroid. 2000;10:587-594.
16. Kawada K, Kitagawa K, Kamei S, et al. The feasibility study of
docetaxel in patients with anaplastic thyroid cancer. Jpn J Clin Oncol.
2010;40:596-599.
17. Dowlati A, Robertson K, Cooney M, et al. A phase I pharmacokinetic
and translational study of the novel vascular targeting agent combretastatin
a-4 phosphate on a single-dose intravenous schedule in patients with advanced cancer. Cancer Res. 2002;62:3408-3416.
18. Mooney CJ, Nagaiah G, Fu P, et al. A phase II trial of fosbretabulin in
advanced anaplastic thyroid carcinoma and correlation of baseline serumsoluble intracellular adhesion molecule-1 with outcome. Thyroid. 2009;19:
233-240.
19. Shimaoka K, Schoenfeld DA, DeWys WD, et al. A randomized trial of
doxorubicin versus doxorubicin plus cisplatin in patients with advanced
thyroid carcinoma. Cancer. 1985;56:2155-2160.
20. Ha HT, Lee JS, Urba S, et al. A phase II study of imatinib in patients
with advanced anaplastic thyroid cancer. Thyroid. 2010;20:975-980.
21. Foote RL, Molina JR, Kasperbauer JL, et al. Enhanced survival in
locoregionally confined anaplastic thyroid carcinoma: a single-institution
experience using aggressive multimodal therapy. Thyroid. 2011;21:25-30.
22. Troch M, Koperek O, Scheuba C, et al. High efficacy of concomitant
treatment of undifferentiated (anaplastic) thyroid cancer with radiation and
docetaxel. J Clin Endocrinol Metab. 2010;95:E54-57.
23. Carr LL, Mankoff DA, Goulart BH, et al. Phase II study of daily
sunitinib in FDG-PET-positive, iodinerefractory differentiated thyroid cancer
and metastatic medullary carcinoma of the thyroid with functional imaging
correlation. Clin Cancer Res. 2010;16:5260-5268.
AVOIDING OVERDIAGNOSIS AND

OVERTREATMENT OF COMMON CANCERS
CHAIR
Ian M. Thompson Jr, MD
University of Texas Health Science Center at San Antonio
San Antonio, TX
SPEAKERS
Ruth D. Etzioni, PhD
Seattle, WA
San Francisco, CA
Overdiagnosis and Overtreatment of

Prostate Cancer
By Ian M. Thompson Jr., MD
Overview: Prostate cancer is a ubiquitous disease, affecting

as many as two-thirds of men in their 60s. Through widespread
prostate-specific antigen (PSA) testing, increasing rates of
prostate biopsy, and increased sampling of the prostate, a
larger fraction of low-grade, low-volume tumors have been
detected, consistent with tumors often found at autopsy.
These tumors have historically been treated in a manner
similar to that used for higher-grade tumors but, more recently, it has become evident that with a plan of active
surveillance that reserves treatment for only those patients
whose tumors show evidence of progression, very high
disease-specific survival can be achieved. Unfortunately, the
frequency of recommendation of an active surveillance strat-
ITH THE diffusion of PSA screening that first began

in the mid-1980s and with subsequent changes in
the way we diagnose prostate cancer, the tumors of today
are remarkably different from those of the 1980s or even
1990s. At its most basic level, the opportunity for overdiagnosis of prostate cancer (overdiagnosis can be defined as
detection of cancers that will ultimately not harm the host)
must be understood to be related to the background rate of
the disease. The best insight into this potential opportunity
for diagnosis of prostate cancer comes from autopsy studies
of men who died as a result of other causes. Careful sectioning of prostates in these men gives us an idea as to the lower
bound estimate for the risk of the disease. (It is the lower
bound estimate because men who have been treated with
surgery or radiation are effectively censored from these
analyses.) Nonetheless, these rates are substantial, as summarized in Table 1.1 Given that a mans life expectancy
approaches 80 years in the United States, it would seem the
infrequent aging man who does not harbor at least a small
tumor in his prostate that, if struck with a prostate biopsy
needle, would result in a prostate cancer diagnosis. Indeed,
to make a diagnosis of prostate cancer, which is generally
asymptomatic until metastases develop, there are three
prerequisites: 1) the physician must suspect cancer, 2) the
patient must accept a prostate biopsy, and 3) the biopsy
needle must strike the tumor. As will be seen, these rates
have changed variably over the years.
How Is the Prostate Cancer of 2012 Different from
the Prostate Cancer of Prior Decades?
Physician Suspicion of Prostate Cancer
Before PSA, the only method for a physician to suspect the

presence of prostate cancer (at least, early prostate cancer)
was through an abnormal digital rectal examination (DRE).
This was generally uncommon. In one series we conducted
prospectively in the early 1980s, in 2,005 men undergoing
biopsy, only 65 abnormal examinations were identified.2
Of these, only 17 proved to be cancer. With the advent of
PSA testing, it was not necessarily a substantially greater
percentage of men in whom prostate cancer was suspected
(approximately 8% of the population has a PSA greater than
4.0 ng/mL) but a far greater fraction of the population was
interested in PSA testing, likely as a result of the poor
egy in the United States is low. An alternative strategy to

improve prostate cancer detection is through selected biopsy
of those men who are at greater risk of harboring high-grade,
potentially lethal cancer. This strategy is currently possible
through the use of risk assessment tools such as the Prostate
Cancer Prevention Trial Risk Calculator (www.prostate.cancer.
risk.calculator.com) as well as others. These tools can predict
with considerable accuracy a mans risk of low-grade and
high-grade cancer, allowing informed decision making for the
patient with a goal of detection of high-risk disease. Ultimately, other biomarkers including PCA3, TMPRSS2:ERG, and
[-2]proPSA will likely aid in discriminating these two types of
cancer before biopsy.
acceptance of DRE by U.S. males. It was then with this

testing that we witnessed a spike in prostate cancer diagnosis in the early 1990s, which subsequently fell back (after a
harvest of prevalent tumors) to a rate approximately double
that of the pre-PSA 1980s. Presently, approximately 75
percent of at-risk men have undergone a PSA test and
approximately 50% are tested regularly.
Further changes in suspicion of prostate cancer then
developed. These included 1) the recognition that prostate
cancer was quite prevalent at lower levels of PSA, 2) the
recognition that other factors such as family history, race,
and age could be incorporated into a risk assessment to lead
to a biopsy in a man with a PSA less than 4.0 ng/mL, and
3) the use of changes in PSA (PSA velocity) to prompt biopsy
at even lower levels of PSA.3-5 All of these biopsy prompts
have led to dramatic increases in the use of prostate biopsy.
The sum total of these events has dramatically increased the
likelihood that a man, if a PSA is obtained, will receive a
recommendation to consider prostate biopsy.
Patient Acceptance of Prostate Biopsy
This variable can dampen the rate of prostate cancer

diagnosis because many men who receive PSA results that
are greater than 4.0 ng/mL will opt to not have a prostate
biopsy. In the Prostate, Lung, Colorectal and Ovarian
(PLCO) Cancer Screening Trial in the United States, for
example, at the baseline PSA test, if a mans PSA exceeded
4.0 ng/mL, within 1 and 3 years, the likelihood that he
underwent prostate biopsy was 41% and 64%, respectively.6
Almost certainly, the way the PSA data are presented by the
physician will play a part in whether the patient undergoes
prostate biopsy.
From the Cancer Therapy and Research Center, University of Texas Health Science
Center at San Antonio, San Antonio, TX.
Address reprint requests to Ian M. Thompson Jr., MD, Cancer Therapy and Research
Center, University of Texas Health Science Center at San Antonio, 7979 Wurzbach Road,
San Antonio, TX; email: thompsoni@uthscsa.edu.
1092-9118/10/1-10
e35
IAN M. THOMPSON JR.

Table 1. Prevalence of Prostate Cancer at Autopsy
Age (Years)
U.S. PrevalenceWhites (%)
U.S. PrevalenceBlacks (%)
37
44
65
83
43
46
70
81
4150
5160
6170
7180
The Prostate Needle Must Strike the Tumor
Urologists who practiced in the pre-PSA era had the

challenging experience of performing prostate biopsies with
instruments that oftentimes poorly sampled the prostate.
Additionally, it was not uncommon for a physician to obtain
as few as two biopsy cores from the prostate in determining
whether cancer was present. Certainly, because prostate
cancer is multifocal and oftentimes occupies a small fraction
of the gland, this poor sampling of the prostate missed many
small tumors. On the other hand, if cancer was detected, it
often was large. Before the late 1980s, the most common
biopsy scheme was four biopsy cores: two from each side of
the gland (left/right) and, on each side, one core from the
base and one core from the apex (cranial and caudal,
respectively). In the late 1980s, Hodge and colleagues suggested the use of a sextant biopsy, arguing that six cores
were more likely to identify cancer than two or four.7 Over
the ensuing years, authors found that increasing the biopsies to 10 or 12 cores further increased the likelihood of
cancer diagnosis.8 More recently, so-called saturation biopsies have been advocated in which, in some schemes, cores
are obtained from every cubic centimeter of the prostate,
totaling 20 40 biopsy samples.9
To even the most nave observer, this ever-increasing
number of biopsy cores would clearly increase the risk of
KEY POINTS
e36
Prostate cancer is very common in the aging U.S.

male population and is most commonly a low-grade,
low-volume tumor that poses a very low risk of
progression and death.
PSA testing and prostate biopsy have increasingly
resulted in detection of low-risk tumors; nonetheless,
these tumors are very often treated radically with
surgery or radiation.
It is possible, through the use of risk assessment
tools, to identify men who are most likely to have
high-risk cancer (for biopsy) and those who are most
likely to have low-risk, potentially indolent cancer (to
not have a biopsy).
A range of novel biomarkers are in the process of
validation and are predictive of high-risk prostate
cancer.
The future of prostate cancer detection will likely
incorporate prostate-specific antigen and other biomeasures (e.g., digital rectal examination, age, body
mass index), as well as rationally incorporate novel
biomarkers associated with risk of high-grade cancer,
to identify the man who is most likely to benefit from
diagnosis and, ultimately, treatment of cancer.
detection of very small, potentially-inconsequential prostate

cancer. It must be understood, however, that in the early
days of PSA testing, physicians were frustrated by patients
who underwent a prostate biopsy in which a small amount of
cancer was found, and then, on prostatectomy, it was found
that the tumor was large and extraprostatic. It was thus
with good intentions that this increase in the number of
biopsy cores occurred; the natural result, however, was that
smaller and smaller tumors were detected.
The cumulative effect of these changes during the last
30 years has been a growing fraction of patients with
small, low-grade cancers detected. Increasing this rate
have been several other events. First, many patients are
now having multiple sets of biopsies as they see physicians
who continue to react to elevated PSA levels. Effectively
then, a man who has had one negative 12-core biopsy in
2007, followed by another negative 12-core biopsy in 2009,
and then has another 12-core biopsy in 2011 that shows
cancer in one core, has effectively had a 36-core biopsy with
one of 36 cores showing a low-grade cancer. For many of
these men, it is highly likely that the cancer was indeed
present in 2007 and 2009 and just not struck by the needle.
The second event is that high-volume prostate cancers that
were present in the 1990s and early 2000s and detected on a
first biopsy after screening, have generally been treated and
effectively removed from the screened population. Thus,
patients who are currently being screened often have undergone multiple PSA tests and may have had multiple negative biopsies, and thus are at intrinsically lower risk of a
high-volume tumor.
What Is the Evidence that Prostate Cancers Are
Overdetected or Overtreated?
Regarding overdetection, a fairly clear piece of evidence in

support of this is the ratio of annual incidence to mortality.
In 2011, it was expected that 241,740 prostate cancers would
be detected and that 28,170 men would die as a result of
their cancer.10 If no overdetection occurred, the ratio would
be closer to 1:1 (by contrast, the rates for lung cancer are
226,160 and 160,340, respectively). An additional piece of
evidence can be seen through the lifetime risk of prostate
cancer (now 16.48%) as opposed to the lifetime risk of
prostate cancer death (now 2.77%).11 It is important to
recognize that this lifetime risk of prostate cancer is this
high with only approximately 50% of the population being
screened for prostate cancer on a regular basis; one might
expect a rate of 20% to 30% if all men underwent screening
annually.
There is also clear evidence of overtreatment. For the
purposes of this article, we will define overtreatment as an
active treatment (e.g., surgery, radiation, hormones, or a
combination of these) in a patient who would never have
had symptoms nor have died as a result of his prostate
cancer. The primary evidence of this comes from the many
series of patients with prostate cancer who have been
managed with either watchful waiting (WW) or active surveillance (AS). It is important to distinguish these two
management plans. In the case of WW, the patient undergoes no monitoring and is not offered an intervention for
cure; he receives treatment only should symptoms or evidence of metastatic disease develop. In the case of AS, a
low-risk patient is monitored with serial PSA and DRE
testing and on a periodic basis (e.g., every 12 years) and he
OVERDIAGNOSIS AND OVERTREATMENT OF PROSTATE CANCER
undergoes prostate biopsy; if there is evidence of development of a more aggressive or larger tumor, treatment can
then be instituted. Thus, AS seeks to reserve treatment for
patients in whom there is evidence of a developing risk of an
aggressive tumor.
Many early series of WW attested to the fact that even
tumors detected with DRE (often large and extraprostatic
tumors), could be observed and, with extended follow-up,
most patients did not develop metastases nor did they die
as a result of prostate cancer.12 Since the advent of PSA
testing, there has been growing evidence that these lowergrade, lower-volume tumors can be watched carefully,
treated only in the event of evidence of a more aggressive or
larger tumor, thus avoiding treatment in many patients and
still showing low rates of metastasis and cancer death.
There are many examples of these series, including that of
Klotz and colleagues.13 In their series, the risk of prostate
cancer death at 5 years was 0.3% and at 10 years was 2.8%.
It was important to recognize in this series as well that 17%
of the patients undergoing AS actually had cancer with a
Gleason score of 7.
Sealing the evidence of overtreatment are the data from
the CAPSURE prostate cancer registry, a collection of a
range of urologic practices designed to evaluate practice
patterns related to prostate cancer. In one of their studies,
Cooperberg and colleagues found that 92% to 98% of patients who had the lowest tumor risk scores (and presumably were eligible for AS) were treated with radical surgery,
radiation, or hormone therapy.14 This stunning observation
that at least 92% of men who had a 10-year risk of prostate
cancer death of 2.8% received aggressive management,
strongly suggests that overtreatment is indeed occurring in
this disease. On the other hand, it is certain that some of
these men are making an individualized decision that this
risk is high enough to justify treatment.
What Is the Solution to Overdiagnosis and
Overtreatment in Prostate Cancer?
Prostate Cancer Prevention
Many institutions are increasingly focusing on methods to

reduce these problems in the U.S. population that cause
substantial resources to be expended unnecessarily and that
lead to substantial morbidity including sexual dysfunction,
urinary obstruction or incontinence, as well as GI complications. Certainly, one of the methods to reduce this problem is
through prostate cancer prevention. Currently available are
two agents that have been clearly demonstrated to reduce a
mans risk of detection of a low-grade prostate cancer:
dutasteride and finasteride, members of the class of five
alpha reductase inhibitors. These two agents have been
tested in phase III trials and have been found to reduce the
risk of prostate cancer by between 22% and 25%.15,16 Patients in both studies who received active treatment were,
however, more commonly found to have high-grade cancer.
Although the interpretation of some parties has been that
the agents induce high-grade cancer, substantial data show
that both agents facilitate the detection of cancer and
high-grade cancer. This occurs in, for example, the case of
finasteride, through an improved sensitivity of PSA for
overall cancer detection, improved sensitivity of DRE for
overall cancer detection, and likely through a reduction in
prostate volume leading to a greater likelihood of detection
Table 2. Potential Benefits and Harms of Prostate Cancer

Chemoprevention
Benefit
Harm
Reduction in morbidity of
treatment
Reduction in cost of
treatment
Unnecessary exposure to preventive drug of healthy

subjects who will never develop prostate cancer.
Side effects of chemoprevention agent. In the
example of finasteride and dutasteride, these
include gynecomastia, decreased libido, loss of
ejaculate, erectile dysfunction, and potential
increase in risk of high-grade cancer.
Increase in cost. (It is not known whether
chemoprevention would increase or decrease total
cost of the disease to society.)
Reduction in psychologic
burden of diagnosis
of high-grade cancer at the time of prostate biopsy by more

comprehensive sampling of a smaller prostate.17,18,19 Multiple studies have demonstrated that, when these biases
increasing cancer detection with finasteride are taken into
account, the overall impact on the entire range of high-grade
tumors (Gleason scores of 710) is a reduction in risk with
finasteride. Clearly, then, one option for reducing the risk of
overtreatment through a reduction in overdetection is
through chemoprevention. Table 2 lists the potential benefits and harms of chemoprevention of prostate cancer.
Screening and Biopsy of Men Who Are Most Likely to Harbor
Consequential Prostate Cancer
Through the use of risk assessment tools that provide

physicians with not only the risk of cancer detection but the
risk of detection of a high-grade cancer, it is possible to do a
better job of identifying the man who more likely has a
high-grade cancer. It is these men who have the greatest
potential for disease progression and death, best illustrated
by the work of Albertsen and colleagues, who examined the
outcomes of prostate cancer by age and grade.20 Lets examine these risks in two men using the Prostate Cancer
Prevention Trial Risk Calculator (www.prostate.cancer.
risk.calculator.com), which was developed based on PCPT
data and has been validated in a number of external
populations.21-23 Mr. Smith went to see his primary care
physician who felt a prostate nodule and sent him to his
urologist. Every single guideline in urology at this time
recommends biopsy for such a man, regardless of any other
risk factors. If this man is 55 and white, has a PSA of 0.5
ng/mL, has no family history of prostate cancer, and had a
biopsy of the nodule last year, his risk of low-grade cancer is
11.9% and his risk of high-grade cancer is 0.8%. Thus, there
is a 15-fold greater likelihood of finding an inconsequential
cancer in this man and his risk of having an aggressive
cancer is about 1 in 125. Because the detection of a lowgrade cancer probably has a net negative impact (the bulk of
these tumors are of low malignant potential and, even with
surveillance, there is a substantial degree of morbidity,
anxiety, and cost of this strategy), in this particular man,
the net potential benefit (probably 1 in 125) is likely to be far
outweighed by the net potential harm (about 1 in 8). Conversely, Mr. Jones, a healthy 73-year-old black man with no
other comorbidities who had a father with prostate cancer, is
found to have a new prostate nodule and whose PSA is 5.8
ng/mL, has a 56% risk of high-grade disease and a 15% risk
of low-grade disease. In this particular man, his risk is 1 in
2 that he may benefit from detection of an aggressive cancer,
whereas his risk of potential overdetection of a low grade
e37
IAN M. THOMPSON JR.

Table 3. Potential Risks and Benefits of Early Detection of
Prostate Cancer
Risks
Benefits
Detection and ultimate treatment of

cancer never destined to cause
harm to the patient
False-positive test: e.g., elevated PSA
in patient without cancer. This then
leads to unnecessary anxiety,
biopsy, and side effects of biopsy.
Detection and cure of cancer ultimately

destined to cause morbidity and/or
mortality
Identification of very low risk population
who may not require frequent testing.
An example could be man with very
low PSA whose risk of prostate cancer
is so low that PSA testing could be
every 2 to 5 years or longer.
Risk of biopsy: bleeding, 2%-4% risk

of sepsis
Risk of missing prostate cancer due
to sampling error
Risk of missing high-grade prostate
cancer (diagnosing only low-grade
cancer) and placing patient on
surveillance when more aggressive
treatment would be appropriate
Abbreviation: PSA, prostate-specific antigen.
cancer is about 1 in 7. It should be clear that the risk/benefit

ratios are utterly different in these two men. The potential
risks and benefits of early detection are displayed in Table 3,
and the potential side effects of the treatments themselves
are shown in Table 4.
The Future of Early Detection
Ideally, we are most likely to benefit the general population if we can 1) tell men at very low risk that they can skip
several years of screening and 2) consider biopsy only if a
high-grade cancer is present and, ideally, when the tumor
is still confined to the prostate and potentially curable.
Although PSA is directly related to risk of high-grade
disease, as can be seen in the examples above, when we use
this marker, we simply cannot tell in advance which man
will have which disease; as a result, we serendipitously
and probably unfortunatelyfind low-grade disease in the
process.
It will be the inclusion of diagnostic markers that are
highly related to high-grade cancer that will allow us to not
perform a biopsy on the man with low-grade disease while
Table 4. Potential Harms of Treatment

Active surveillance
Radical prostatectomy
Radiotherapy (seeds, beam,

combination)
Quarterly PSA testing: false positives, repeated visits

to lab.
Semi-annual rectal exams: cost, inconvenience,
embarrassment, discomfort.
Every 1- to 2-year prostate biopsy: pain,
bleeding, blood in semen, 2%-4% risk of sepsis.
Intraoperative bleeding or anesthetic complications,
infection, impotence, incontinence, anastomotic
stricture, rectal injury, infertility, loss of ejaculate,
disease recurrence, need for adjuvant or salvage
therapy.
Urinary dysfunction (retention, difficulty emptying,
urgency, and urge incontinence), impotence,
bowel dysfunction, radiation cystitis and proctitis,
blood in stool or urine, urethral stricture,
secondary malignancies of colon or bladder,
disease recurrence, salvage treatment.
recommending biopsy in the man with high-grade disease.

Several biomarkers appear to have this relationship including PCA3 and TMPRSS2:ERG (urine markers) as well as
[-2]proPSA (a serum marker). The Early Detection Research
Network (EDRN)Genitourinary Working Group has conducted several validation trials and will shortly complete the
evaluation of all three biomarkers. It is anticipated that the
EDRN will be able to incorporate these markers, either at
the time of initial screening or as a reflex test after an
intermediate risk is established by traditional PSA testing.
For example, if the man has a 20% risk of a low-grade cancer
and an 8% risk of high-grade cancer, a reflex test may be
used to exclude biopsy in the 20% with the low-grade
tumors. Ultimately, it will likely be a panel of markers and
what we have deemed biomeasures (e.g., age, race, ethnicity,
body mass index), all collected in a rational series of steps so
as to achieve the greatest detection of lethal cancers while
avoiding detection of inconsequential cancers, that will be
the future of prostate cancer detection. This detection process will then minimize both overdiagnosis and overtreatment. It is through the efforts of the biomarker science
community, as seen through the current risk assessment
tools and through upcoming incorporation of new markers,
that this vision will be achieved.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Ian M. Thompson Jr.*

REFERENCES
1. Delongchamps NB, Singh A, Haas GP. The role of prevalence in the
diagnosis of prostate cancer. Cancer Control. 2006;13:158-168.
2. Thompson, Ernst JJ, Spence CR, Gangai MP. Adenocarcinoma of the
prostate: Results of routine urological screening. J Urology. 1984;132:690692.
3. Thompson IM, Pauler DK, Goodman PJ, et al. Prevalence of prostate
cancer among men with a prostate-specific antigen level or 4.0 ng per
milliliter. N Engl J Med. 2004;350:2239-2246.
4. Roobol MJ, Schroder FH, Hugosson J, et al. Importance of prostate
volume in the European Randomised Study of Screening for Prostate Cancer
(ERSPC) risk calculators: results from the prostate biopsy collaborative
group. World J Urol. Epub 2011 Dec 28.
e38
5. Fang J, Metter EJ, Landis P, et al. PSA velocity for assessing prostate
cancer risk in men with PSA levels between 2.0 and 4.0 ng/ml. Urology.
2002;59:889-893.
6. Pinsky PF, Andriole GL, Kramer BS, et al. Prostate biopsy following a
positive screen in the prostate, lung, colorectal and ovarian cancer screening
trial. J Urol. 2005;173:746-750.
7. Hodge KK, McNeal JE, Terris MK, Stamey TA. Random systematic
versus directed ultrasound guided transrectal core biopsies of the prostate.
J Urol. 1989;142:71-74.
8. Levine MA, Ittman M, Melamed J, Lepor H. Two consecutive sets of
transrectal ultrasound guided sextant biopsies of the prostate for the detection of prostate cancer. J Urol. 1998;159:471-475.
OVERDIAGNOSIS AND OVERTREATMENT OF PROSTATE CANCER

9. Abdollah F, Scattoni V, Raber M, et al. The role of transrectal saturation
biopsy in tumour localization: pathological correlation after retropubic radical
prostatectomy and implication for focal ablative therapy. BJU Int. 2011;108:
366-371.
10. Siegel R, Naishadham D, Jemal A. Cancer Statistics, 2012. CA: Cancer
J Clin. 2012;62:10-29.
11. National Cancer Institute. SEER Cancer Statistics Review 1975-2008.
Lifetime Risk (Percent) of Dying from Cancer by Site and Race/Ethnicity:
Males, Total US, 2006-2008 (Table 1.18) and Females, Total US, 2006-2008
(Table 1.19). http://seer.cancer.gov/csr/1975_2008/results_merged/topic_life
time_risk_death.pdf. Accessed on December 8, 2011.
12. Whitmore WF Jr, Warner JA, Thompson IM Jr. Expectant management of localized prostatic cancer. Cancer. 1991;67:1091-1096.
13. Klotz L, Zhang L, Lam A, et al. Clinical results of long-term follow-up
of a large, active surveillance cohort with localized prostate cancer. J Clin
Oncol. 2010;28:126-131.
14. Cooperberg MR, Broering JM, Carroll PR. Time trends and local
variation in primary treatment of localized prostate cancer. J Clin Oncol.
2010;28:1117-1123.
15. Andriole GL, Bostwick DG, Brawley OW, et al. Effect of dutasteride on
the risk of prostate cancer. N Engl J Med. 2010;362:1192-1202.
16. Thompson IM, Goodman PJ, Tangen CM, et al. The influence of
finasteride on the development of prostate cancer. N Engl J Med. 2003;349:

215-224.
17. Thompson IM, Chi C, Ankerst DP, et al. Effect of finasteride on the
sensitivity of PSA for detecting prostate cancer. J Natl Cancer Inst. 2006;98:
1128-1133.
18. Thompson IM, Tangen CM, Goodman PJ, et al. Finasteride improves
the sensitivity of digital rectal examination for prostate cancer detection.
J Urol. 2007;177:1749-1752.
19. Lucia MS, Epstein JI, Goodman PJ, et al. Finasteride and high-grade
prostate cancer in the Prostate Cancer Prevention Trial. J Natl Cancer Inst.
2007;99:1375-83.
20. Lu-Yao GL, Albertsen PC, Moore DF et al. Outcomes of localized
prostate cancer following conservative management. JAMA. 2009;302:12021209.
21. Thompson IM, Ankerst DP, Chi C, et al. Assessing prostate cancer risk:
results from the Prostate Cancer Prevention Trial. J Natl Cancer Inst.
2006;98:529-534.
22. Parekh DJ, Ankerst DP, Higgins BA, et al. External validation of the
Prostate Cancer Prevention Trial risk calculator in a screened population.
Urology. 2006;68:1152-1155.
23. Eyre SJ, Ankerst DP, Wei JT, et al. Validation in a multiple urology
practice cohort of the Prostate Cancer Prevention Trial calculator for predicting prostate cancer detection. J Urol. 2009;182:2653-2658.
e39
Overdiagnosis and Overtreatment of

Breast Cancer
By Michael Alvarado, MD, Elissa Ozanne, PhD, and Laura Esserman, MD, MBA
Overview: Breast cancer is the most common cancer in

women. Through greater awareness, mammographic screening, and aggressive biopsy of calcifications, the proportion of
low-grade, early stage cancers and in situ lesions among
all breast cancers has risen substantially. The introduction of
molecular testing has increased the recognition of lower
risk subtypes, and less aggressive treatments are more commonly recommended for these subtypes. Mammographically
detected breast cancers are much more likely to have lowrisk biology than symptomatic tumors found between screenings (interval cancers) or that present as clinical masses.
Recognizing the lower risk associated with these lesions
and the ability to confirm the risk with molecular tests
should safely enable the use of less aggressive treatments.
Importantly, ductal carcinoma in situ (DCIS) lesions, or what
have been called stage I cancers, in and of themselves are
not life-threatening. In situ lesions have been treated in a

manner similar to that of invasive cancer, but there is little
evidence to support that this practice has improved mortality. It is also being recognized that DCIS lesions are
heterogeneous, and a substantial proportion of them may in
fact be precursors of more indolent invasive cancers. Increasing evidence suggests that these lesions are being overtreated. The introduction of molecular tests should be able
to help usher in a change in approach to these lesions.
Reclassifying these lesions as part of the spectrum of highrisk lesions enables the use of a prevention approach. Learning from the experience with active surveillance in prostate
cancer should empower the introduction of new approaches,
with a focus on preventing invasive cancer, especially given
that there are effective, United States Food and Drug Administration (FDA)-approved breast cancer preventive interventions.
What Is the Evidence that Breast Cancer May Be

Overdiagnosed or Overtreated?
REAST CANCER incidence has changed substantially

over the past 25 years. The risk has gone from one
in 12 to one in eight women being diagnosed in their
lifetimes. In contrast, the absolute risk of dying of breast
cancer has remained somewhat the same. As a proportion of
the women in whom breast cancer is diagnosed, mortality
has gone down considerably. But in absolute terms, the
number of women who die each year has changed less. In
the late 1980s, the number of women dying ranged from
45,000 50,000. Today, we expect about 42,000 women to die
annually.
The biology of the mammographically detected disease
appears to have changed, particularly among postmenopausal women. In a retrospective study in which nodenegative tumors diagnosed in Europe in the era before the
advent of mammography (1980 1991) were compared
with those detected during the era when screening was
prevalent (2004 2006), the chance of low-risk biology did
not change for women younger than age 40 but changed
substantially for those age 49 60.1 Using a 70-gene signature, 25% to 30% of women younger than 40 were diagnosed with low-risk disease, regardless of the time period in
which they were diagnosed. In contrast, among women
49 60, 40% were diagnosed with low-risk disease before
screening, but 58% were diagnosed as having low-risk disease during the later time period. Of the women with
mammographically detected tumors, 67% were classified as
having low risk by the 70-gene signature. A number of
factors are likely to contribute to the differences in the
cancers found today. Screening is clearly one factor that
enables the identification of small tumors; greater awareness is another. Hormone-replacement therapy also contributed to the greater chance of being diagnosed with breast
cancer, although with the publication of the Womens Health
Study in 2002, the widely publicized finding that combined
estrogen and progesterone increased breast cancer incidence
resulted in the plummeting of the number of prescriptions
for hormone-replacement therapy, accompanied by a rapid
decline in incidence.2
e40
From 1980 to the present, the incidence of breast cancer

has risen substantially.3 Much like prostate cancer, the
proportion of early-stage disease accounts for the bulk of the
increase in incidence, whereas the rates of locally advanced
cancers have not changed considerably. Screening is likely
to be a contributor to this phenomenon (Fig. 1).
Interestingly, we are also seeing a change in the likelihood
of local recurrence after breast cancer treatment. Among
women with node-negative tumors who were randomly assigned to radiation therapy or no radiation therapy after
quadrantectomy, the local recurrence rate was 10% for
women with mammographically detected tumors in the
group who did not receive radiation.4 Over the past decade,
the recurrence rate for postmenopausal women who receive
radiation therapy has dropped substantially, to the 1% to 2%
range.5,6 It is likely that low-risk tumors are contributing to
the drop in local recurrence rates reported in many trials. In
the recent update of the Canadian trial of women treated
with hormone therapy and randomly assigned to radiation
therapy or no radiation therapy, the recurrence risk was 2%
for women with luminal A tumors, even in the absence of
radiation.7
We are clearly detecting more low-risk disease. Some
argue that screening has led to the detection of tumors that
would never have come to clinical attention.8 Observational
evidence has shown that screening mammography is associated with increases in the incidence of breast cancer for
women of screening age but does not appear to contribute to
any decrease in the incidence of breast cancer among older
women after sustained screening for specific time periods.9
From the Department of Surgery, University of California, San Francisco.

Address reprint requests to Laura Esserman, MD, University of California, San Francisco, Department of Surgery, 1600 Divisadero St., 2nd Floor, Box 1710, San Francisco, CA
94115; email: laura.esserman@ucfsmedctr.org.
1092-9118/10/1-10
OVERDIAGNOSIS AND OVERTREATMENT OF BREAST CANCER
Fig 1. The SEER incidence of ductal carcinoma in situ (DCIS), stage I

(localized), stage II (regional), and stage IV (distant) breast cancer
over time. The red line at the bottom represents the incidence of DCIS
in 1980. The large red arrow represents the change in incidence in
DCIS by the year 2002. A concomitant drop in the incidence of
invasive cancer (top red arrow of equal magnitude) has not been
observed.
This evidence has even led to the hypothesis that some

screen-detected invasive breast cancers would likely regress
and be clinically unimportant in the future.8,10 Some note
that screening of all kinds has led to an epidemic of overdiagnosis.11 It is difficult to say with absolute certainty that
the increase in cancers detected would never become a
clinical problem, but there is ample evidence that the biology
of screen-detected tumors differs substantially from that of
interval cancers or nonscreening-detected cancers,12,13 although some say that the improvement in outcome is explained by lead-time bias.14
What is clear is that the lower risk biology associated with
these tumors should safely allow less treatment. In particular, there are groups of women for whom local therapy, in
particular, can be less, and for whom breast conservation
with adjuvant hormone therapy alone or with intraoperative
radiation therapy is associated with a low risk for local
recurrence and low mortality.5,7,15 Surveillance Epidemiology and End Results (SEER) data indicate that as many as
50% of all cancers occur in postmenopausal women and are
KEY POINTS
SEER data suggest that a meaningful amount of both

invasive and noninvasive breast cancers are likely to
be overdiagnosed and overtreated.
Further risk stratification is needed in the treatment
of breast cancer to avoid overtreatment.
Focusing on the prevention of invasive disease may
be one option to reduce the overtreatment of DCIS.
Learning from experiences in prostate cancer, active
surveillance should be considered for some cases of
DCIS.
Additional efforts towards the prevention of breast
cancer are greatly needed.
node-negative and low or intermediate grade.16 Given that

the inclusion of external-beam radiation therapy is considered a marker of quality care for breast cancer, there is
clearly evidence of overtreatment.
For many years, it has been recognized that there are
various pathologic characteristics and a spectrum of changes
in breast cell morphology, which range from hyperplasia to
atypia to in situ carcinoma.17,18 Many researchers believed
that each of these changes carried a distinct increase in risk
for future invasive breast cancer. The classic precursor
pathway, was accepted and therefore it was assumed that
all low-risk in situ cancers should be treated aggressively to
decrease the risk of a future invasive breast cancer. However, it has been shown that this is likely not the case, both
in epidemiology studies as well as in cohort studies of
specific high-risk lesions.
Considering that evidence suggests overdiagnosis in the
setting of invasive breast cancer, it is reasonable to consider
the possibility of overdiagnosis of DCIS.9,19 In the context of
DCIS, overdiagnosis may occur if DCIS lesions are precursors of invasive breast cancer that turn out to be indolent
disease or if the diagnosed DCIS is indolent itself without
the potential to progress to invasive cancer even in the
absence of treatment.
Similar to the situation with invasive breast cancer, SEER
data show a dramatic rise in the incidence of DCIS since the
introduction of screening mammography. However, there is
no corresponding drop in the rate of invasive breast cancer
that could account for successful detection and removal of
DCIS. In contrast, there is an overall increase in the incidence of invasive breast cancer since the introduction of
screening mammography. After more than 25 years of widespread efforts to increase rates of screening mammography,
the rate of invasive breast cancer incidence has continued to
increase despite widespread treatment of DCIS, except for a
recently observed leveling in incidence rates. It is possible
that this recent change in breast cancer incidence rates is
due to effective screening programs. However, a number of
epidemiologic studies have shown that this change is directly associated with the decreased use of hormone replacement therapy among postmenopausal women.20-22
These observations raise the question of whether DCIS is
an obligate precursor to invasive breast cancer and would
progress to invasive cancer if left untreated.23 A recent
modeling analysis explored the relationship between DCIS
progression rates and the assumed increase in the background rate of invasive breast cancer.24 The results of this
study demonstrated that different sets of assumptions related to DCIS and its progression rates could produce
similar projected trends. The study found that, if most DCIS
lesions were destined to progress to invasive breast cancer,
it would then be reasonable to assume that there is a
substantial increase in the baseline rate of invasive breast
cancer over what we have seen. In contrast, if there is only
a modest increase in the baseline rate of invasive breast
cancer after the introduction of screening mammography,
more in line with the observed trends before screening, it
then follows that most DCIS lesions would not have progressed to invasive breast cancer, even in the absence of
treatment.
Similarly, there is evidence in support of regression for
invasive breast cancer, which may be more relevant for
DCIS than for invasive cancer.8,11,19 This conclusion has
e41
ALVARADO, OZANNE, AND ESSERMAN
been reached by other modeling efforts designed to examine

trends in the incidence of breast cancer. With their model,
Mandelblatt and colleagues25 consistently underestimated
rates of DCIS as compared with rates drawn from SEER
data, suggesting that not all DCIS will become clinically
relevant. In a similar model, Fryback and colleagues26
identified a substantial proportion of tumors having limited
potential to become malignant in order to account for the
observed DCIS rate.
Although there are few data, evidence indicates that
low-grade DCIS treated by biopsy alone does not progress to
invasive breast cancer.27,28 Additionally, autopsy data suggest that untreated DCIS may not develop into invasive
breast cancer in a time period that will affect a womans
life.29 It appears feasible that there exists a reservoir of
DCIS in the population that is never diagnosed and never
attains clinical relevance, compelling us to ask if we have
overestimated the potential of DCIS to progress to invasive
breast cancer. If so, are we overdiagnosing and therefore
overtreating DCIS?
Researchers have tried to identify which in situ cancers
are the likely precursors of invasive disease and, of those,
which are of such low-risk that they might require minimal
local treatment (i.e., surgery alone or even active surveillance). Until recently, researchers have tried to identify
these low-risk groups by standard clinical-pathologic features.30 For example, high-grade palpable lesions have been
shown to carry a higher risk for recurrence in the first 5
years after diagnosis.31 A number of molecular markers
have been proposed to stratify risk but have not yet been
validated in large studies.32 The Van Nuys score, based on
combinations of patient age, tumor size, tumor grade, and
surgical margins, has been used to determine the value of
radiation therapy. A particular problem however, is the
current use of nuclear grade as overall grade, causing as
much as a 50% increase in the classification of high-grade
lesions.
Interestingly, a recent analysis of grade and molecular
profiling of in situ lesions suggested that low-grade lesions
were not the precursors of low-grade tumors, but high-grade
DCIS lesions were consistently predictive of high-grade
invasive tumors. However, when molecular subtyping was
performed, the vast majority (85%) of the lobular carcinoma
in situ (LCIS) and DCIS lesions were luminal A subtype.
The molecular subtype of the subsequent invasive tumor
was largely maintained: 85% for LCIS, and 69% for DCIS.33
Although the numbers in the study were small, the findings
suggest the possibility that grade is not a reliable indicator
of risk and that many DCIS lesions are in fact precursors of
more indolent cancers.
The unmet need for better risk stratification is now being
served with molecular profiling in the same way it has for
invasive cancers. Researchers at Genomic Health have now
applied their expertise in molecular profiling to DCIS. Recent data from their studies showed that they are able to
identify a low-risk category for which surgery alone would be
adequate local therapy. In fact, not only were they able to
identify risk for local recurrence but they were also able to
distinguish between total recurrences (in situ plus invasive)
and invasive cancer recurrence only. These data suggest
that a substantial number of low-risk lesions have, on
average, a 5% risk of an invasive local recurrence at 10 years
after lumpectomy alone. This risk is the equivalent of a Gail
e42
Risk Score of 2.5, indicating a 2.5% risk of invasive cancer at

5 years. This risk is also similar to the average risk of
invasive cancer for a woman in her mid-60s. Again, these
findings mean that DCIS, with excision alone, is appropriately categorized as a high-risk lesion such as atypia. These
new data are extremely important because they may finally
provide a more precise way of identifying women who are
being overtreated for a diagnosis that may have little bearing on their life for the next 20 30 years.
It is important to recognize that when screening was first
introduced, the goal was to identify invasive cancers. Over
the years, the focus has broadened to include calcifications
and the identification of DCIS. In the United States, the
biopsy rates are higher than they are in Europe,34 and the
cancer-to-biopsy ratios are in the range of 25% and 50%,
respectively. The proportion of DCIS detected is higher, and
more cases of detected DCIS are lower grade. These findings
are important to keep in mind when evaluating results of
different trials. For example, in the [United Kingdom New
Zealand trial], the majority of DCIS detected was high
grade, which may explain why tamoxifen was not found to be
effective in reducing the risk of ipsilateral invasive cancer.
The introduction of more standard molecular tools may
prove important in comparing DCIS types and in allowing a
different approach to DCIS in the future.
Despite the recognition that some breast cancers we
detect are indolent in nature, there have been almost no
trials of active surveillance, as there have been in prostate
cancer. Even DCIS, which is thought to be a precursor of
invasive cancer, has rarely been approached with active
surveillance. It is hoped that the availability of tools to
profile DCIS lesions in a standard manner will usher in a
more progressive approach to DCIS, not unlike what has
been adopted in the treatment of prostate cancer, especially
given the availability of preventive interventions.
What Is the Solution?
Change the Approach to DCIS from Cancer Treatment to
Prevention of Invasive Cancer
The evidence for overtreatment of DCIS is extremely

compelling, given the associated risk. Classic treatment
strategies have almost always included either breast conservation with whole breast radiation or mastectomy. It is now
becoming apparent that this type of precursor lesion is
being overtreated. With continued evidence that we are not
only overdetecting but also overtreating breast cancer in
this country and abroad, researchers need to think seriously
about developing protocols for women at the lowest risk.
With regard specifically to DCIS, it would not be unreasonable to develop algorithms to identify women who could have
active surveillance after carcinoma in situ is found on
evaluation of a biopsy specimen.35 Many groups have discussed ways in which to stratify risk, and with the availability of molecular markers, it should be an immediate
action item to finally address the overtreatment of earlystage breast cancer. Previous research attempted to identify
risk groups that were typically based on classic, clinicalpathologic features such as patient age, tumor size, and
tumor grade. Numerous data with molecular profiling have
indicated that standard clinical-pathologic features are a
poor way of truly risk stratifying newly diagnosed breast
cancer. In fact, grade appears to be nonreproducible across

Table 1. Short- and Long-term Risk of Breast Cancer by Lesion/Risk Factor. 36
Abbreviations: LCIS, lobular carcinoma in situ; DCIS, ductal carcinoma in situ.
multiple laboratories and different pathologists. Therefore,

one would expect that future treatment decisions are based
on molecular markers and not subjective data such as grade,
age, and tumor size.
Focus on Prevention
The Athena Breast Health Network is an innovative

collaboration across the five University of California medical
centers and includes the UCSF Institute for Health Policy
Studies and the Graduate School of Public Health at Berkeley. Athena is creating a 21st century knowledge economy
that will integrate clinical care and research to drive innovation in prevention, screening, treatment, and management of breast cancer and, at the same time, revolutionize
the delivery of care. By working together as a community,
the University of California medical centers, their affiliates,
and primary care and specialty physicians will work to
change the options for patients today and create a better
future for all women at increased risk for breast cancer.
Athena will implement a comprehensive Web-based informatics strategy to modernize the way in which clinical
information is collected, tracked, and integrated with research. Athena will also develop and implement standard
Web-based tools that will be integrated with the existing
clinical information infrastructure to optimize the capture of
structured clinical information at the point of care from
clinicians and patients. The development of an automated
risk assessment tool integrated into Web-based decision
support (BreastHealthDecisions.org, or BHD) is the culmination of 10 years of collaborative work by faculty involved
in Athena. BHD was designed to widely disseminate
evidence-based personalized options for breast cancer risk
reduction and to increase awareness and use among the
primary care community and hard-to-reach populations in a

cost-efficient manner. This tool, which includes all of the
validated breast cancer risk models, was built on an awardwinning decision-making framework15,16 and is a decision
aid that our team has developed and tested with patients
and experts in breast cancer risk and prevention over the
past 5 years.17 The tool is designed for use by breast
specialists, gynecologists, and primary care physicians, and
incorporates the standard and improved models for breast
cancer risk assessment that include breast density, hereditary risk, and atypia.18
Opportunities to Change Now
Given that many cases of DCIS are associated with a 5%

risk of invasive cancer after 10 years (the equivalent of a
Gail Risk score of 2.5 or the average risk of a woman in her
mid-60s) as shown in Table 1, these lesions too should likely
be considered to be markers of elevated risk and should
prompt thoughts of chemoprevention strategies. Coopey and
colleagues37 recently reported the outcome for nearly 3000
patients with atypical breast lesions, ranging from atypical
lobular hyperplasia to borderline DCIS lesions. The data
demonstrated that these lesions conferred a risk of cancer in
both breasts and that for this group of patients, chemoprevention with tamoxifen or raloxifene reduced the risk by
about two-thirds, on average. This finding is consistent with
data from the National Surgical Adjuvant Breast and Bowel
Project (NSABP) P-01 trial showing that women with atypia
had an 85% reduction in risk with chemoprevention.38
Ironically, the only lesions we approach with surveillance
are those with greater risk than that posed by many DCIS
cases. Unlike the case in the prostate community, we in the
e43
ALVARADO, OZANNE, AND ESSERMAN
breast cancer community have FDA-approved risk-reducing

agents that can be offered.
It is possible that the lesions associated with the equivalent of a low DCIS score according to molecular profiling
with Oncotype DX (Genomic Health, Redwood City, CA)
could be treated with a chemoprevention approach rather
than surgery. The ALLIANCE is opening a trial for nonhighgrade DCIS in postmenopausal women, with initial treatment of an aromatase inhibitor for 6 months. Magnetic
resonance imaging (MRI) will be done before therapy is
initiated, to rule out the presence of invasive cancer and to
document the extent of disease by MRI-measured volume.
MRI will be repeated after 3 months of therapy, and if there
is no evidence of progression, therapy will continue for an
additional 3 months. The endpoint is the change in MRImeasured volume. For high-grade DCIS, we should be thinking about screening compounds for their biologic effect, with
the goal of generating neoadjuvant trials for the compounds
that have the greatest effect on mitotic activity or other
surrogate markers. Targets should not only be the tumor
itself but changes in the stroma (e.g., density), as these may
be measures of impact.
Conclusion: Future of Early Detection
Our goal should be to learn who is at risk for what type of

cancer and to use the available predictors of risk to determine the type and frequency of screening. In addition, risk
assessment, using existing and emerging models, should be
part of screening. Our efforts should be targeted to preventing breast cancer in the highest risk groups. Not only should
we work on making interventions available where appropriate and informing women of their options (eg, with tools
such as BHD) but we should also be using emerging markers
of risk, such as breast density, as a potential surrogate for
effectiveness of prevention agents, as has been shown for
tamoxifen.39 Other markers should be developed to help
guide the testing of new interventions, especially for women
at risk for aggressive tumor biology. In situ lesions should be
used as markers of risk and tools to evaluate the effect of
therapies. Breast cancer is clearly a heterogeneous disease.

Our treatment approaches need to reflect the differences in
biology even more than they do today, and in the future, we
will need to tailor our screening and prevention recommendations as well. We can improve the use of early detection
efforts in the United States by following the U.S. Preventive
Services Task Force guidelines. Screening every other year
is not associated with a substantial increase in the number
of locally advanced cancers, reduces the false-positive results of biopsy evaluations by about a third,40 and could save
billions of dollars annually.41 Efforts are already underway
to develop risk-based screening guidelines.42 Not only would
risk-based screening be more effective and a much better
way to apply resources it may also be an important way to
move past the controversy about annual or biennial screening and encourage women to accept that there may be a
more appropriate way to apply screening.
We can learn from the prostate cancer experience with
active surveillance to recognize the safety of following more
indolent disease and apply that to the setting of DCIS, which
is not invasive cancer. The first step is to recognize that most
of the DCIS lesions detected are not likely to progress. Even
if they do, the period of risk spans several years, and many
lesions are likely to be precursors of relatively indolent
disease. More importantly, the subsequent invasive events
appear to be largely preventable. We should be thinking
more like how the cardiology community in how they view
the opportunity to reduce the incidence of cardiac and stroke
events. We have the markers of risk that can be used; these
markers should include not just atypia but lower risk DCIS.
We now have preventive agents that we can use. We need to
refine our ability to determine whether specific patients are
benefiting from chemoprevention interventions, but we have
much that we can do today that is likely better than what we
have been doing. The emergence of molecular tools to more
characterize DCIS in a more standard manner should be
used as an opportunity to reframe options and work together
to reduce the morbidity associated with early detection.
Author
Michael Alvarado*
Elissa Ozanne*
Laura Esserman
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Agendia
REFERENCES
1. Esserman LJ, Shieh Y, Rutgers EJ, et al. Impact of mammographic
screening on the detection of good and poor prognosis breast cancers. Breast
2. Chlebowski RT, Kuller LH, Prentice RL, et al. Breast cancer after use of
estrogen plus progestin in postmenopausal women. N Engl J Med. 2009;360:
573-587.
3. Lin C, Moore D, DeMichele A, et al. Detection of locally advanced breast
cancers in the I-SPY TRIAL (CALGB 150007/150012, ACRIN 6657) in the
interval between routine screening. J Clin Oncol. 2009;27:15s (suppl; abstr
1503).
4. Malmstrom P, Holmberg L, Anderson H, et al. Breast conservation
surgery, with and without radiotherapy, in women with lymph node-negative
breast cancer: a randomised clinical trial in a population with access to public
mammography screening. Eur J Cancer. 2003;39:1690-1697.
5. Vaidya JS, Joseph DJ, Tobias JS, et al. Targeted intraoperative radio-
e44
therapy versus whole breast radiotherapy for breast cancer (TARGIT-A trial):
an international, prospective, randomised, non-inferiority phase 3 trial.
Lancet. 2010;376:91-102.
6. START Trialists Group, Bentzen SM, Agrawal RK, et al. The UK
Standardisation of Breast Radiotherapy (START) Trial A of radiotherapy
hypofractionation for treatment of early breast cancer: a randomised trial.
Lancet. 2008;9:331-341.
7. Fyles A, McCready D, Pintilie M, et al. Luminal A subtype predicts
radiation response in patients with T1N0 breast cancer enrolled in a randomized trial of tamoxifen with or without breast radiation. Paper presented at:
San Antonio Breast Cancer Symposium; December 2011; San Antonio, TX.
8. Zahl PH, Maehlen J, Welch HG. The natural history of invasive breast
cancers detected by screening mammography. Arch Intern Med. 2008;168:
2311-2316.
9. Jrgensen, KJ, Gtzsche PC. Overdiagnosis in publicly organised mam-

mography screening programmes: systematic review of incidence trends.
BMJ. 2009;339:b2587.
10. Welch HG, Frankel BA. Likelihood that a woman with screen-detected
breast cancer has had her life saved by that screening. Arch Intern Med.
2011;171:2043-2046.
11. Welch HG, Black WC. Overdiagnosis in cancer. J Natl Cancer Inst.
2010;102:605-613.
12. Mook S, Van t Veer LJ, Rutgers EJ, et al. Independent prognostic value
of screen detection in invasive breast cancer. J Natl Cancer Inst. 2011;103:
585-597.
13. Lin C, Buxton MB, Moore D, et al. Locally advanced breast cancers are
more likely to present as interval cancers: results from the I-SPY 1 TRIAL
(CALGB 150007/150012, ACRIN 6657, InterSPORE Trial). Breast Cancer Res
Treat. Epub 2011 Jul 28.
14. Allgood PC, Duffy SW, Kearins O, et al. Explaining the difference in
prognosis between screen-detected and symptomatic breast cancers. Br J
Cancer. 2011;104:1680-1685.
16. Esserman LJ, Mohan AJ, Park C, et al. Projecting the impact of
adopting trial results. Poster presented at: San Antonio Breast Care Symposium; December 2011; San Antonio, TX.
17. Allred DC, Mohsin SK, Fuqua SA. Histological and biological evolution
of human premalignant breast disease. Endocr Relat Cancer. 2001;8:47-61.
18. Dupont WD, Page DL. Risk factors for breast cancer in women with
proliferative breast disease. N Engl J Med. 1985;312:146-151.
19. Morrell S, Barratt A, Irwig L, et al. Estimates of overdiagnosis of
invasive breast cancer associated with screening mammography. Cancer
Causes Control. 2010;21:275-282.
20. Ravdin PM, Cronin KA, Howlader N, et al. The decrease in breastcancer incidence in 2003 in the United States. N Engl J Med. 2007;356:16701674.
21. Kerlikowske K, Miglioretti DL, Buist DS, et al. Declines in invasive
breast cancer and use of postmenopausal hormone therapy in a screening
mammography population. J Natl Cancer Inst. 2007;99:1335-1339.
22. Glass AG, Lacey JV Jr, Carreon JD, et al. Breast cancer incidence,
1980-2006: combined roles of menopausal hormone therapy, screening mammography, and estrogen receptor status. J Natl Cancer Inst. 2007;99:11521161.
23. Welch HG. Should I Be Tested for Cancer? Maybe Not and Heres Why.
Berkeley: University of California Press; 2004.
24. Ozanne EM, Shieh Y, Barnes J, et al. Characterizing the impact of 25
years of DCIS treatment. Breast Cancer Res Treat. 2011;129:165-173.
25. Mandelblatt J, Schechter CB, Lawrence W, et al. The SPECTRUM
population model of the impact of screening and treatment on U.S. breast
cancer trends from 1975 to 2000: principles and practice of the model
methods. J Natl Cancer Inst Monogr. 2006;(36):47-55.
26. Fryback DG, Stout NK, Rosenberg MA, et al. The Wisconsin Breast
Cancer Epidemiology Simulation Model. J Natl Cancer Inst Monogr. 2006;
(36):37-47.
27. Sanders ME, Schuyler PA, Dupont WD, et al. The natural history of
low-grade ductal carcinoma in situ of the breast in women treated by biopsy
only revealed over 30 years of long-term follow-up. Cancer. 2005;103:24812484.
28. Lagios MD, Margolin FR, Westdahl PR, et al. Mammographically
detected duct carcinoma in situ. Frequency of local recurrence following
tylectomy and prognostic effect of nuclear grade on local recurrence. Cancer.
1989;63:618-624.
29. Welch HG, Black WC. Using autopsy series to estimate the disease
reservoir for ductal carcinoma in situ of the breast: how much more breast
cancer can we find? Ann Intern Med. 1997;127:1023-1028.
30. Silverstein MJ, Lagios MD. Choosing treatment for patients with
ductal carcinoma in situ: fine tuning the University of Southern California/
Van Nuys Prognostic Index. J Natl Cancer Inst Monogr. 2010; 2010(41):193196.
31. Kerlikowske K, Molinaro A, Cha I, et al. Characteristics associated
with recurrence among women with ductal carcinoma in situ treated by
lumpectomy. J Natl Cancer Inst. 2003;95:1692-1702.
32. Kerlikowske K, Molinaro AM, Gauthier ML, et al. Biomarker expression and risk of subsequent tumors after initial ductal carcinoma in situ
diagnosis. J Natl Cancer Inst. 2010;102:627-637.
33. King TA, Sakr RA, Muhsen S, et al. Is there a low-grade precursor
pathway in breast cancer? Ann Surg Oncol. Epub 2011 Sep 21.
34. Esserman L, Cowley H, Eberle C, et al. Improving the accuracy of
mammography: volume and outcome relationships. J Natl Cancer Inst.
2002;94:369-375.
35. Meyerson AF, Lessing JN, Itakura K, et al. Outcome of long term active
surveillance for estrogen receptor-positive ductal carcinoma in situ. Breast.
2011;20:529-533.
36. Esserman L, Sepucha K, Ozanne EM, Hwang ES. Applying the neoadjuvant paradigm to ductal carcinoma in situ. Annals of Surgical Oncology.
2004;11(1 Suppl):28S-36S.
37. Coopey SB, Mazzola E, Buckley JM, et al. Clarifying the risk of breast
cancer in women with atypical breast lesions. Paper presented at: San
Antonio Breast Cancer Symposium; December 2011; San Antonio, TX.
38. Fisher B, Costantino JP, Wickerham DL, et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and
Bowel Project P-1 Study. J Natl Cancer Inst. 1998;90:1371-1388.
39. Cuzick J, Warwick J, Pinney E, et al. Tamoxifen-induced reduction in
mammographic density and breast cancer risk reduction: a nested casecontrol study. J Natl Cancer Inst. 2011;103:744-752.
40. Hubbard RA, Kerilkowske K, Flowers CI, et al. Cumulative probability
of false-positive recall or biopsy recommendation after 10 years of screening
mammography: a cohort study. Ann Intern Med. 2011;155:481-492.
41. Burnside E, Belkora J, Esserman L. The impact of alternative practices
on the cost and quality of mammographic screening in the United States. Clin
Breast Cancer. 2001;2:145-152.
42. Schousboe JT, Kerilkowske K, Loh A, et al. Personalizing mammography by breast density and other risk factors for breast cancer: analysis of
health benefits and cost-effectiveness. Ann Intern Med. 2011;155:10-20.
e45
COSTS OF CANCER CARE: AFFORDABILITY,

ACCESS, AND POLICY
CHAIR
Thomas J. Smith, MD
Baltimore, MD
SPEAKERS
Richard Sullivan, PhD, MBBS
Kings Health Partners Integrated Cancer Centre, Guys Hospital
Sean R. Tunis, MD, MSc
Center for Medical Technology Policy
Baltimore, MD
Reducing the Cost of Cancer Care: How to

Bend the Curve Downward
By Thomas J. Smith, MD, Bruce E. Hillner, MD, and Ronan J. Kelly, MD, MBA
Overview: Health care and cancer care costs are rising

unsustainably such that insurance costs have doubled in 10
years. Oncologists find themselves both victims of high costs
and the cause of high-cost care by what we do and what we
do not do. We previously outlined five ways that oncologists
could personally bend the cost curve downward and five
societal attitudes that would require change to lower costs.
Here, we present some practical ways to reduce costs while
HE RISING cost of cancer care is unsustainable. Medical care costs more in the United States than anywhere
else in the world, as shown in Fig. 1. The impact of this
rising cost is felt throughout the health care system, from
patients and families to insurers and the government. The
high cost of medical care is a concern for manufacturers who
must build that cost into their goods. We illustrate some of
the impact in human terms in Sidebar 1.
Cancer doctors often think we are the victims of the rising
cost of cancer care, when we are both victims and causal
agents. In fact, we are responsible for what we do and what
we do not do, and the consequences of that action or inaction.
We order the tests and prescribe the chemotherapy and
supportive care drugs, and make the decisions to continue
chemotherapy, involve palliative care or hospice early, have
discussions about goals of care and advance directives or
not. In our review4 we explored five changes in oncologist
behavior that would bend the cost curve (Sidebar 2), and
five attitudes that needed to change for this to happen
(Sidebar 3).
Here, we want to think about recognizing the coming
problems, and propose some concrete solutions. First, the
world is changing from fee-for-service to bundled payments
to take away the incentive to administer more profitable
chemotherapy. We have never maintained that oncologists
administer chemotherapy just to make money. Interestingly, patients in countries like Sweden and Portugal, where
oncologists do not make money giving chemotherapy, still
receive chemotherapy near the end of life.7 However, there is
no way to generate an income of nearly $400,000/year from
cognitive services alone, and there is an inherent conflict of
interest when we must choose between chemotherapy that
gives us little profit compared with major profit. Second,
more people will be uninsured or have less coverage with
higher copays and deductibles, and shift between plans as
insurance becomes more portable. Third, value is missing
in some of our spending, if we define value as quality/cost.
Finally, all of us have to realize that change is risky and
disruptive, with major implications for our salaries and
ability to support an enterprise.
We propose practical ways to improve health, quality of
care, and value. First, we propose redesign of the National
Comprehensive Cancer Network (NCCN) and other pathways to incorporate cost and value. Second, we propose an
audit of current patterns of care for under- and overuse. We
can help improve electronic medical record (EMR) prompts
that promote best practices. Finally, we want to redesign
NCCN and other pathways to incorporate palliative care
e46
maintaining or improving quality, including: 1) evidence-based

surveillance after curative therapy; 2) reduced use of white
cell stimulating factors (filgrastim and pegfilgrastim); 3) better
integration of palliative care into usual oncology care; and
4) use of evidence-based, cost-conscious clinical pathways
that allow appropriate care and lead to equal or better
outcomes at one-third lower cost.
early and concurrently to provide the best care at a cost we

can afford.
This is part of a major national effort to restrain costs
while maintaining quality. The American Board of Internal
Medicine is promoting Choose Wisely to have each specialty find at least five ways to reduce costs8 has been endorsed by at least eight major organizations (http://choosing
wisely.org/wp-content/uploads/2011/12/about_choosingwisely.
pdf). Accountable care organizations and medical homes
are part of all new health care legislation and attempts to
restrain costs while maintaining quality.9 The American
College of Physicians has recognized the inherent tension
between doing all that one can for an individual patient
compared with careful use of societal resources: Physicians
have a responsibility to practice effective and efficient health
care and to use health care resources responsibly. Parsimonious care that utilizes the most efficient means to
effectively diagnose a condition and treat a patient respects
the need to use resources wisely and to help ensure that
resources are equitably available. Physicians considered
judgments should reflect the best available evidence including data on the cost-effectiveness of different clinical approaches.10
Some Specific Examples of Actions Under
Oncologist Control
Target Surveillance Tests or Imaging to Those
Situations Where a Benefit Has Been Shown
This one should be easy. There are no data suggesting

better medical outcomes for patients with breast or ovary
cancer treated with curative intent who are followed by
other than routine exams. For breast cancer, two large
European studies showed that routine follow-up compared
with more intense schedules of scans gave equal outcomes
and equal quality of life. There are no data that suggest
early identification of metastatic breast cancer leads to
better outcomes. In one small study, carcinoembryonic antigen (CEA) could detect breast cancer recurrence, but early
From the Palliative Medicine Program, Sidney Kimmel Comprehensive Cancer Center,
The Johns Hopkins University, Baltimore, MD; Massey Cancer Center, Virginia Commonwealth University, Richmond, VA.
Address reprint requests to Thomas J. Smith, MD, Director of Palliative Medicine for
Johns Hopkins Medicine, Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins,
600 N. Wolfe Street, Blaylock 369, Baltimore, MD 21287-0005; email: tsmit136@jhmi.edu.
1092-9118/10/110
REDUCING THE COST OF CANCER CARE
Sidebar 1. Some Facts about the Cost of Cancer

Care in the United States
Fig. 1. Expenditure per person on health care, in 2009 U.S. dollars. From OECD Health Data 2011, http://stats.oecd.org/Index.
aspx?DataSetCodeSHA.1
treatment with tamoxifen in estrogen receptorpositive

patients made absolutely no difference compared with treatment when patients were symptomatic.11 Early identification of ovarian cancer recurrence with cancer antigen (CA)125 testing did not lead to better medical outcomes in a
large, well-designed randomized clinical trial.12 There are
no data suggesting benefit in lung or prostate cancer for
early detection of recurrent disease, and screening is not
recommended. In fact, the only common disease in which one
can make a compelling argument is in colorectal cancer CEA
testing.13
KEY POINTS
Cancer costs are rising at an unsustainable rate by

any measure.
Oncologists are directly responsible for some of the
increase in costs, by what we do and what we do not
do, and are one key to bending the cost curve downward.
We can reduce the cost of care by restricting surveillance of patients who have received treatment to
those tests that have been shown to improve outcomes; this will save several billions of dollars in
breast cancer alone.
We can save several billions of dollars a year in
patients with metastatic solid tumors by reducing
the use of white cell growth factors to the indications
approved by ASCO, European Organisation for Research and Treatment of Cancer, and National Comprehensive Cancer Network, and with dose reduction
instead of colony-stimulating factor use where appropriate.
We can improve care near the end of life by involving
palliative care and hospice earlier3 to 6 months
before deathand, for most diseases by not administering chemotherapy to patients with poor performance status, or after progression despite three lines
of chemotherapy.
Medical care costs more in the United States than

any other country often twice as muchwith no
better survival. We spend $8,000/person/year versus Canadas $4500.1
Nearly one million14 families suffered medical
bankruptcy in 2010. (Commonwealth Fund) Over
half of those bankruptcies happened to people
with insurance, and most families are middle
class.2 Approximately 15% to 20% of insurance
payments are for cancer, so we caused several
them. Eight percent of families with a member
with lung cancer are bankrupt because of the
disease.15
The cost of insurance for a family of four has risen
from $6,000 to more than $15,000/year in the last
10 years.16 No wonder more people are underinsured up to nearly 60 millionand others are
underinsured.
Drug costs are rising at a rate of 3%, but represent only 13% of the total cost of care.3
Much of the rest is under our control including
imaging, chemotherapy choices, integration of palliative care, use of hospice, and avoiding chemotherapy and hospitalization near the end of life.4
Oncologists salaries (median $381,992) are
among the highest for medical specialists compared with primary care doctors ($202,392) and
rose 4% in 2010. This disparity is unique in
developed countries.5
At least half of oncologists income, 50% to 80%,
comes from drug sales which represent an inherent potential conflict of interest, and is unique in
U.S. medicine.5
Hospitals have an inherent interest in keeping
beds full and profit margins high on drugs and
services.
Approximately 25% of all Medicare funds are
spent in the last year of life, and more than 9%
(upwards of $50 billion) in the last month of life,6
with similar patterns in commercial insurance.
The reason for the lack of effect on early recurrence has

been stated eloquently by Dr. Tito Fojo: Until we have
treatments that kill essentially all the recurrent cancer,
rather than just some of it, we will not have a chance of cure
(written communication, December 2011).
The reasons patients and maybe doctors want these tests
are complicated. Some patients want to be doing something even if something is not of proven benefit. Explaining
the reasons behind not testing is more difficult and time
consuming than ordering the test. Using breast cancer as an
example, the ASCO printed guidelines that state no testing can be reviewed in less than 10 minutes17 and patients
can understand the rationale for not testing,18 but it takes a
conversation that may be difficult for some oncologists.
ASCO is taking the lead in advocating for best practices
in cancer surveillance. Adherence to the ASCO and NCCN
e47
SMITH, HILLNER, AND KELLY
Sidebar 2. Five Changes in Oncologist Behavior

that Will Bend the Cancer Cost Curve
Sidebar 3. Five Attitudes that Must Change for

Better Cost Consciousness
1. Target surveillance tests or imaging to those

situations in which a benefit has been shown.
2. For most solid tumors limit second- and thirdline treatment for metastatic cancer to sequential monotherapies. Based on the evidence,
single drugs are indicated in breast, lung and
prostate but not colorectal cancer.
3. For patients with cancer that has progressed
despite treatment, limit future chemotherapy to
patients with good performance status.
4. Replace the routine use of white-cellstimulating factors in the treatment of metastatic solid
cancers with chemotherapy dose reduction.
5. For patients not experiencing response to three
consecutive regimens, limit further chemotherapy to patients entering clinical trials.
1. Recognition that oncologists drive the costs of

care by what we do and do not do.
2. Both doctors and patients need more realistic
expectations.
3. Realignment of compensation to rebalance cognitive services with chemotherapy use.
4. Better integration of end-of-life non chemotherapy-oriented palliative care.
5. Acceptance of the necessity for cost-effectiveness
analysis and the need for some limits on care.
breast cancer guidelines with minimal testing will likely be

one of ASCOs five topics in the American Board of Internal
Medicine Choose Wisely campaign.
Replace the Routine Use of White-Cell Stimulating Factors in the
Treatment of Metastatic Solid Cancers with Chemotherapy
Dose Reduction
The use of colony stimulating factors (CSFs) is one area

in which the United States and other countries have very
disparate patterns of care but no difference in mortality.19
The United States represents 3% of the world population but
purchases 75% of the granulocyte (G-) CSF and pegylated
G-CSF produced by Amgen (Thousand Oaks, CA). We have
recently explored some of the reasons why this disparity
occurs, including dislike of febrile neutropenia, marketing,
fear of malpractice, and the profits made often several
hundred dollars for each injection. The only proven recommended uses are in dose-dense treatment for estrogenpositive breast cancer and for lymphoma treatment when
the risk of febrile neutropenia is high.20 A recent study
showed that for patients older than age 65, there was no
proven clinical benefit to the use of primary prophylactic
CSFs and that the cost-effectiveness ratio was more than
$900,000 per quality-adjusted life year saved,21 which may
make us question current practice or demand lower CSF
prices.
We continue to recommend that the United States follow
the ASCO, NCCN, and European Organisation for Research
and Treatment of Cancer (EORTC) guidelines and use CSFs
for curative care. However, for the treatment of metastatic
solid tumors in which no clinical benefit has been shown, we
should follow ASCO and NCCN guidelines and use less toxic
regimens, not use primary prophylaxis even in diseases such
as small cell lung cancer (in which it is not recommended by
NCCN guidelines), and if needed reduce the doses per the
original protocol. We simply cannot afford $2200 4800 each
cycle for supportive care that does not improve survival
without the consequences in Sidebar 1.
e48
Changing Attitudes
There are oncologist attitudes that also drive the cost of

care. We highlight two: recognition that oncologists drive
the costs of care by what we do and not do, and better
integration of end-of-life, non chemotherapy-oriented palliative care.
For instance, having a discussion about impending death
of a patient improves the pattern of care. Data show clearly
that although we discuss curability or not,22 only 37% of the
time do we discuss the fact of an impending death.23 Data
also show clearly that having this discussion was associated
with no more depression or anxiety in the patient; less
depression in the caregiver; far less end-of-life intubation,
resuscitation, and intensive care unit use; and longer hospice use. Just having that discussion lead to better medical
care, better outcomes for the family, and $1,000 less spent in
the last week of life.24
Sidebar 4. Integrating Best Practices: Use the

Medical Record to Help Your Practice
1. List the treatments used, so that we can know

when it is time to switch to non-chemotherapy
based care. Build in some prompts to trigger
consultation.
2. Put a prompt to remind you to discuss
Goals of care
Prognosis
Advance medical directives
Code status
Hospice referral
3. Have all patients with incurable cancer receive
a hospice information visit when they have 3 to 6
months to live, to make the transition smoother.
4. If you are not comfortable discussing these difficult issues, appoint someone in your practice
(an advance practice nurse or social worker) or
get training.
5. After a patient is referred to hospice, pencil in
appointments to call them every week, just to
check on them and make sure they do not feel
abandoned.
This recognition of when it is time to switch treatments

away from chemotherapy is essential to best quality of care,
but we are not good at it. At the University of Iowa, 60% of
the patients who died in the hospital were eligible for
hospice on their penultimate admission, yet this was recognized only 14% of the cases and few were enrolled.25 Sixty
percent of oncologists prefer to wait until there are no more
chemotherapy options left to discuss hospice, advance medical directives, and code status.22 This may explain why
no doctor has mentioned hospice to half of all patients with
lung cancer when they have 8 weeks left to live26 and why
one-third of all patients with cancer enter hospice with less
than a week to live, according to the National Hospice and
Palliative Care Organization.27 Data are also clear that
hospice does not worsen survival of patients with nonsmall
cell lung cancer,28,29 chemotherapy in the last 2 weeks of life
does not improve survival,28 and fourth-line chemotherapy
has not worked with a response rate of 0%.26 Despite this,
a large percentage of people with incurable solid tumors
receive chemotherapy within 2 weeks of their death, including at academic centers.31
We can improve care by incorporating palliative care
early, as recommended by ASCO in the Provisional Clinical
Opinion.32 Studies clearly show no harm, substantial benefit
in most patient reported outcomes, and even better survival
in lung cancer. The strongest correlations with increased
survival in that study were not more chemotherapy but less
intravenous chemotherapy in the last 60 days of life, and
better understanding of prognosis and goals of treatment.33
Of note, the usual oncology arm used hospice for only 4 days
if they used it at all.
How can we improve care? We have put some simple
recommendations in Sidebar 4.
Bringing it All Together to Improve Care and Reduce
Cost at the Same Time
Nonsmall cell lung cancer is a good example of how we

can modify existing guidelines to improve care. There are
Fig. 2. Changing our practice to incorporate evidence-based

guidelines and best practices about palliative care, using lung cancer
as an example. These suggestions are superimposed on the survival
curve of lung cancer patients treated on-pathway or off-pathway;
note that the survival curves are identical but the pathway patients
care cost 33% less. See Neubauer M et al.35
Abbreviations: ADs, advance directives; DPMA, durable power of
medical attorney; PC, palliative care.
several important points that would improve care; we have

illustrated them on the survival curve from U.S. Oncology
patients on and off their clinical pathways in Figure 2. Note
that the survival curves are superimposable, but the cost
was 35% less with the use of the pathways.34,35 The figure
shows how this might work.
In the first phase, the pathway restricts choice for patients
with nonadenocarcinoma lung cancer to carboplatin plus
paclitaxel with bevacizumab if indicated. This is the standard treatment arm of most cooperative groups. It does not
allow the use of high-cost regimens such as carboplatin plus
pemetrexed plus bevacizumab before proof of effectiveness
in a randomized clinical trial, which has been the standard
for guidelines.36
In the second phase, in the first several visits, palliative
care and hospice concepts are introduced by someone on the
team. This can be the oncologist, social worker, or nurse, and
is in line with best current practices. This ensures that the
patient will have time to plan for when the disease grows,
whether in weeks or years. We can discuss the regimens,
outcomes, and costs with patients. Many patients will have
satisfied their $5,000 deductible and not care about additional costs after that, but some may be paying high dollar
amounts and we could manage their care with generic drugs
and a chest film or standard computed tomography (CT)
scan rather than a positron emission tomography CT scan at
three times the cost.37
In the third phase we can follow ASCO and NCCN
guidelines and not provide treatment to patients whose
nonsmall cell lung cancer has grown despite three treatments, or whose Eastern Cooperative Oncology Group performance status greater than 2 (remembering that 3 is in
bed or chair over half the time). If we have started palliative
care (PC) early, it will be easier to transition to full hospice
care. PC consultation improves appropriate hospice enrollment. Morrison and colleagues compared 1,427 patients who
did not have a PC consult with 296 patients who did,
matched in every other respect, and found that if the PC
team consulted, 30% of hospice-eligible patients were discharged to hospice compared with just 1% of patients not
seen by PC.38 Patients enrolled in PC programs at Sutter
Health enrolled in hospice 47% of the time, compared with
20% of similar patients.39,40 Appropriate transition to hospice not only improves care,41 but is associated with longer
survival29 and lower costs.42 PC offloads PC service provision from too-busy oncologists (e.g., surgeons, radiation
oncologists, leukemia specialists, etc.) to PC specialist providers,43 and the model can save money for oncology practices and be financially self-sustaining.44 As noted, PC
consultation leads to increased hospice referrals, and exposure to palliative care before hospice is strongly associated
with better caregiver quality of life, likely resulting from
symptom management and better understanding of prognosis and goals.45 The use of PC, with its attendant planning
about goals of care and follow-up, reduced readmissions to
the hospital per patient from 1.15 to 0.7 (a decrease of 36%)
in the last 6 months of life.46
To make this reality will require an audit of charts and
provision of feedback to the doctor and practice. We propose
that we add these overuse criteria to ASCOs Quality
Oncology Practice Initiative (QOPI), which clearly works to
reduce end-of-life chemotherapy.47
e49
SMITH, HILLNER, AND KELLY

Conclusion
ACKNOWLEDGMENT
Costs are rising at an unsustainable rate and, if continued, will lead to more uninsured, less access, and more
disparity. We can bend the cost curve downward by changing our practice patterns to provide evidence-based care,
use less-expensive drugs and tests, and increase the use of
palliative care and hospice. This will maintain quality and
decrease costs to pay for new and expensive advances.
The authors gratefully acknowledge the National Cancer

Institute, who supported this research through ACS Grant
#PEP-10-174-01 (T.S.), R01CA116227-01 (T.J.S.), 2R01CA10637005A1 (T.J.S.), and RC2CA148259 (B.E.H.).
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Thomas J. Smith
Research
Funding
Expert
Testimony
Other
Remuneration
American
Cancer Society
Bruce E. Hillner*
Ronan J. Kelly*
REFERENCES
1. Organization for Economic Co-Operation and Development. OECD
Health at a Glance 2011: Health expenditure and financing. http://stats.oecd.
org/Index.aspx?DataSetCodeSHA. Acessed February 24, 2012.
2. Himmelstein DU, Thorne D, Warren E, et al. Medical bankruptcy in the
United States, 2007: results of a national study. Am J Med. 2009;122:741-746.
3. Hillner BE, Smith TJ. Efficacy does not necessarily translate to cost
effectiveness: a case study in the challenges associated with 21st-century
cancer drug pricing. J Clin Oncol. 2009;27:2111-2113.
4. Smith TJ, Hillner BE. Bending the cost curve in cancer care. N Engl
J Med. 2011;364:2060-2065.
5. Gatesman ML, Smith TJ. The shortage of essential chemotherapy drugs
in the United States. N Engl J Med. 2011;365:1653-1655.
6. Riley GF, Lubitz JD. Long-term trends in Medicare payments in the last
year of life. Health Serv Res. 2010;45:565-576.
7. Braga S. Why do our patients get chemotherapy until the end of life? Ann
Oncol. 2011;22:2345-2348.
8. Brody H. Medicines ethical responsibility for health care reform: the top
five list. N Engl J Med. 2010;362:283-285.
9. Emanuel EJ, Pearson SD. Physician autonomy and health care reform.
JAMA. 2012;307:367-368.
10. Snyder L. American College of Physicians Ethics Manual. Ann Intern
Med. 2012;156:73-104.
11. Merimsky O, Kovner F, Inbar M, et al. Tamoxifen for disease-negative
but MCA-positive breast cancer patients. Oncol Rep. 1997;4:843-847.
12. Rustin GJ, van der Burg ME, Griffin CL, et al. Early versus delayed
treatment of relapsed ovarian cancer (MRC OV05/EORTC 55955): a randomised trial. Lancet. 2010;376:1155-1163.
13. Desch CE, Benson AB, Somerfield MR, et al. Colorectal cancer surveillance: 2005 update of an American Society of Clinical Oncology practice
guideline. J Clin Oncol. 2005;23:8512-8519.
14. Himmelstein DU, Thorne D, Warren E, et al. Medical bankruptcy in the
United States, 2007: results of a national study. Am J Med. 2009;122:741-746.
15. Ramsey SD, McCune JS, Blough DK, et al. Colony-stimulating factor
prescribing patterns in patients receiving chemotherapy for cancer. Am J
Manag Care. 2010;16:678-686.
16. Claxton G, DiJulio B, Whitmore H, et al. Health benefits in 2010:
premiums rise modestly, workers pay more toward coverage. Health Aff
(Millwood). 2010; Epub 2010 Sept. 2.
17. Smith TJ. The American Society of Clinical Oncology recommended
breast cancer surveillance guidelines can be done in a routine office visit.
J Clin Oncol. 2005;23:6807.
18. Loprinzi CL, Hayes D, Smith T. Doc, shouldnt we be getting some
tests? J Clin Oncol. 2003;21:108-111.
19. Meropol NJ, Schulman KA. Cost of cancer care: issues and implications. J Clin Oncol. 2007;25:180-186.
20. Smith TJ, Hillner BE. A way forward on the medically appropriate use
of white cell growth factors (CSFs). J Clin Oncol. In press.
21. Chan KKW, Siu E, Krahn MD, et al. A cost-utility analysis of primary
prophylaxis versus secondary prophylaxis with granulocute colony-
e50
stimulating factor in elderly patients wiht diffuse aggressive lymphoma

receiving curative-intent chemotherapy. J Clin Oncol. Epub 2012 Mar 5.
22. Keating NL, Beth Landrum M, Arora NK, et al. Cancer patients roles
in treatment decisions: do characteristics of the decision influence roles?
J Clin Oncol. 2010;28:4364-4370.
23. Wright AA, Zhang B, Ray A, et al. Associations between end-of-life
discussions, patient mental health, medical care near death, and caregiver
bereavement adjustment. JAMA. 2008;300:1665-1673.
24. Zhang B, Wright AA, Huskamp HA, et al. Health care costs in the last
week of life: associations with end-of-life conversations. Arch Intern Med.
2009;169:480-488.
25. Freund K, Weckmann MT, Casarett DJ, et al. Hospice eligibility in
patients who died in a tertiary care center. J Hosp Med. Epub 2011 Nov 15.
26. Huskamp HA, Keating NL, Malin JL, et al. Discussions with physicians
about hospice among patients with metastatic lung cancer. Arch Intern Med.
2009;169:954-962.
27. NHPCO Facts and Figures: Hospice Care in America. Alexandria, VA:
National Hospice and Palliative Care Organization. http://www.nhpco.org/
files/public/Statistics_Research/2011_Facts_Figures.pdf. Accessed January
2012.
28. Saito AM, Landrum MB, Neville BA, et al. The effect on survival of
continuing chemotherapy to near death. BMC Palliative Care. 2011;10:14.
29. Connor SR, Pyenson B, Fitch K, et al. Comparing hospice and nonhospice patient survival among patients who die within a three-year window.
J Pain Symptom Manage. 2007;33:238-246.
30. Massarelli E, Andre F, Liu DD, et al. A retrospective analysis of the
outcome of patients who have received two prior chemotherapy regimens
including platinum and docetaxel for recurrent non-small-cell lung cancer.
Lung Cancer. 2003;39:55-61.
31. Dy SM, Asch SM, Lorenz KA, et al. Quality of end-of-life care for
patients with advanced cancer in an academic medical center. J Palliat Med.
2011;14:451-457.
32. Smith TJ, Temin S, Alesi E, et al. American Society of Clinical Oncology
provisional clinical opinion: the integration of palliative care into standard
oncology care. J Clin Oncol. Epub 2012 Feb 6.
33. Temel JS, Greer JA, Admane S, et al. Longitudinal perceptions of
prognosis and goals of therapy in patients with metastatic non-small-cell lung
cancer: results of a randomized study of early palliative care. J Clin Oncol.
2011;29:2319-2326.
34. Hoverman JR, Cartwright TH, Patt DA, et al. Pathways, outcomes, and
costs in colon cancer: retrospective evaluations in two distinct databases.
J Oncol Pract. 2011;7:52s-59s.
cancer in the community setting. J Oncol Pract. 2010;6:12-18.
36. Azzoli CG, Baker S, Temin S, et al. American Society of Clinical
Oncology clinical practice guideline update on chemotherapy for stage IV non
small-cell lung cancer. J Clin Oncol. 2009;27:6251-6266.
37. McFarlane J, Riggins J, Smith TJ. SPIKE$: A six-step protocol for

delivering bad news about the cost of medical care. J Clin Oncol. 2008;26:
4200-4204.
38. Morrison RS, Dietrich J, Ladwig S, et al. Palliative care consultation
teams cut hospital costs for medicaid beneficiaries. Health Aff. 2011;30:454-463.
39. Meyers FJ, Linder J, Beckett L, et al. Simultaneous care: a model
approach to the perceived conflict between investigational therapy and
palliative care. J Pain Symptom Manage. 2004;28:548-556.
40. Meyers FJ, Carducci M, Loscalzo MJ, et al. Effects of a problem-solving
intervention (COPE) on quality of life for patients with advanced cancer on
clinical trials and their caregivers: Simultaneous Care Educational Intervention (SCEI): linking palliation and clinical trials. J Palliat Med. 2011;14:465473.
41. Peppercorn JM, Smith TJ, Helft PR, et al. American Society of Clinical
Oncology statement: toward individualized care for patients with advanced
42. Pyenson B, Connor S, Fitch K, et al. Medicare cost in matched hospice

and non-hospice cohorts. J Pain Symptom Manage. 2004;28:200-210.
43. Alesi E, Fletcher D, Muir C, et al. Palliative care and oncology
partnerships in real practice. Oncology. 2011;25:1287-1193.
44. Muir JC, Daly F, Davis MS, et al. Integrating palliative care into the
outpatient, Private Practice Oncology Setting. J Pain Symptom Manage.
2010;40:126-135.
45. Wittenburg-Lyles EM. What patients and families dont hear: Backstage communication in hospice interdisciplinary team meetings. J Hous
Elderly. 2009;23:92-105.
46. Nelson C, Chand P, Soratais J, et al. Inpatient palliative care consults
and the probability of hospital readmission. Perm J. 2011;15:48-51.
47. Blayney DW, McNiff K, Hanauer D, et al. Implementation of the
Quality Oncology Practice Initiative at a university comprehensive cancer
center. J Clin Oncol. 2009;27:3802-3807.
e51
NEW DIAGNOSTICS AND DEVICES IN THE ERA OF

COMPARATIVE EFFECTIVENESS
CHAIR
Daniel F. Hayes, MD
University of Michigan Medical Center
Ann Arbor, MI
SPEAKERS
Muin J. Khoury, MD, PhD
Office of Public Health Genomics, Centers for Disease Control and Prevention
Atlanta, GA
David Ransohoff, MD
University of North Carolina at Chapel Hill
Chapel Hill, ND
Why Hasnt Genomic Testing Changed the

Landscape in Clinical Oncology?
By Daniel F. Hayes, MD, Muin J. Khoury, MD, PhD, and David Ransohoff, MD
Overview: The omics revolution produced great optimism

that tumor biomarker tests based on high-order analysis of
multiple (sometimes thousands) of factors would result in truly
personalized oncologic care. Unfortunately, 10 years into the
revolution, the promise of omics-based research has not yet
been realized. The factors behind the slow progress in omicsbased clinical care are many. First, over the last 15 years,
there has been a gradual recognition of the importance of
conducting tumor biomarker science with the kind of rigor that
has traditionally been used for therapeutic research. However,
this recognition has only recently been applied widely, and
therefore most tumor biomarkers have insufficiently high
levels of evidence to determine clinical utility. Second, omicsbased research offers its own particular set of concerns,
UMOR BIOMARKERS are used to determine a patients current status and more importantly to predate
future events that might be modified by intervention.1
Tumor biomarkers can be analyzed in cancer or healthy
tissue, in secretions, and circulating in blood. Tumor biomarkers usually represent somatic changes that have
emerged during the process of carcinogenesis. However, it is
also reasonable to consider inherited germ-line differences
between individuals that predict higher risk of developing a
new malignancy or for estimating differential distribution,
metabolism, or response to a drug (pharmacogenomics).2,3
Assays for tumor biomarkers can identify changes, or
individual differences, in nucleic acids (DNA, RNA), proteins, lipids, whole cells, or tissue processes. Until recently,
an assay for a biomarker usually analyzed a single analyte,
or substance. Perhaps one of the best examples is the
development of assays for the estrogen receptor (ER) to
predict both prognosis and likelihood of responding to antiestrogen, or endocrine therapies.4,5 The biology of ER was
determined in the 1960s, and the first assay was a cumbersome test to biochemically measure binding of radioactively
labeled estrogen to the receptor. Subsequent tests using
specific antibodies to perform either enzyme-linked immunosorbent assays (ELISAs) and more recently immunohistochemistry (IHC) replaced the original ligand-binding
assays, and are now almost uniformly used in clinical
medicine. However, recently, newer assays that measure
RNA expression have been introduced. Regardless, these
are all tests that assay for a single analyte, ER, and not for
several analytes that might be combined into a unified index
designed to make a clinical decision.
High-Dimensional Biomarkers
In addition to measuring a single analyte, assessment of

complex processes, such as counting vessels to determine
levels of angiogenesis, or development of a multifactoral
index, such as tumor grade (which combines estimates of
relative gland formation, nuclear appearance, and mitotic
rate), represent higher-order forms of a single test. In this
regard, during the last decade, advances in molecular biology, technology, and bioinformatics have led to a new field
loosely described as omics. This field encompasses several
disciplines, generating high-dimensional data from global
e52
especially in regard to overfitting computational models and

false discovery rates. Researchers and clinicians need to
understand the importance of analytic validity, and the difference between clinical/biologic validity and clinical utility. The
latter is required to introduce a tumor biomarker test of any
kind (single analyte or omics-based), and are ideally generated by carefully planned and properly conducted prospective retrospective or truly prospective clinical trials. Only
carefully planned studies, which take all three of these into
account and in which the investigators are aware and recognize the enormous risk of unintended bias and overfitting
inherent in omics-based test development, will ultimately
result in translation of the exciting new technologies into
better care for patients with cancer.
sets of biologic molecules such as DNAs (genomics), RNAs

(transcriptomics), proteins (proteomics), and metabolites
(metabolomics).6 Massive amounts of data are used to
produce a profile, or signature, generated by a computational mathematical function model. This model may be
unsupervised, meaning that specimens are grouped computationally by apparent similarities in the omics patterns,
without regard to preconceived biologic or clinical associations. Alternatively, generation of the profiles can be supervised; in this case, the signature is pegged to some sort of
prospectively defined biologic or clinical characteristic of
interest. Ultimately, at least in regard to clinical care, one or
more omics-based test that reputedly has clinical utility in
guiding patient care is generated.
In an online continuous horizon scanning review of the
literature from 2009 to the present, researchers from the
Centers for Disease Control and Prevention found more than
400 new genomic and other omics-based tests in transition
from bench to bedside, of which the vast majority are related
to cancer.7 However, in 2012, few if any omics-based tests
have actually been widely adopted or embraced in the clinic.
Why not? There are several obstacles that block introduction and use of an omics-based test. These relate to generation and validation of tumor biomarker tests in general, but
in addition omics-based tests have special considerations
that have impeded progress in the field.8 It is essential that
basic, translational, clinical, and computational scientists,
and importantly clinicians caring for patients with cancer,
understand these obstacles and work to overcome them so
that patients receive better, more personalized oncologic
care than they do now.
From the University of Michigan Comprehensive Cancer Center, Ann Arbor, MI; Epidemiology and Genomics Research Program, Division of Cancer Control and Population
Sciences, National Cancer Institute, Bethesda, MD; Office of Public Health Genomics,
Centers for Disease Control and Prevention, Atlanta, GA; Departments of Medicine and
Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill, NC.
Address reprint requests to Daniel F. Hayes, MD, Breast Oncology Program, University
of Michigan Comprehensive Cancer Center, 6312 Cancer Center, 1500 E. Medical Center
Drive, Ann Arbor, MI 48109-0942; email: hayesdf@umich.edu.
1092-9118/10/1-10
GENOMIC TESTING IN CLINICAL ONCOLOGY

What Are the Problems?
A Matter of Semantics
The word validation means many things to many people.8,9 As in all scientific endeavors, carefully used semantics and precisely defined meanings are critical to
development of tumor markers and incorporation into clinical care. In this regard, the independent multidisciplinary
Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Working Group, convened by the Centers for
Disease Control and Prevention, has defined three important semantics that have greatly helped to organize the
field.10 They have defined three types of validation.
Analytic validity. Analytic validity refers to the preanalytic and analytic factors that contribute to accuracy, reliability, and reproducibility of the specific assay test for the
biomarker
Clinical validity. Clinical validity implies that the biomarker assay distinguishes two or more subgroups within a
population that have different biologic or clinical outcomes.
Clinical validity does not imply that a tumor biomarker
assay should be used to care for patients, but it is unlikely
the assay will be useful if it does not at least have clinical
validity.
Clinical utility. Clinical utility requires that an assay
with analytic and clinical validity be shown to be useful in
improving patient outcomes to the extent that the patients
care should differ than if the marker results were not
available.
What is Clinical Utility? Every clinician makes clinical
decisions on a daily basis. He/she carefully weighs the
relative benefits of a planned intervention against the relative risks and costs, in terms of both dollars and inconvenience to the patient. Diagnostics tests are used to narrow
the funnel of differential diagnosis so that any therapeutic
plan is focused as carefully as possible on the individual
characteristics of the patient at hand. For example, administering chemotherapy to every person in our society would
probably result in benefiting those who happened to have
cancer, but would induce prohibitive toxicity and costs for
the vast majority of individuals who do not. So, of course,
clinicians use history, physical examination, blood tests,
radiologic evaluation, and pathology to focus our therapy on
those who have cancer, and even more so, on those who have
a life-threatening cancer and for whom the benefits of a
specific therapy are likely to outweigh the risks. For an
individual who has a cancer and experiences disease response, the benefit is the same (100%), regardless of what we
KEY POINTS
Tumor biomarker research requires a careful understanding of analytical and clinical validity and clinical utility.
Omics-based research is fraught with special problems related to over-fitting and lack of well-designed
studies to generate validity and utility.
Specific pathways have been proposed to generate
clinical utility by either conducting prospective, retrospective, or truly prospective clinical trials.
know about that patient or whether everyone else is treated.

However, the magnitude of the overall benefit for society is
enhanced when the therapy is administered only to these
patients and not to everyone else. Thus, biomarkers help
identify those patients who either do not need or will not
benefit from therapy. Clinical utility of a tumor marker can
be established if the marker identifies those patients who
are so unlikely to experience an event (a new cancer, or a
recurrence or death as a result of an established cancer), or
those so unlikely to respond or benefit, that the risks are the
same or outweigh the benefits of treatment and they should
forgo therapy. Determination of clinical utility requires
judgment, on the part of the caregiver, the patient, and
society, in regard to the relative benefits comapred with the
risks and costs of the therapeutic plan under consideration.
For example, adjuvant endocrine therapy for patients
with breast cancer clearly reduces the odds of recurrence
and death.11 In general, adjuvant endocrine therapy is
reasonably well tolerated with few life-threatening toxicities, but it is relatively expensive, and it does produce
annoying adverse effects, which occasionally can be intolerable. ER is a powerful predictive factor for endocrine therapy.5 However, ER does not predict who will benefit. Indeed,
adjuvant endocrine therapy decreases relapse in only approximately one-half of ER-positive patients who are destined to experience recurrence. ER does, though, strongly
identify patients who will not benefit. The overall survival
curves for adjuvant tamoxifen compared with nil in ER-poor
patients are almost completely overlapping.11 Are there
some patients with ER-negative cancers who benefit? Almost certainly, but clinicians, patients, and guideline panels
feel that the marker is sufficiently strong that the potential
benefits of adjuvant endocrine therapy for these few patients
with ER-negative breast cancer do not outweigh the risks
that would be accrued for all of the others.12
Demonstrate Clinical Utility of a Tumor Biomarker
In medicine, adoption of a new therapeutic strategy,

especially related to a new drug, requires high levels of
evidence of a sufficiently large benefit that the intervention
outweighs the risk. In most cases, to recommend adoption
into standard of care, guideline bodies require one or more
prospective randomized trials in which use of the new
strategy is compared with standard of care without it.
Unfortunately, although it should be, the same has not been
true for clinical research of diagnostic devices, especially
related to tumor biomarkers. Therefore, several markers
that might be very helpful to personalize oncologic care have
languished in uncertainty regarding their clinical utility.
Perhaps just as unfortunately, markers with unproven clinical utility have been used to manage care, often with
adverse outcomes. Both of these circumstances lead to poor
disease management, and highlight the phrase that a bad
tumor marker is as bad as a bad drug.
In 1996, members of the ASCO Tumor Marker Guideline
Committee proposed a tumor marker utility grading system,
which included a level of evidence (LOE) scale that could be
used to determine the relative quality of data supporting
claims for clinical utility.13 At that time, almost all tumor
biomarker research fell into the LOE III category or worse,
principally because most studies were performed out of
convenience of having available specimens and a novel
assay. Ideally, to generate higher levels of evidence, inves-
e53
HAYES, KHOURY, AND RANSOHOFF
tigators should develop prospectively designed and conducted trials to directly test the test. Indeed, a few such
trials have been or are being conducted in North America,
principally in breast and colorectal cancers. Several different trial designs to test tumor biomarkers have been proposed, but the details of these are outside the scope of this
review and the reader is referred to two very well-written
reviews of this subject.14,15
Prospective trials are expensive and time consuming. One
advantage of tumor biomarker compared with therapeutic
agent research is the opportunity to use archived specimens
to generate high levels of evidence that might support, or
refute, clinical utility of a tumor biomarker. Indeed, the
appeal of this apparent advantage has worked to the disfavor of many tumor biomarkers, leading to lower-level evidence studies using conveniently available specimens
without regard to trial design or proper statistical analysis.13 However, the ASCO Tumor Marker Guidelines Committee and others, such as the National Cancer Center
(NCCN) committees, have relied as much as possible on
having LOE I data to recommend use of a tumor biomarker
to direct care. This publically stated strategy has led to
better and more rigorous studies addressing the true clinical
utility (as opposed to clinical validity) of specific tumor
biomarker tests, including those that have been generated
from omics research. Recently, Simon, Paik, and Hayes have
proposed a refinement of the original ASCO LOE scale that
establishes a hierarchy of tumor biomarker studies, ranging
from very poor (retrospective studies of convenience using
archived specimens that were not collected, processed, or
stored with clinical trial grade annotated data) to highlevel evidence prospective retrospective studies. The latter type of studies use archived specimens from previously
conducted clinical trials, and although not as well-regarded
as true prospective studies, can, if performed carefully and
validated properly, generate LOE I data to determine clinical utility.16
Embedded in such studies are the fundamental concepts
of the Evaluation of Genomic Applications in Practice and
Prevention (EGAPP) semantics.10 An assay for a biomarker
must be analytically validated in regard to technical accuracy, reproducibility, and reliability, especially in regard to
the types of specimens to be used in the clinic. Furthermore,
if the assay does not at least have clinical/biologic validity,
it is unlikely to have clinical utility. However, proper study
design must be used to generate the LOE needed to show
clinical utility. To develop clinically useful tumor biomarkers, it is essential to carefully plan the experiment with
well-developed, prospectively stated hypotheses regarding
the intended use: risk assessment, detection of occult tumor,
prognosis, prediction of response or resistance to specific
therapy, or monitoring.17 Moreover, analytic strategies
should be determined in writing before the investigation
is begun. These mandates do not preclude exploratory analyses and hypothesis-generating discovery studies, but the
latter do not constitute high LOEs that demonstrate clinical
utility.
How Do Omics-Based Tests Differ from Other Tumor Biomarkers?
In addition to the generic considerations related to tumor

biomarker research in general, research of omics-based
tests is saddled with several specific concerns. First, highdimensional data are particularly prone to overfitting.9,18-20
e54
Computational models often appear to perform beautifully

during the discovery phase of omics-based research, but
frequently fail when applied to an independent specimen/
data set.9 There are several factors that lead to these types
of false discovery. Not uncommonly the profile, or signature,
has been generated without supervision of a clinical outcome
or biologic hypothesis. Even if the discovery phase is performed with supervision to a biologic or clinical factor, the
enormous number of factors that have been put into the
original model (for example, expression of 10,000 genes)
and the relatively small data sets that are often used, make
the lack of subsequent validation the exception rather than
the rule.
Has There Been Any Success in Translating Omics-Based Tests to
Clinical Oncology?
Although disappointing, the field has not been bereft of

success. Perhaps the best examples are illustrated in two
separate, although linked, investigations that span the
Atlantic Ocean. In the late 1990s, by using gene expression
microarray technology, investigators from Amsterdam reported generation of a 70-gene signature that divided patients with early-stage breast cancer into very good versus
poor prognosis.21 Subsequent single and multi-institutional
studies demonstrated clinical validity of this assay.22,23
Unfortunately, the original and follow-up validation studies
were conducted using specimens that had been banked
and stored as part of routine care (as opposed to part of a
prospective clinical trial), without regard to a specifically
identifiable question that would demonstrate clinical utility,
and use of this assay has not been recommended by clinical
guidelines committees (ASCO or NCCN).12,24 A prospective
randomized trial (the MINDACT trial) is now underway in
Europe to determine whether this omics-based test should
or should not be used to guide administration of adjuvant
chemotherapy in node-negative patients.
A second omics-based test was developed, in part, by
incorporating some of the genes identified in the 70-gene
signature, and in part by choosing logical and biologically
based candidate genes. In this manner, investigators from
the National Surgical Adjuvant Bowel and Breast Project
(NSABP) generated a 21-gene assay specifically to identify a
group of women with node-negative, ER-positive breast
cancer treated with adjuvant tamoxifen whose prognosis
was so good that, even if chemotherapy were effective, so few
patients would benefit that its risks would outweigh its
utility.25 This test was shown to have superb analytic and
clinical validity in a blinded analysis of a prestated hypothesis applying the assay to formalin-fixed, paraffin embedded
tissue.26,27 Moreover, the investigations used to determine
clinical utility were prospective retrospective studies using
material collected from women who participated in two
prospective clinical trials addressing the benefits of tamoxifen in this patient group (NSABP B20 and B14). Thus,
although the Simon-Paik-Hayes criteria had not yet been
proposed at the time of the original publications of this
assay, the ASCO Tumor Marker Guidelines Panel as well as
the NCCN Breast Cancer Guidelines Panel have both recommended the use of the 21-gene assay for this clinical
use.12,24
The predictive role of either of these assays for benefit or
resistance to chemotherapy has been proposed on the basis
of data from other prospective retrospective studies.28,29
GENOMIC TESTING IN CLINICAL ONCOLOGY
However, the specimen/data sets to validate the clinical

utility of these exploratory observations are not available.
Thus, the Eastern Cooperative Oncology Group and Southwest Oncology Group have completed or are conducting
prospective clinical trials (TailoRX and RxPonder, respectively) within the North American Breast Cancer Group to
address the predictive role of the 21-gene recurrence score in
ER-positive patients with breast cancer.
Conclusion
Unfortunately, the field of tumor biomarker research has

been chaotic and haphazard, leading to many published
papers in the peer-reviewed literature, but very few markers
that truly have clinical utility or that can be recommended

for routine patient care. This situation has led to two
problematic circumstances: 1) use of biomarkers in the
absence of high levels of evidence supporting their clinical
utility; and 2) lack of biomarkers that can lead to truly
personalized cancer care with high confidence. It is imperative that the field take major actions to break the vicious
cycle that has led to these circumstances and to do the
rigorous research needed to provide proper assessments,
because a bad tumor marker is as bad as a bad drug. When
these factors are recognized and incorporated into tumor
biomarkers studies, the dream of personalized oncologic care
will be more likely to become a reality.
Author
Daniel F. Hayes
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Biomarker
Strategies;
Chugai Pharma
OncImmune
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Novartis; Pfizer;
Veridex
Muin J. Khoury*
David Ransohoff*
REFERENCES
1. Khleif SN, Doroshow JH, Hait WN. AACR-FDA-NCI Cancer Biomarkers
Collaborative consensus report: advancing the use of biomarkers in cancer
drug development. Clin Cancer Res. 2010;16:3299-3318.
2. Weinshilboum R. Inheritance and drug response. N Engl J Med. 2003;
348:529-537.
3. Wang L, McLeod HL, Weinshilboum RM. Genomics and drug response.
N Engl J Med. 2011;364:1144-1153.
4. Hammond ME, Hayes DF, Dowsett M, et al. American Society of Clinical
Oncology/College of American Pathologists guideline recommendations for
immunohistochemical testing of estrogen and progesterone receptors in
breast cancer (unabridged version). Arch Pathol Lab Med. 2010;134:e48-e72.
5. Hammond ME, Hayes DF, Dowsett M, et al. American Society of Clinical
Oncology/College Of American Pathologists guideline recommendations for
immunohistochemical testing of estrogen and progesterone receptors in
6. Field D, Glockner FO, Garrity GM, et al. Meeting report: the fourth
Genomic Standards Consortium (GSC) workshop. OMICS. 2008;12:101-108.
7. Gwinn M, Grossniklaus DA, Yu W, et al. Horizon scanning for new
genomic tests. Genet Med. 2011;13:161-165.
8. Ioannidis JP, Khoury MJ. Improving validation practices in omics
research. Science. 2011;334:1230-1232.
9. Ransohoff DF. Rules of evidence for cancer molecular-marker discovery
and validation. Nat Rev Cancer. 2004;4:309-314.
10. Teutsch SM, Bradley LA, Palomaki GE, et al. The Evaluation of
Genomic Applications in Practice and Prevention (EGAPP) Initiative: methods of the EGAPP Working Group. Genet Med. 2009;11:3-14.
11. Early Breast Cancer Trialists Collaborative Group (EBCTCG). Effects
of chemotherapy and hormonal therapy for early breast cancer on recurrence
and 15-year survival: an overview of the randomised trials. Lancet. 2005;365:
1687-1717.
12. Harris L, Fritsche H, Mennel R, et al. American Society of Clinical
Oncology 2007 update of recommendations for the use of tumor markers in
13. Hayes DF, Bast RC, Desch CE, et al. Tumor marker utility grading
system: a framework to evaluate clinical utility of tumor markers. J Natl
Cancer Inst. 1996;88:1456-1466.
14. Sargent DJ, Conley BA, Allegra C, et al. Clinical trial designs for
predictive marker validation in cancer treatment trials. J Clin Oncol. 2005;
23:2020-2027.
15. Freidlin B, McShane LM, Korn EL. Randomized clinical trials with
biomarkers: design issues. J Natl Cancer Inst. 2010;102:152-160.
16. Simon RM, Paik S, Hayes DF. Use of archived specimens in evaluation
of prognostic and predictive biomarkers. J Natl Cancer Inst. 2009;101:14461452.

17. Henry NL, Hayes DF. Uses and abuses of tumor markers in the
diagnosis, monitoring, and treatment of primary and metastatic breast
cancer. Oncologist. 2006;11:541-552.
18. Simon R. Roadmap for developing and validating therapeutically relevant genomic classifiers. J Clin Oncol. 2005;23:7332-7341.
19. Simon R. Development and validation of therapeutically relevant
multi-gene biomarker classifiers. J Natl Cancer Inst. 2005;97:866-867.
20. Simon R. Development and evaluation of therapeutically relevant
predictive classifiers using gene expression profiling. J Natl Cancer Inst.
2006;98:1169-1171.
21. vant Veer LJ, Dai H, van de Vijver MJ, et al. Gene expression profiling
predicts clinical outcome of breast cancer. Nature. 2002;415:530-536.
22. van de Vijver MJ, He YD, vant Veer LJ, et al. A gene-expression
signature as a predictor of survival in breast cancer. N Engl J Med.
2002;347:1999-2009.
23. Buyse M, Loi S, vant Veer L, et al. Validation and clinical utility of a
70-gene prognostic signature for women with node-negative breast cancer.
24. Carlson RW, Allred DC, Anderson BO, et al. Invasive breast cancer.
J Natl Compr Canc Netw. 2011;9:136-222.
25. Paik S, Shak S, Tang G, et al. A multi-gene RT-PCR assay using fixed,
paraffin-embedded tumor tissue to predict the likelihood of breast cancer
recurrence in node negative, estrogen receptor positive, tamoxifen-treated
patients. N Engl J Med. 2004;351:2817-2826.
26. Cronin M, Pho M, Dutta D, et al. Measurement of gene expression in
archival paraffin-embedded tissues: development and performance of a 92gene reverse transcriptase-polymerase chain reaction assay. Am J Pathol.
2004;164:35-42.
27. Cronin M, Sangli C, Liu ML, et al. Analytical validation of the Oncotype
DX genomic diagnostic test for recurrence prognosis and therapeutic response
prediction in node-negative, estrogen receptor-positive breast cancer. Clin
Chem. 2007;53:1084-1091.
28. Albain KS, Barlow WE, Shak S, et al. Prognostic and predictive value
of the 21-gene recurrence score assay in postmenopausal women with nodepositive, oestrogen-receptor-positive breast cancer on chemotherapy: a retrospective analysis of a randomised trial. Lancet Oncol. 2010;11:55-65.
29. Straver ME, Glas AM, Hannemann J, et al. The 70-gene signature as a
response predictor for neoadjuvant chemotherapy in breast cancer. Breast
e55
MANAGEMENT OF CHRONIC LYMPHOCYTIC

LEUKEMIA
CHAIR
Neil E. Kay, MD
Mayo Clinic
Rochester, MN
SPEAKERS
Peter Hillmen, PhD
Leeds General Infirmary
Leeds, United Kingdom
John G. Gribben, DSc, MD
St. Bartholomews Hospital
Predicting Clinical Outcome in B-Chronic

Lymphocytic Leukemia
By Neil E. Kay, MD
Overview: B-Chronic lymphocytic leukemia (CLL) is a relatively common B-cell malignancy that has a very heterogeneous clinical course, despite carrying the designation of
chronic, which is a gross oversimplification. Being able to
give some estimate of the rates of disease progression and
overall survival (OS) at first diagnosis is, therefore, important
in CLL. The ability to accurately predict response to therapy,
as well as subsequent duration of response to therapy, is
required given the variability of current therapies to induce
and sustain treatment responses. The holy grail of prognostics
HE MAJOR features of CLL that dictate need for

prognostics include the following points: it is a relatively common leukemia (approximately one in 100,000
patients in North America),1 the disease is incurable with
the exception of transplant, clinical courses are notoriously
variable, and the patients have considerable anxiety with
this diagnosis. The latter is now well documented and it is
attributable to the fact that we often do not treat these
patients with Rai stage 0 1 on diagnosis. This practice is
based on clinical trial data suggesting treatment is not a
benefit in the early stages of disease.2 It is also partially
because we do not have a ready portfolio of relatively
nontoxic agents for treating our early-stage patients with
CLL. The anxiety found in CLL is pervasive and has been
quantified, with clear evidence that the anxiety may exist for
years postdiagnosis.3,4 The cause of this psychologic distress
is multifactorial but one causal aspect is the current dogma
to watch and evaluate early-stage CLL as best as is possible
for a given patient. The case for prognostics, therefore, is
that it provides useful guides for initial patient assurance
(or not) and subsequent treatment. It is this need for
prognostic assistance in early-stage CLL that primarily
drove the research to develop more effective and powerful
methods that inform us. The result of this research is now
evident in that there are prognostic parameters used alone
or in model systems that assist us in counseling and/or
management of the majority of patients with CLL. These
models can incorporate either clinical-based factors or molecular parameters that reflect the biology of CLL B cell and
its microenvironment. These prognostic features or models
are helpful for predicting time to first therapy, extent of
response to treatments, and duration of response. In some
cases, these features or models guide us in limited ways to
treatment choices. Although not ideal, the wise use of these
prognostics is and should be a major assist to us in CLL
practice.
In this article, I will survey the available maneuvers and
tests that can be used for most patients and that will provide
for the practitioner and the patient a guidepost to risk
stratification. These tests can tell us whether a given patient
has a lower or higher risk of progressing over time. In
addition, it is possible to use these tests to guide patient
counseling, including determining how often a patient
should be seen by the practitioner and potentially some
treatment selection.
394
would be to state with accuracy which therapy or types of

therapy are best for a given patient. Although there is no
complete answer to prognostic counseling, there is a continued development of markers specific to the CLL B cell and/or
to its environment, as well as of testing of prognostic models.
These models use both traditional and novel prognostic markers that can aid in the dissection of outcome for early-stage
CLL in terms of progression risk and time to therapy. This has
resulted in significant enhancement of our ability to guide and
predict outcome for our patients with CLL.
Clinical Course and Prognostic Parameters
The current dogma is that for every 100 patients with

CLL, approximately one-third will not progress to treatment
even over decades, one-third will eventually progress, and
one-third will need more urgent treatment than the rest.5 In
addition, because of intensive research on early-stage CLL it
is known that approximately 50% of those patients will have
high risk based on the presence of adverse prognostic features.6,7 Indeed, we can now identify a cohort of patients
with high-risk CLL at diagnosis who will have rapid disease
progression, poor response to treatment, and poor survival
based on prognostic methods developed from an improved
understanding of the biology of CLL. The prognostic parameters that are used to define this risk can be subdivided into
both traditional/clinical and novel prognostic factors. The
traditional and clinical factors are usually based on quantifiable plasma factors (beta-2 microglobulin, lactic dehydrogenase [LDH]), Rai stage, or hematologic features such as
bone marrow features (diffuse infiltration) or levels of blood
lymphocyte counts over time. The novel prognostic parameters that are in routine practice are leukemic cell based.
Examples of the latter include the presence of membrane
proteins such as CD38 or CD49 days, cytoplasmic presence
of ZAP-70, and prognostic nuclear features including immunoglobulin variable heavy chain (IgVH) gene mutation status and cytogenetic abnormalities on fluorescent in situ
hybridization (iFISH). In the assessment of these prognostic
factors for CLL it is critical to consider and clarify the
clinical features that are studied for association with the
particular prognostic factor(s). Although the focus has been
on using prognostics for previously untreated CLL, there is
a special need for more information on subsequent course for
relapsed patients. The most helpful information in these
cases is the time of the patients relapse from initial therapy.
Here we also need better prognostic variable to predict their
subsequent clinical outcomes. However, for now, the treatment responses following initial therapy can be best predicted by clinical features that include extent of response
From the College of Medicine, Mayo Clinic, Rochester, MN.

Address reprint requests to Neil E. Kay, MD, College of Medicine, Mayo Clinic, Stabile
6-28, 200 First Street SW, Rochester, MN 55905; email: kay.neil@mayo.edu.
1092-9118/10/1-10
PREDICTIVE PARAMETERS IN CLL

Table 1. Prognostic Factors That Are of Help in Defining
Risk of Progression
Factor
Biologic Role
CD38
ZAP-70
CD49d
iFISH panel
Cell activation
Cell signaling
Cell adhesion
Common recurring
genetic defects
Test Assay1
Flow cytometry
Flow cytometry
Flow cytometry
Probes detected by microscopic
fluorescence
1. None of the prognostic factors are approved by the U.S. Food and Drug
Administration for use in CLL with the exception of some iFISH probes.
High Risk factors for adverse clinical outcomes
Presence of del(17p13) or del(11q22) or complex iFISH (more than one iFISH
detectable defect)
o predict for shorter time to therapy (del(17p13)) except if seen close to
diagnosis
o del(11q22) is associated with bulky disease and less response to therapy
Unmutated status with or without ZAP-70
o predict for shorter time to therapy
o predict for less durable response to therapy
del(17p13) with or without inactivating p53 mutations
o predict for poor response to chemoimmunotherapy
o often need stem cell transplant ultimately or experimental drug approaches
to first-line therapy and the first remission duration. This

article primarily focuses on the most recent work on prognostic parameters that best predict the clinical course for
newly diagnosed patients, their subsequent response to
upfront therapy, and its duration.
Useful Prognostic Variables in CLL
Traditional Prognostic Features
The most useful prognostic features incorporate clinical

features of the patient at diagnosis and have been used for
decades as the Rai and Binet staging systems. Although
these are still helpful, they do not identify good from bad
prognostic cohorts in the patient with early-stage CLL.
Because we now identify at least 70% of patients in earlystage disease, these staging systems are no longer satisfactory as sole predictors (Table 1). To improve on earlier
predictors for patients with CLL, a number of routine blood
and/or serum factors have been assessed for their utility.
The lymphocyte doubling time (LDT), calculated by determining the number of months it takes the absolute lymphocyte count (ALC) to double in number, is a marker of disease
kinetics that has been found to correlate with both
progression-free (PFS) and OS rates.8 Beyond this, the ALC
KEY POINTS
Prognostic information is very helpful in initial treatment of patients with B-chronic lymphocytic leukemia.
Biomarkers that reflect key biologic aspects of the
CLL B cell and/or its environment are highly informative of disease course.
Prognosis for time to treatment from first diagnosis is
now feasible for a given patient.
Models for prediction of durability of response to
standard therapies are available.
Targeted therapy based on prognostics are emerging
but still immature.
has also been shown to predict both PFS and OS.9,10 LDT
values have been shown to be predictive of outcome for
patients with early-stage CLL. A median PFS of 20 months
for Binet stage A patients with an LDT of 12 months or less
was found, compared with 75 months for those with an LDT
of more than 12 months. The same study found a median
PFS of 17 months for patients with an ALC of greater than
30 109/L compared with 88 months for those with an
ALC of less than 30 109/L9. The serum level of beta-2
microglobulin, when the level is greater than 4.0 mg/dL at
presentation, can be an adverse prognostic feature,9,10 but
this measure is not valuable as a sequential tool and may be
confounded by other clinical variables such as renal disease
and infectious or inflammatory conditions. One study found
that the median PFS of patients with beta-2 microglobulin
levels greater than 3.5 mg/L was 13 months compared with
75 months for those with a beta-2 microglobulin levels less
than 3.5 mg/L9. Similarly the use of a LDT can be difficult
because of associated disease conditions that will also raise
lymphocyte counts.
Novel Prognostic Features
Here we refer to prognostic features that are clearly

related to leukemic B-cell biology. The novel prognostic
markers that are in widespread use in clinical practice
include: immunoglobulin heavy-chain variable region
(IgVH) mutational status, interphase fluorescence in situ
hybridization (iFISH) abnormalities, CD38, and zetaassociated protein (ZAP)-70. These markers on an individual
basis can be used to predict clinical outcome where patients
with mutated IgVH genes, low CD38 expression, low ZAP-70
expression, and the absence of del(17p13) del(11q22) are all
associated with a good prognosis.11-20 Each of these markers
has been shown to be a predictor of time to treatment and
of OS on univariate analysis. Of special interest, it is only
iFISH defects of novel prognosis that are known to change
over time in CLL. This latter aspect is of importance as
clonal evolution (acquisition of new genetic changes or of
increasing percentages of abnormal leukemic cells by iFISH)
is associated with more aggressive disease. Clonal evolution
is associated with short survival, but the difference in
survival was limited to patients who acquired a del(17p13)
or del(11q22) where the median survival postclonal evolution was only 1.3 years.21 In addition newer information on
iFISH defects initially associated with good clinical outcome,
such as 13q-, may not always hold true. Thus for patients
with CLL with 13q- those with higher percentages or those
with structural defects larger than the typical defect (shown
by single nucleotide polymorphisms [SNP] chips) have a
greater tendency to have worse clinical outcome.22
Important issues regarding these prognostic parameters
are that the pivotal studies identifying these prognostic
markers used cohorts of patients with all stages of CLL. Also
there is no currently accepted or agreed on mandate for use
of these markers for prediction of disease outcome for CLL
despite their usefulness. The need for prognostic evaluation
approaches, however, is now buttressed by many studies
conducted in patients with recently diagnosed CLL. The use
of these novel prognostics can help better define the heterogeneity in outcome and can at least promote enhanced
counseling information and, thus, potentially alleviate significant anxiety for many patients.
The following can be most useful in predicting more
395
NEIL E. KAY
Table 2. Prognostic Models
Factors in Model
Risk Groups Defined
Clinical Application Focus
*
*
*
*
*
*
*
*
*
*
*
*
*
Low
Intermediate
High
a) To predict survival (5 and 10 yr)

b) Useful for all Rai stages
Ref 10
Multivariable models developed for early stage
Useful for predicting TTFT and OS
Ref 30
Nomograms Developed
a) Predict response to initial therapy

b) Useful for prediction of CR, TTFT, and OS
Ref 29
Age
Beta-2 microglobulin
Absolute lymphocyte count
Sex
Rai stage
Number of involved lymph node groups
Lymphocyte doubling time
IGHV Mutation Status
CD 38
Age at Diagnosis
Treatment Regimen
Beta-2 Microglobulin
Age
Reference
Abbreviations: CR, complete response; OS, overall survival, TTFT, time to first therapy.
aggressive disease for early-stage disease: presence of highrisk iFISH (del(17p13) del(11q22) or complex iFISH (multiple iFISH defects), p53 mutations, unmutated IgVH status,
or combinations of these defects. The studies of IgVH status
are most useful in showing the relative power of a single
prognostic factor related to the biology of the CLL B cell.
Thus the pivotal study of IgVH demonstrated that patients
with CLL who had mutated immunoglobulin IgVH had a
survival of more than 20 years, whereas those with unmutated IgVH had median survival of around 8-years.11,12 The
ability of IgVH mutation status to stratify OS remains when
applied exclusively to patients with early-stage disease.12
For prediction of response to therapy, patients with
del(17p13) or del(11q22) defects are less likely to have a
vigorous response to even the most aggressive upfront treatments. For patients with unmutated and/or del(17p13), the
durability of response to therapy is markedly reduced.
Recent work has shown that clinical course in patients with
iFISH-detectable del(17p13) is profoundly affected by the
presence or absence of TP53 mutations.23 If TP53 mutations
(detected by SNP microarray or by sequencing of the p53
gene) are present along with the del(17p13), OS is adversely
affected.24 Most recently there have been advances in putting together prognostic models for prediction of clinical
courses and response to therapy.
Overview of Prognostic Models in CLL
Most patients in the 21st century are diagnosed with

early-stage disease and within this early-stage group, the
vast majority do not require immediate treatment. There is,
however, a small subgroup of patients who do present with
early-stage disease and with symptoms such as fatigue or
night sweats or with advanced-stage disease, requiring
treatment. Although the latter cohort is not a clinical dilemma in terms of treatment decisions, the majority of
patients with early-stage CLL do need better prognostication and subsequent enhanced counseling with regard to
follow-up times and to their individual risk of progression.
There have been a limited but important set of published
work on the development of prognostic models for important
clinical end points such as time to first treatment, clinical
levels of responses, and duration of responses. These models
are important to develop and validate because they may be
very useful for patient clinical trial stratification and for
providing patient-comparison platforms across clinical trials
(Table 2).
396
A pivotal and pioneering study of clinical outcome was

conducted in an initial large series of patients with CLL
patients. This analysis identified six factors including age,
stage, sex, ALC, beta-2 microglobulin, and number of lymph
node regions involved that were independently associated
with patient survival.10 From this set of data these various
factors were then combined in a prognostic index that was
able to predict OS more accurately than clinical stage alone.
Further work validating the prognostic index was subsequently published by two groups indicating that this is a
valuable and practical model.25,26 In one of these studies it
was extended to show that the index remained useful even
when applied just to Rai stage 0 for the prediction of time to
treatment and of OS.25
Subsequent prognostic models have been developed that
also incorporated the biologic factors shown to reflect the
pathobiology of the CLL B cell. The initial work on this
included an analysis of four novel markers: ZAP-70, CD38,
iFISH, and IgVH mutation status for predicting time to
treatment in a study of more than 1,000 patients with
CLL.19 The analysis found three groups: low-, intermediate-,
and high-risk groups based on ZAP-70 and IgVH mutation
status. In summary ZAP-70 positive patients were high
risk irrespective of IgVH mutation status; patients with CLL
who are ZAP-70 negative can be classified as low (mutated
IgVH) or intermediate (unmutated IgVH) risk based on
IgVH status. However widespread application of this model
is difficult because of the lack of standardization of ZAP-70
measurements.
Further work on the ability of iFISH detectable del(17p13)
and IgVH mutation status in a cohort of 99 nontreated
early-stage patients found surprisingly better time to treatment and survival in some.27 However if Rai stage was 1 or
greater and/or the patient also had an unmutated IgVH
status, the favorable clinical course dropped off sharply.
A European study found that Binet A patients could be
categorized into low-, intermediate-, and high-risk groups
based on the combination of iFISH and IgVH mutation
status.15 Here patients with del(17p13) were high risk
regardless of mutation status, whereas patients with
del(11q22) and/or unmutated IgVH genes were intermediate
risk. Finally patients with mutated IgVH without del(17p13)
or del(11q22) were low risk.
Another recent study analyzed 930 patients with CLL for
time to first treatment where both traditional and new
prognostic factors were measured.28 For this study the
PREDICTIVE PARAMETERS IN CLL
patients did not have active CLL requiring initiation of

treatment within 3 months of first visit and were observed
for time to first treatment. The results from this study
showed a mix of both types of prognostic parameters were
found to be independently associated with a shorter time to
first therapy. These parameters included: three involved
lymph node sites, increased size of cervical lymph nodes,
presence of del(17p13) or del(11q22) increased serum lactate
dehydrogenase, and unmutated IgVH mutation status.
From a subset of patients a multivariable model was constructed and a nomogram was developed using the latter
prognostics to predict the risk of time to first treatment. The
authors make the point that this nomogram model system
was constructed from only patients with early-stage CLL
and where they had very prolonged clinical follow-up using
strict criteria to decide on treatment initiation.
The same group of investigators has also attempted to
develop prognostic features to better predict response to first
therapy and subsequent clinical course. Rationale for this
is that responses to upfront therapy can be very heterogeneous. Knowledge of critical patient features that are
strongly associated with clinical courses posttherapy will
again aid in counseling and even in performing relevant
clinical trials. For 595 patients who underwent upfront
therapy, researchers looked for predictors of three aspects:
complete response, time to treatment failure, and OS.29 In
patients who achieved a complete response there was a more
favorable durability and survival, but having received combination chemotherapy with antibody regimen was very
significant for all three clinical outcomes. The use of various
clinical parameters (i.e., age, serum beta-2 microglobulin)
generated two nomograms that could be used to predict 5and 10-year OS.
Finally there was a very recent and quite large study of

1,154 patients with Binet stage A CLL recently reported
that studied both traditional and novel prognostic factors in
association with clinical outcome.30 This study found that
LDT was most significantly associated with time to first
therapy but that IgVH was strongest in association with OS.
In addition only LDT, mutation status, CD38, and age at
diagnosis were independent prognostic variables for time to
first therapy and OS. Their recommendation was to assess
IgVH mutation status and CD38 expression at initial evaluation as they have independent prognostic value in earlystage CLL.
Conclusion
The use of both traditional and novel prognostic parameters can be a very valuable ally in determining the relative
risk of the individual patients clinical course of CLL. There
are multiple parameters other than individual prognostic
factors that can be used at initial prognostic evaluation and
can inform the practitioner about disease progression risk
and time to therapy. However, the routine use of these
parameters, either in single use or in models, is not absolutely mandated in the care of the patients with CLL. In
addition, prognostic parameters should include the recognition of other influences on the course of all patients with
CLL, such as advanced age and poor biologic fitness (defined
as impaired physical fitness and organ function), which
considerably increase the risk for adverse consequences of
progressive disease and contribute to the decreased survival
for patients with CLL compared with the age-matched
population.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Neil E. Kay
Research
Funding
Expert
Testimony
Other
Remuneration
Genentech;
Hospira
REFERENCES
1. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2008. CA Cancer
J Clin. 2008;58(2):71-96. Epub 2008 Feb 20.
2. Dighiero G, Maloum K, Desablens B, et al. Chlorambucil in indolent
chronic lymphocytic leukemia. French Cooperative Group on Chronic Lymphocytic Leukemia. N Engl J Med. 1998;338(21):1506-1514.
3. Shanafelt TD, Bowen D, Venkat C, et al. Quality of life in chronic
lymphocytic leukemia: An international survey of 1482 patients. Br J Haematol. 2007;139(2):255-264.
4. Shanafelt TD, Bowen DA, Venkat C, et al. The physician-patient
relationship and quality of life: Lessons from chronic lymphocytic leukemia.
Leuk Res. 2009;33(2):263-270. Epub 2008 Jul 25.
5. Rozman C, Bosch F, Montserrat E. Chronic lymphocytic leukemia: A
changing natural history? Leukemia. 1997;11(6):775-778.
6. Zent CS, Kay NE. Management of patients with chronic lymphocytic
leukemia with a high risk of adverse outcome: The Mayo Clinic approach.
Leuk Lymphoma. 2011;52(8):1425-1434. Epub 2011 Jun 8.
7. Shanafelt TD, Rabe KG, Kay NE, et al. Age at diagnosis and the utility
of prognostic testing in patients with chronic lymphocytic leukemia. Cancer.
2010;116(20):4777-4787.
8. Molica S, Alberti A. Prognostic value of the lymphocyte doubling time in
chronic lymphocytic leukemia. Cancer. 1987;60(11):2712-2716.
9. Bergmann MA, Eichhorst BF, Busch R, et al. Prospective Evaluation of
Prognostic Parameters in Early Stage Chronic Lymphocytic Leukemia (CLL):
Results of the CLL1-Protocol of the German CLL Study Group (GCLLSG).

ASH Annual Meeting Abstracts. 2007;110(11):625.
10. Wierda WG, OBrien S, Wang X, et al. Prognostic nomogram and index
for overall survival in previously untreated patients with chronic lymphocytic
leukemia. Blood. 2007;109(11):4679-4685. Epub 2007 Feb 13.
11. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38
expression as novel prognostic indicators in chronic lymphocytic leukemia.
Blood. 1999;94(6):1840-1847.
12. Hamblin TJ, Davis Z, Gardiner A, et al. Unmutated Ig V(H) genes are
associated with a more aggressive form of chronic lymphocytic leukemia.
Blood. 1999;94(6):1848-1854.
13. Hamblin TJ, Orchard JA, Ibbotson RE, et al. CD38 expression and
immunoglobulin variable region mutations are independent prognostic variables in chronic lymphocytic leukemia, but CD38 expression may vary during
the course of the disease. Blood. 2002;99(3):1023-1029.
14. Dohner H, Stilgenbauer S, Benner A, et al. Genomic aberrations and
survival in chronic lymphocytic leukemia. N Engl J Med. 2000;343(26):19101916. Epub 2008 Mar 26.
15. Krober A, Seiler T, Benner A, et al. V(H) mutation status, CD38
expression level, genomic aberrations, and survival in chronic lymphocytic
leukemia. Blood. 2002;100(4):1410-1416.
16. Crespo M, Bosch F, Villamor N, et al. ZAP-70 expression as a surrogate
for immunoglobulin-variable-region mutations in chronic lymphocytic leukemia. N Engl J Med. 2003;348(18):1764-1775.
397
NEIL E. KAY
17. Orchard JA, Ibbotson RE, Davis Z, et al. ZAP-70 expression and
prognosis in chronic lymphocytic leukaemia. Lancet. 2004;363(9403):105-111.
18. Rassenti LZ, Huynh L, Toy TL, et al. ZAP-70 compared with immunoglobulin heavy-chain gene mutation status as a predictor of disease progression in chronic lymphocytic leukemia. N Engl J Med. 2004;351(9):893-901.
19. Rassenti LZ, Jain S, Keating MJ, et al. Relative value of ZAP-70, CD38,
and immunoglobulin mutation status in predicting aggressive disease in
chronic lymphocytic leukemia. Blood. 2008;112(5):1923-1930. Epub 2008 Jun
24.
20. Del Principe MI, Del Poeta G, Buccisano F, et al. Clinical significance of
ZAP-70 protein expression in B-cell chronic lymphocytic leukemia. Blood.
2006;108(3):853-861. Epub 2006 Apr 6.
21. Shanafelt TD, Witzig TE, Fink SR, et al. Prospective evaluation of
clonal evolution during long-term follow-up of patients with untreated earlystage chronic lymphocytic leukemia. J Clin Oncol. 2006;24(28):4634-4641.
22. Dal Bo M, Rossi FM, Rossi D, et al. 13q14 deletion size and number of
deleted cells both influence prognosis in chronic lymphocytic leukemia. Genes
Chromosomes Cancer. 2011;50(8):633-643. Epub 2011 May 11.
23. Zenz T, Krober A, Scherer K, et al. Monoallelic TP53 inactivation is
associated with poor prognosis in chronic lymphocytic leukemia: Results from
a detailed genetic characterization with long-term follow-up. Blood. 2008;
112(8):3322-3329. Epub 2008 Aug 8.
24. Gonzalez D, Martinez P, Wade R, et al. Mutational Status of the TP53
398
Gene as a Predictor of Response and Survival in CLL Patients with and

without 17p Deletion. ASH Annual Meeting Abstracts. 2008;112(11):784.
25. Shanafelt TD, Jenkins G, Call TG, et al. Validation of a new prognostic
index for patients with chronic lymphocytic leukemia. Cancer. 2009;115(2):
363-372.
26. Gonzalez Rodriguez AP, Gonzalez Garcia E, Fernandez Alvarez C, et al.
B-chronic lymphocytic leukemia: Epidemiological study and comparison of
MDACC and GIMENA pronostic indexes. Med Clin (Barcelona). 2009;133(5):
161-166. Epub 2009 Jun 18.
27. Tam CS, Shanafelt TD, Wierda WG, et al. De novo deletion 17p13.1
chronic lymphocytic leukemia shows significant clinical heterogeneity:
The M. D. Anderson and Mayo Clinic experience. Blood. 2009;114(5):957-964.
Epub 2009 May 4.
28. Wierda WG, OBrien S, Wang X, et al. Multivariable model for time to
first treatment in patients with chronic lymphocytic leukemia. J Clin Oncol.
2011;29(31):4088-4095. Epub 2011 Oct 3.
29. Wierda WG, OBrien S, Wang X, et al. Characteristics associated with
important clinical end points in patients with chronic lymphocytic leukemia
at initial treatment. J Clin Oncol. 2009;27(10):1637-1643. Epub 2009 Feb 17.
30. Pepper C, Majid A, Lin TT, et al. Defining the prognosis of early stage
chronic lymphocytic leukaemia patients. Br J Haematol. 2012;156(4):499-507.
Epub 2011 Dec 15.
Transplant in Chronic Lymphocytic

Leukemia: To Do It or Not and If So, When
and How?
By John G. Gribben, MD, DSc
Overview: Most patients with chronic lymphocytic leukemia

(CLL) have an indolent clinical course, but the disease remains
incurable with standard therapy and the prognosis is dismal
for those patients with disease refractory to available treatment options. The only potentially curative treatment is allogeneic hematopoietic stem cell transplantation (SCT), but
since CLL is a disease of elderly patients, few patients are
candidates for myeloablative allogeneic SCT. Although autologous SCT is feasible and has low treatment-related mortality,
it is not curative. The widespread adoption of reducedintensity conditioning (RIC) allogeneic SCT has made this
approach applicable to the elderly patient population with
CLL. This approach relies on the documented graft-versus-
CT IS NOT a suitable treatment option for the majority

of patients with CLL in whom the disease follows an
indolent course, and many patients never require therapy.
The outcome has improved dramatically over the past decade for patients whose disease progresses to require treatment.1 Most patients with CLL are elderly and not
sufficiently fit for SCT. However, it is possible to identify
suitable candidates for SCT using a number of clinical and
biologic features.2 The role of SCT in a number of other
hematologic malignancies has been established in prospective studies, but no studies in CLL have compared the
outcome after standard chemotherapy with allogeneic SCT.
The biggest challenges remain the decision of which patients
are eligible for SCT and when in their disease course SCT
should occur.
Patient Selection for SCT
CLL is an extremely heterogeneous disease, with the

clinical course varying from patients who never require
therapy to patients with rapidly progressive and fatal malignancy. Treatment guidelines state that therapy should be
reserved for those with advanced, symptomatic, or progressive disease.3 For those patients whose disease requires
therapy, the results of randomized clinical trials have demonstrated significant improvement over the past decade
with the use of combination chemotherapy and now chemoimmunotherapy.4-9 When assessing the potential role of
transplantation, these approaches must be considered in
addition to the many exciting novel agents currently in
clinical trials, which may alter our approach as to when and
to whom transplant should be offered.
Major advances have been made in our understanding
of CLL pathophysiology, which has led to the emergence of
a large number of prognostic biomarkers cytogenetics, immunoglobulin heavy chain (IgVH) gene mutational status,
zeta-associated protein 70 (ZAP70) expression, and CD38
expression.10-13 The emergence of prognostic biomarkers
has implications for the selection for which patients might
merit SCT, but it is not yet fully clear how we should use
these factors in CLL management.14
leukemia (GVL) effect and is strong in CLL. Steps to further

decrease the morbidity and mortality of the RIC SCT and in
particular to reduce the incidence of chronic extensive graftversus-host disease (GVHD) remain a major focus. Many
potential treatments are available for CLL, and appropriate
patient selection and SCT timing remain controversial and the
focus of ongoing clinical trials. The use of SCT must always
be weighed against the risk of the underlying disease, particularly in a setting where improvements in treatment are leading to improved outcome. The major challenge remains how
to identify which patients with CLL merit this approach and
where in the treatment course this treatment can be applied
optimally.
Autologous SCT
The role of autologous SCT in CLL remains highly controversial and there is currently no role for autologous SCT for
CLL except in the setting of a clinical trial. A number of
phase II studies have reported outcome following autologous
SCT for CLL, demonstrating that this approach is feasible
with a transplant-related mortality (TRM) of 1% to 10%,
with most toxicity occurring late.15-17 Such studies need to
be considered in terms of intention to treat, and this can be
difficult to determine in transplant centers where only
responding might be referred. In a pilot study to assess the
feasibility of performing autologous SCT, 115 previously
untreated patients with CLL prospectively enrolled, and
only 65 (56%) proceeded to transplant.16 TRM was low,
complete remission (CR) rate was 74%, the 5-year estimated
overall survival (OS) was 77.5%, and progression-free survival (PFS) was 51.5%. Of concern, 8% of patients developed
post-transplant acute myeloid leukemia/myelodysplastic
syndrome, a complication also seen in other series.15 In a
single-center study, among 137 patients who underwent
autologous transplantation, the one-year TRM was 4% but
rose to 10% when late events were taken into account. At the
median follow-up time of 6.5 years, OS was 58% after
autologous SCT. There was no TRM among 72 patients who
underwent autologous SCT in five Finnish centers.17 Initial
enthusiasm for autologous SCT has been tempered since the
observed results demonstrated no plateau in either eventfree survival (EFS) or OS and because of concerns regarding
the risk of secondary malignancies.18
A retrospective matched-pair analysis was performed including 66 patients who had undergone a uniform high-dose
therapy and autologous SCT with a database of 291 patients
treated conventionally and suggested a survival advantage
From the Barts Cancer Institute, Queen Mary University of London, Charterhouse
Square, London.
Address reprint requests John G. Gribben, MD, DSc, Barts Cancer Institute, Queen Mary
University of London, Charterhouse Square, London EC1M 6BQ, United Kingdom; email:
j.gribben@qmul.ac.uk.
1092-9118/10/1-10
399
JOHN G. GRIBBEN
19
for autologous SCT over conventional therapy. Survival

was significantly longer for patients who had undergone
autologous SCT compared with patients conventionally
treated.
Results have now been reported for three phase III randomized trials examining the role of autologous transplant
in CLL.20-22 The European Intergroup Study randomly
selected 223 patients, 83% after first-line therapy and 17%
after second-line therapy. Only 59% of patients were in CR.
Patients were randomly selected to receive autologous SCT
(112 patients) and observation (111 patients). Autologous
SCT significantly improved EFS from 24.4 months in the
observation group to 51.2 months in the patients receiving
autologous SCT, but there was no difference in OS at 5
years.20 In a second study, 241 patients received three
courses of cyclophosphamide, hydroxydaunorubicin, vincristine, and prednisone/prednisolone (mini-CHOP), and then
three courses of fludarabine. Patients in CR were then
randomly selected to receive autologous SCT or observation,
whereas patients not in CR were randomly selected to
receive salvage therapy followed by either autologous SCT
or three courses of fludarabine plus cyclophosphamide (FC).
The patients achieving CR were also included for analysis in
the European Intergroup Study. Autologous SCT improved
EFS in patients achieving CR to initial therapy, but no
differences were observed among patients who required
salvage therapy. No difference in OS was found in either
group.21 The GOELAMS LLC 98 trial compared six monthly
courses of cyclophosphamide, doxorubicin hydrochloride,
vincristine, and prednisone (CHOP) followed by six CHOP
courses every 3 months in responding patients with highdose therapy with autologous SCT used as consolidation in
responding patients after three CHOP courses. A total of 86
patients were enrolled, with 43 patients evaluable after
autologous SCT. On an intent-to-treat basis and with a
median follow-up time of 77.1 months, median PFS was 22
months after conventional therapy and 53 months after
autologous SCT (p 0.0001). There was no difference in OS
between the two arms.25 Concerns regarding this trial
include the relatively small number of patients enrolled; the
trial had to be closed early because of concerns of emerging
better induction therapies, leading to a low statistical power.
The results obtained from these randomized studies all
demonstrate that patients with chemotherapy-sensitive disease had prolongation in EFS but not OS. The question
KEY POINTS
400
Autologous stem cell transplantation (SCT) is not

curative for chronic lymphocytic leukemia (CLL) and
should be performed only in the setting of clinical
trials.
Myeloablative allogeneic SCT is applicable only to a
small population and is associated with unacceptably
high treatment-related morbidity and mortality.
Reduced-intensity conditioning allogeneic SCT is potentially curative but complicated by chronic GVHD.
Identifying patients with CLL who should receive
transplantation and when in their clinical course this
should occur remains a major challenge.
Table 1. Outcome after FCR or ASCT

Study
FCR
M. D. Anderson8
CLL89
Autologous SCT
European Intergroup20
SFGM-TC/GFLLC21
GOELAMS22
Number of
Patients
Median
Age (y)
EFS
(%)
OS
(%)
300
408
57
61
51 at 5 y
65 at 3 y
77 at 6 y
87 at 3 y
112
52 in CR
46 not CR
43
54
56
42 at 5 y
79.8 at 3 y
48.9 at 3 y
PFS 53 at
6.5 y
86 at 5 y
96 at 3 y
82 at 3 y
107.4 at 6.5 y
Upper age
limit 60
Abbreviations: FCR, fludarabine, cyclophosphamide, and rituximab; ASCT,

autologous stem cell transplantation; y, years; EFS, event-free survival; OS,
overall survival; SFGM-TC, Societe Francaise de Greffe de Moelle et de Therapie
Cellulaire; GFLLC, Groupe Franais dtude de la Leucmie Lymphode Chronique; CR, complete remission; GOELAMS, Groupe Ouest Est dEtude des
Leucmies et Autres Maladies du Sang.
remains if these patients would have had similar results if

they had been offered the type of chemoimmunotherapy that
is the standard of care today. Whereas great caution must be
taken comparing the outcome of separate studies, the outcomes in terms of EFS and OS might be similar for patients
treated with fludarabine, cyclophosphamide, and rituximab
(FCR) and those receiving autologous SCT (Table 1). This
does not address whether improved outcome would be seen
with autologous SCT after FCR. In a phase II study, those
patients who received autologous SCT after front-line FCR
had an inferior outcome to patients who underwent autologous SCT before chemoimmunotherapy was available.15
Monoclonal antibodies have been used to increase the
likelihood of elimination of minimal residual disease (MRD)
after autologous SCT, ex vivo or by in vivo treatment with
alemtuzumab or rituximab.15 Alemtuzumab was used in the
conditioning regimen for autologous SCT in one arm of the
German CLL Study Group CLL3 trial, and 12 of 16 patients
(87%) developed a skin rash between 43 and 601 days
post-SCT. In seven patients, biopsy confirmed GVHD that
persisted for a median duration of 517 (range, 60 to 867)
days.23 The trial was discontinued because of the TRM, but
addition of alemtuzumab led to improved disease control.
When alemtuzumab was used at modified dose (10 mg
subcutaneously three times per week for 6 weeks) in 34
patients who had a clinical response to a fludarabine-based
regimen, the CR rate improved from 35% to 79.5% with 56%
achieving eradication of MRD.24 Peripheral blood stem cell
collection was subsequently successfully performed in 92%.
Most studies reported relatively short follow-up and therefore focus mostly on TRM early post-SCT, but late consequencesparticularly
development
of
secondary
myelodysplasia and acute myeloid leukemia (MDS/AML)
are of concern. Among 65 previously untreated patients who
were treated with fludarabine followed by autologous SCT,
eight developed MDS/AML, with a 5-year actuarial risk of
12% developing MDS/AML after autologous SCT.16 Longterm follow-up reports a high incidence of other solid tumors
in 31 (19%) patients.15
Myeloablative Allogeneic SCT
The major advantage of allogeneic SCT is the potential

for a GVL effect. There is strong evidence for a GVL effect in
CLL as demonstrated by a decreased risk of relapse in
patients with chronic GVHD, increased risk of relapse in
WHEN TO OFFER TRANSPLANTATION IN CLL

Table 2. RIC Allogeneic SCT for CLL
Number of
Patients
82
77
46
41
39
30
Age
Years
(range)
Prior
Regimens
(range)
Chemorefractory
(%)
Prior
Auto-SCT
82
(4272)
54
(3066)
53
(3567)
54
(3767)
57
(3470)
50
(1263)
87%
3
(08)
5
(110)
3
(18)
3
(28)
3
(08)
33%
10
57%
10
27%
11
Not stated
47%
GVHD
Donor
(includes
mismatch)
TRM
Acute
Grade 24
Chronic
Extensive
63% related
37% unrelated
81% related
25% overall
55%
18% 12 m
34%
49% related
53% unrelated
58%
33%
67%
58%
42%
90%
10%
50%
50%
17% overall
34%
43%
10%
(grade 34)
45%
33%*
*after DLI
58%
56%
21%
related
unrelated
related
unrelated
related
unrelated
related
unrelated
5% at 100 d
26% overall
2% at 100 d
13% overall
Survival
OS 50% 5 yr
PFS 45%
OS 72% 2 yr
PFS 56%
OS 54% 2 yr
PFS 34%
OS 51 2 yr
PFS 45%
OS 48% 4 yr
PFS 44%
OS 72% 2 yr
PFS 67%
Reference
Sorror et al 2008
Dreger et al 2003
Brown et al 2006
28
29
30
Delgado et al 2006
Khouri et al 2006
31
32
Schetelig et al 2003
33
Abbreviations: RIC, reduced-intensity conditioning; SCT, stem cell transplantation; CLL, chronic lymphocytic leukemia; TRM, transplant-related mortality; GVHD,
graft-versus-host disease; OS, overall survival; PFS, progression-free survival; m, months; d, days.
patients who have undergone T-cell depletion, and clinical

responses to removal of immune suppression or to donor
lymphocyte infusion (DLI).15 Allogeneic SCT has significant
morbidity and TRM, from regimen-related toxicity, GVHD,
and infection, but surviving patients have long-term disease
control.15,25-27 In registry data, TRM following allogeneic
SCT in patients with CLL was unacceptably high at 46%,
with mortality from GVHD of 20%.28 Currently there is only
a very limited role for myeloablative allogeneic SCT in the
setting of very young patients with particularly aggressive
disease. Among 25 patients with CLL who underwent allogeneic SCT at the Fred Hutchinson Cancer Research Center
(FHCRC), grades 2 through 4 acute GVHD was seen in 14
patients, 10 developed clinically extensive chronic GVHD,
and estimated OS at 5 years was 32%.29
No randomized studies have compared the outcome of
autologous SCT with allogeneic SCT. Studies from University of Texas M. D. Anderson Cancer Center (MDACC)
demonstrate improved outcome after allogeneic SCT compared with autologous SCT, suggesting that allogeneic SCT
can induce durable remission even in patients with refractory disease.30 At Dana-Farber Cancer Institute, 162 patients with high-risk CLL were enrolled in a biologic
randomization in which 25 patients with a human leukocyte antigen (HLA)-matched sibling donor underwent T
cell depleted myeloablative allogeneic SCT, while 137 with
no HLA-matched sibling donor underwent B cellpurged
autologous SCT, with both groups receiving identical conditioning regimen using high-dose cyclophosphamide and total
body irradiation.15 The 100-day TRM was 4% after autologous or allogeneic SCT, but later TRM had a major effect
on outcome. At the median follow-up of 6.5 years, PFS
was significantly longer following autologous than T cell
depleted allogeneic SCT, but no significant differences were
observed in disease recurrence or deaths without recurrence
by type of transplant. There was no difference in OS between
the two groups, and at the median follow-up time of 6.5
years, OS was 58% after autologous and 55% after allogeneic
SCT.
RIC SCT for CLL
A major advance in reducing the short-term morbidity and

mortality of allogeneic SCT has been the introduction of
nonmyeloablative or RIC regimens to allow engraftment of

allogeneic stem cells. This approach is much more applicable
to the age group with CLL who are potential candidates for
SCT. Most patients reported have been treated on experimental treatment protocols with enrollment of many patients with chemo-refractory end-stage disease.
RIC regimens allow transplantation in older patients,
making this approach more applicable to increased numbers
of patients with CLL and results from the larger reported
studies are shown in Table 2.29-34 Most patients were
heavily pretreated and refractory to therapy, but despite
these issues, the majority demonstrated donor engraftment
and had a high CR rate. The ability of such approaches to
eradicate MRD in patients with advanced CLL and the
observation of late remissions in patients treated with low
doses of chemotherapy provide direct evidence for a powerful
GVL in CLL.35 The outcome from the FHCRC multiinstitutional protocol after RIC allogeneic SCT was reported
for 82 patients with advanced fludarabine-refractory CLL
using related (52 patients) or unrelated donors (30 patients)
median age 56 (range, 42 to 72).29 TRM was 23% at 5 years,
with significant GVHD a remaining problem. Five-year OS
was 50% and EFS was 39%. Among 46 patients who underwent RIC transplantation at Dana-Farber Cancer Institute 67% using unrelated donorsfactors associated with
increased risk of relapse include low levels of donor chimerism at day 30, chemotherapy-refractory disease, increased
number of previous therapies, and adverse cytogenetics.31
No formal assessment of RIC compared with myeloablative allogeneic SCT has been undertaken, but the outcome
after RIC allogeneic SCT of 73 patients who had undergone
RIC was compared with that of 82 matched patients who
had undergone standard myeloablative conditioning for CLL
from the European Blood and Marrow Transplantation
(EBMT) registry database during the same time period.
Patients undergoing RIC transplants had significantly reduced TRM but higher relapse incidence, and there was no
significant difference in OS or PFS between these two
groups.36 Of particular interest is the report of 44 patients
with CLL with deletion of 17p and loss of p53 in whom
allogeneic SCT has the potential to induce long-term remission in these very high-risk patients.37
401
JOHN G. GRIBBEN
Addition of Monoclonal Antibodies to RIC SCT
GVHD remains the major concern after RIC SCT and

attempts have been made to utilize monoclonal antibodies to
reduce the incidence of GVHD without increasing the subsequent risk of relapse. Excellent results have been obtained
at MDACC using RIC based on a combination of fludarabine
and cyclophosphamide with the addition of rituximab, an
approach designed to maximize GVL by early tapering of
immune suppression with use of rituximab and DLI. Among
39 patients treated, median age 57 (range, 34 to 70), median
time from diagnosis to transplantation was 4.5 years.33 All
patients had recurrent advanced disease, were heavily pretreated with a median of three (range, 2 to 8) chemotherapy
regimens, and had been previously treated with fludarabine/
rituximab-based regimens. At transplant, 34 patients (87%)
had active disease, including nine (23%) with evidence of
Richters transformation. In this series, only four of the
donors were unrelated. Fourteen patients required immunomodulation with rituximab and DLI for persistent disease
after SCT. Only one patient died early, and among the 38
evaluable patients, 27 (71%) achieved CR with a 48% estimated OS at 4 years and current PFS of 44%. Acute grade 2
through 4 GVHD was observed in 45%, but chronic extensive
GVHD was reduced without concomitant increased risk of
relapse.
GVHD can be decreased using alemtuzumab in the conditioning regimen to reduce donor lymphocytes, but this is
associated with delayed immune reconstitution, increases
the risk of infective complications, and appears to impair
GVL. In 41 consecutive patients with CLL (24 HLA-matched
sibling donors and 17 unrelated volunteer donors, including
4 mismatched) treated with the conditioning regimen alemtuzumab with fludarabine and melphalan had impressive
antitumor effects with 100% of patients with chemotherapysensitive disease and 86% with chemotherapy-refractory
disease responding.32 The TRM rate was 26%, OS was 51%,
and relapse risk was 29% at 2 years. GVHD rates were
relatively low with acute GVHD occurring in 17 (41%) and
chronic GVHD in 13 (33%). The unexpectedly high TRM
rate was because of a high incidence of fungal and viral
infections.
How to Select Who and When to Offer
Allogeneic SCT
All studies of SCT have enrolled younger patients with

high-risk disease. This term is very loosely defined and it is
difficult to determine precisely the risk factors used in each
of the reported studies. EBMT guidelines have been established outlining indications for SCT in CLL, which conclude
that there is a evidence base for the efficacy of allogeneic
SCT in CLL and that this procedure is indicated in patients
with high-risk CLL.38 Patients at high risk are defined in
Table 3 and include those requiring treatment who have
TP53 abnormalities (who merit allogeneic SCT in first
response), patients who fail to achieve CR, who progress
within 12 months after fludarabine, who relapse within 24
months after having achieved a response with combination
therapy, those who have relapsed after prior autologous
SCT, or those patients who are fludarabine refractory. It
should be noted that the only category that requires assessment of biologic risk is detection of TP53 abnormalities.
Ongoing prospective clinical studies will determine the ef-
402
Table 3. EBMT Guidelines for Transplantation in CLL 38

Allo-SCT is a reasonable treatment option in poor-risk CLL including:
Fludarabine resistancenonresponse or early relapse ( 12 months) after
purine analogue-based therapy
Relapse 24 months after purine analogue combinations or auto-SCT (plus
high-risk genetics)
p53 mutation with treatment indication
Auto-SCT indicated in clinical trial only
Abbreviation: EBMT, European Blood and Marrow Transplantation; Allo-SCT,
allogeneic stem cell transplantation; CLL, chronic lymphocytic leukemia; autoSCT, autologous SCT.
fect of biomarkers including IgVH mutational status and

other cytogenetic abnormalities in identification of patients
at sufficiently high risk to merit use of allogeneic SCT in first
CR.
The definition of refractory CLL is of particular importance and the terms refractory CLL and fludarabinerefractory CLL are often used interchangeably. But with
the large number of treatment options, the consideration of
treatment context is important. The type of therapy to which
patients fail to respond and previous therapies received are
of major importance. The clinical importance of refractory
CLL is based on the fact that these patients have very poor
prognosis (median 12 years OS in most studies) despite
various and intense salvage therapy strategies.2,39-40 Most
trials of investigational agents use this definition as an
entry point for early drug development. The most recent
CLL guidelines define refractory CLL as treatment failure
(less than partial remission [PR]) or disease progression
within 6 months of the last antileukemic therapy.3 These
still correspond to the definition coined when CLL treatment
was based on chlorambucil or fludarabine monotherapy. In
current standard practice, patients eligible to consider allogeneic SCT are most likely to have received combination
chemoimmunotherapy. There is accumulating evidence that
patients who are formally not refractory based on the current definition but relapse within 3 years after chemoimmunotherapy also have very poor outcome and are unlikely to
have durable responses to subsequent chemotherapy.2
The guidelines suggest three groups of patients who may
be offered therapy. The first group is those with TP53 loss,
and these are the patients who have the poorest response
and shortest duration of remission. Allogeneic SCT is considered appropriate in first response after initial therapy for
these patients. Those patients who failed to achieve CR to
FCR and/or have very short ( 24 months) response to prior
FCR can be added to this very high-risk group. These
patients are prime candidates for drugs with proven activity
in TP53 deleted/mutant cells, investigational agents in clinical trials, and then for allogeneic SCT if they respond.
A second scenario of high-risk CLL is identification of
subgroups destined to relapse relatively early after standard
treatment and it is in this group that biomarker analysis
will prove useful. Candidates within this group include
patients with high beta-2-microglobulin or thymidine kinase, unmutated immunoglobulin heavy chain variable region (IgVH) or 11q deletion.9,41 These patients are the ones
who gained most by the addition of rituximab to FC.9
Biomarker analysis will also be useful to identify those
patients in whom allogeneic SCT should not be offered (no
11q deletion, no TP53 deletion/mutation, mutated IGHV,
low beta-2-MG, no prior therapy) who have very favorable
outcome despite progressing to indication for treatment.
WHEN TO OFFER TRANSPLANTATION IN CLL
The results of phase II studies of RIC allogeneic SCT

suggest that whereas patients with refractory disease may
respond to the GVL effect, optimal outcome will be achieved
by consideration of transplant for those patients with highrisk disease before they become truly refractory and those
patients in whom minimal disease state cannot be achieved.
We have demonstrated that CLL cells are inherently immunosuppressive even to allogeneic donor T cells, and this may
explain why better results are obtained when patients
receive RIC allogeneic SCT with low tumor bulk, which can
only be achieved if patients are offered this approach before
truly refractory disease occurs.42
Conclusion
SCT has a role to play in selected patients with CLL, with

major focus now on the use of RIC allogeneic SCT. Future
approaches to the management of this disease must take

into account the balance between the increased morbidity
and mortality of SCT in CLL with the curative potential that
these approaches potentially offer, in the setting of the
outcome improvements that can now be seen using chemoimmunotherapy. Although RIC allogeneic SCT results in
high response rates and eradication of polymerase chain
reaction detectable MRD and is potentially curative, the
follow-up of most clinical trials remains too short to assess
whether SCT can really cure CLL. In the absence of any
other treatment modalities currently capable of improving
outcome in this disease, SCT should be considered as a
treatment approach for younger patients with high-risk CLL
early in the course of the disease, ideally in the setting of
well-designed clinical trials assessing the treatments effect
on outcome in these patients. Several such trials are underway.
Author
John G. Gribben
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Celgene; Merck
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Mundipharma;
Roche
REFERENCES
1. Gribben JG, OBrien S. Update on therapy of chronic lymphocytic
leukemia. J Clin Oncol. 2011;29:544-550.
2. Stilgenbauer S, Zenz T. Understanding and managing ultra high-risk
chronic lymphocytic leukemia. Hematology Am Soc Hematol Educ Program.
2010;2010:481-488.
3. Hallek M, Cheson BD, Catovsky D, et al. Guidelines for the diagnosis
and treatment of chronic lymphocytic leukemia: a report from the International Workshop on Chronic Lymphocytic Leukemia updating the National
Cancer Institute-Working Group 1996 guidelines. Blood. 2008;111:5446-5456.
4. Rai KR, Peterson BL, Appelbaum FR, et al. Fludarabine compared with
chlorambucil as primary therapy for chronic lymphocytic leukemia. N Engl
J Med. 2000;343:1750-1757.
5. Eichhorst BF, Busch R, Hopfinger G, et al. Fludarabine plus cyclophosphamide versus fludarabine alone in first-line therapy of younger patients
with chronic lymphocytic leukemia. Blood. 2006;107:885-891.
6. Flinn IW, Neuberg DS, Grever MR, et al. Phase III trial of fludarabine
plus cyclophosphamide compared with fludarabine for patients with previously untreated chronic lymphocytic leukemia: US Intergroup Trial E2997.
J Clin Oncol. 2007;25:793-798.
7. Catovsky D, Richards S, Matutes E, et al. Assessment of fludarabine
plus cyclophosphamide for patients with chronic lymphocytic leukaemia (the
LRF CLL4 Trial): a randomised controlled trial. Lancet. 2007;370:230-239.
8. Tam CS, OBrien S, Wierda W, et al. Long-term results of the fludarabine, cyclophosphamide, and rituximab regimen as initial therapy of chronic
lymphocytic leukemia. Blood. 2008;112:975-980.
9. Hallek M, Fischer K, Fingerle-Rowson G, et al. Addition of rituximab
to fludarabine and cyclophosphamide in patients with chronic lymphocytic
leukaemia: a randomised, open-label, phase 3 trial. Lancet. 2010;376:11641174.
10. Dohner H, Stilgenbauer S, Benner A, et al. Genomic aberrations and
survival in chronic lymphocytic leukemia. N Engl J Med. 2000;343:1910-1916.
11. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38
expression as novel prognostic indicators in chronic lymphocytic leukemia.
Blood. 1999;94:1840-1847.
12. Rassenti LZ, Huynh L, Toy TL, et al. ZAP-70 compared with immunoglobulin heavy-chain gene mutation status as a predictor of disease progression in chronic lymphocytic leukemia. N Engl J Med. 2004;351:893-901.
13. Rassenti LZ, Jain S, Keating MJ, et al. Relative value of ZAP-70, CD38,
and immunoglobulin mutation status in predicting aggressive disease in
chronic lymphocytic leukemia. Blood. 2008;112:1923-1930.
14. Gribben JG. Are prognostic factors in CLL overrated? Oncology (Williston Park). 2011;25:703, 706.
15. Gribben JG, Zahrieh D, Stephans K, et al. Autologous and allogeneic
stem cell transplantation for poor-risk chronic lymphocytic leukemia. Blood.

2005;106:4389-4396.
16. Milligan DW, Fernandes S, Dasgupta R, et al. Results of the MRC pilot
study show autografting for younger patients with chronic lymphocytic
leukemia is safe and achieves a high percentage of molecular responses.
Blood. 2005;105:397-404.
17. Jantunen E, Itala M, Siitonen T, et al. Autologous stem cell transplantation in patients with chronic lymphocytic leukaemia: the Finnish experience. Bone Marrow Transplant. 2006;37:1093-1098.
18. Montserrat E, Gribben JG. Autografting CLL: the game is over! Blood
2011;117:6057-6058.
19. Dreger P, Stilgenbauer S, Benner A, et al. The prognostic impact of
autologous stem cell transplantation in patients with chronic lymphocytic
leukemia: a risk-matched analysis based on the VH gene mutational status.
Blood. 2004;103:2850-2858.
20. Michallet M, Dreger P, Sutton L, et al. Autologous hematopoietic stem
cell transplantation in chronic lymphocytic leukemia: Results of European
intergroup randomized trial comparing autografting versus observation.
Blood. 2011;117:1516-1521.
21. Sutton L, Chevret S, Tournilhac O, et al. Autologous stem cell transplantation as a first-line treatment strategy for chronic lymphocytic leukemia: a multicenter, randomized, controlled trial from the SFGM-TC and
GFLLC. Blood. 2011;117:6109-6119.
22. Brion A, Mahe B, Kolb B, et al. Autologous transplantation in CLL
patients with B and C Binet stages: Final results of the prospective randomized GOELAMS LLC 98 trial. Bone Marrow Transplant. Epub 2011 Jul 4.
23. Zenz T, Ritgen M, Dreger P, et al. Autologous graft-versus-host diseaselike syndrome after an alemtuzumab-containing conditioning regimen and
autologous stem cell transplantation for chronic lymphocytic leukemia. Blood.
2006;108:2127-2130.
24. Montillo M, Tedeschi A, Miqueleiz S, et al. Alemtuzumab as consolidation after a response to fludarabine is effective in purging residual disease in
patients with chronic lymphocytic leukemia. J Clin Oncol. 2006;24:2337-2342.
25. Michallet M, Archimbaud E, Bandini G, et al. HLA-identical sibling
bone marrow transplantation in younger patients with chronic lymphocytic
leukemia. European Group for Blood and Marrow Transplantation and the
International Bone Marrow Transplant Registry. Ann Internal Med. 1996;
124:311-315.
26. Khouri I, Champlin R. Allogenic bone marrow transplantation in
chronic lymphocytic leukemia. Ann Intern Med. 1996;125:780-787.
27. Doney KC, Chauncey T, Appelbaum FR. Allogeneic related donor
hematopoietic stem cell transplantation for treatment of chronic lymphocytic
leukemia. Bone Marrow Transplant. 2002;29:817-823.
403
JOHN G. GRIBBEN
28. Khouri IF, Przepiorka D, van Besien K, et al. Allogeneic blood or
marrow transplantation for chronic lymphocytic leukaemia: timing of transplantation and potential effect of fludarabine on acute graft-versus-host
disease. Br J Haematol. 1997;97:466-473.
29. Sorror ML, Storer BE, Sandmaier BM, et al. Five-year follow-up of
patients with advanced chronic lymphocytic leukemia treated with allogeneic
hematopoietic cell transplantation after nonmyeloablative conditioning.
J Clin Oncol. 2008;26:4912-4920.
30. Dreger P, Brand R, Hansz J, et al. Treatment-related mortality and
graft-versus-leukemia activity after allogeneic stem cell transplantation for
chronic lymphocytic leukemia using intensity-reduced conditioning. Leukemia. 2003;17:841-848.
31. Brown JR, Kim HT, Li S, et al. Predictors of improved progression-free
survival after nonmyeloablative allogeneic stem cell transplantation for
advanced chronic lymphocytic leukemia. Biol Blood Marrow Transplant.
2006;12:1056-1064.
32. Delgado J, Thomson K, Russell N, et al. Results of alemtuzumab-based
reduced-intensity allogeneic transplantation for chronic lymphocytic leukemia: a British Society of Blood and Marrow Transplantation Study. Blood.
2006;107:1724-1730.
33. Khouri IF. Reduced-intensity regimens in allogeneic stem-cell transplantation for non-hodgkin lymphoma and chronic lymphocytic leukemia.
Hematology Am Soc Hematol Educ Program. 2006:390-397.
34. Schetelig J, Thiede C, Bornhauser M, et al. Evidence of a graft-versusleukemia effect in chronic lymphocytic leukemia after reduced-intensity
404
conditioning and allogeneic stem-cell transplantation: the Cooperative German Transplant Study Group. J Clin Oncol. 2003;21:2747-2753.
35. Moreno C, Villamor N, Colomer D, et al. Clinical significance of
minimal residual disease, as assessed by different techniques, after stem cell
transplantation for chronic lymphocytic leukemia. Blood. 2006;107:4563-4569.
36. Dreger P, Brand R, Milligan D, et al. Reduced-intensity conditioning
lowers treatment-related mortality of allogeneic stem cell transplantation
for chronic lymphocytic leukemia: a population-matched analysis. Leukemia.
2005;19:1029-1033.
37. Schetelig J, van Biezen A, Brand R, et al. Allogeneic hematopoietic
stem-cell transplantation for chronic lymphocytic leukemia with 17p deletion:
a retrospective European Group for Blood and Marrow Transplantation
analysis. J Clin Oncol. 2008;26:5094-5100.
38. Dreger P, Corradini P, Kimby E, et al. Indications for allogeneic stem
cell transplantation in chronic lymphocytic leukemia: the EBMT transplant
consensus. Leukemia. 2007;21:12-17.
39. Montserrat E, Moreno C, Esteve J, et al. How I treat refractory CLL.
Blood. 2006;107:1276-1283.
40. Tsimberidou AM, Keating MJ. Treatment of fludarabine-refractory
chronic lymphocytic leukemia. Cancer. 2009;115:2824-2836.
41. Oscier D, Wade R, Davis Z, et al. Prognostic factors identified three risk
groups in the LRF CLL4 trial, independent of treatment allocation. Haematologica. 2010;95:1705-1712.
42. Ramsay AG, Johnson AJ, Lee AM, et al. Chronic lymphocytic leukemia
T cells show impaired immunological synapse formation that can be reversed
with an immunomodulating drug. J Clin Invest. 2008;118:2427-2437.
NEW DEVELOPMENTS IN MYELOPROLIFERATIVE

NEOPLASMS
CHAIR
Srdan Verstovsek, MD, PhD
Houston, TX
SPEAKERS
Jason Gotlib, MD, MS
Stanford University School of Medicine
Stanford, CA
Francesco Passamonti, MD
Fondazione IRCCS Policlinico San Matteo
Pavia, Italy
Therapeutic Advances in Myeloproliferative

Neoplasms: The Role of New-Small
Molecule Inhibitors
By Srdan Verstovsek, MD, PhD
Overview: The discovery that a somatic point mutation

(JAK2V617F) in the Janus kinase 2 (JAK2) is highly prevalent in
patients with myeloproliferative neoplasms (MPNs) has been a
crucial breakthrough in our understanding of the underlying
molecular mechanisms of these diseases. Therefore, preclinical and clinical research in recent years has focused intensely
on the development of new therapies targeted to JAK2. These
efforts culminated in recent approval of ruxolitinib as the first
official therapy for patients with intermediate- or high-risk
myelofibrosis (MF). Therapy with JAK2 inhibitors substantially
improves quality of life and reduces organomegaly in MF with
or without JAKV617F mutation. Recent results suggest that
patients with advanced MF may live longer when receiving
RIMARY MYELOFIBROSIS (PMF) is a Philadelphia

chromosome (Ph)-negative myeloproliferative neoplasm
(MPN) characterized by diffuse bone marrow fibrosis and
osteosclerosis leading to bone marrow failure, extramedullary hematopoiesis, massive splenomegaly, very poor quality
of life as a result of debilitating MF-related systemic symptoms, weight loss, decrease in performance status, and an
increased risk of transformation to acute myeloid leukemia.1,2 Among the classic Ph-negative MPNs (the other two
are polycythemia vera [PV] and essential thrombocythemia
[ET]) PMF has the worse outcome, with a median survival of
approximately 5 to 7 years.1,2 MF can also develop secondary to disease transformation from PV and ET (then called
post-PV or -ET MF). The clinical course and outcome of
patients with primary and secondary MF appears to be
similar. MF presents a great burden to patients and a
challenge for their treating physicians, and is a focus of
intense clinical research to develop new effective therapies.1,2 Until recently, no medication has been approved as
therapy for MF, and our efforts in helping patients battling
the disease have been largely palliative.1,2
Starting with the discovery of the JAKV617F mutation in
approximately 50% of patients with MF in 2005,3 research to
find new treatment options for MF has expanded greatly
over recent years to include not just Janus kinase (JAK) 2
tyrosine kinase but also several other promising molecular
targets.2,4 These efforts led to a recent approval of ruxolitinib, an oral inhibitor of JAK1 and JAK2 tyrosine kinases, as therapy for patients with intermediate- and highrisk MF.5 Ruxolitinib therapy leads to a reduction in
splenomegaly, improvement in systemic MF-related symptoms and signs, and improved quality of life in patients with
advanced MF.6 Recent evidence suggests that ruxolitinib
may also prolong the life of patients with advanced MF.7
Clinical benefits are seen in patients with and without
JAK2V617F mutation (most common among many known
mutations in MPN), which reflects emerging notion that
dysregulated JAK/signal transducers and activators of
transcription (STAT) pathway is a common pathogenetic
abnormality in MPN.8 Ruxolitinib, being nonspecific for
JAK2V617F mutation, therefore, may benefit all patients
with MF regardless of their mutation status.6 Many other
406
therapy with ruxolitinib. However, JAK2 inhibitors do not

eliminate the disease and new medications are needed to
expand on the benefits seen with JAK2 inhibitors. Although
many agents are still in the early stages of development, the
wealth of publications and presentations has continued to
support our growing understanding of the pathophysiology of
MF as well as the potential short- and long-term outcomes of
these new and diverse approaches to treatment. Focus of
ongoing efforts is particularly on the improvements in anemia
and fibrosis, as well as on rational combination trials of JAK2
inhibitors and other potentially active agents. Therapeutic
potential and limitations of JAK2 inhibitors and other novel
medications in clinical studies are reviewed.
JAK2 inhibitors, as well as histone deacetylase inhibitors

(HDACIs), mammalian target of rapamycin (mTOR) inhibitors, transforming growth factor (TGF)-beta inhibitors, and
others are in clinical studies (Table 1). An overview of
published information for these medications (including published abstracts) is provided here; we focus on medications
in active clinical development.
One important caveat when analyzing results of clinical
trials for patients with MF is that studies use distinct
response criteria, thus hampering comparisons between
different therapeutic approaches. Uniform criteria for response in MF were developed several years ago (International Working Group for Myelofibrosis Research and
Therapy [IWG-MRT] response criteria),9 but have already
been questioned, and modifications proposed (already in use
in some studies; e.g., anemia response as it relates to the
definition of transfusion dependency and independency in
patients with MF).10 Indeed, the importance of clinically
relevant and objective trial end points that would support
claims of efficacy cannot be emphasized enough.11 To that
end, use of magnetic resonance imaging (MRI) instead of
physical exam to assess a response in splenomegaly and the
development of Myelofibrosis Symptom Assessment Form
(MFSAF), which is an inventory to measure the symptom
burden in MPNs, have been largely adopted in ongoing
clinical trials.
JAK2 inhibitors
Ruxolitinib
Efficacy and safety of ruxolitinib has been evaluated in

two phase III clinical trials: the Controlled Myelofibrosis
Study with Oral JAK1/JAK2 Inhibitor Treatment I and II
(COMFORT-I and COMFORT-II).7,12 Results of these studies led to a recent approval of ruxolitinib in the United
From the Department of Leukemia, M. D. Anderson Cancer Center, Houston, TX.

Address reprint requests to Srdan Verstovsek, MD, PhD, M. D. Anderson Cancer Center,
Department of Leukemia, Unit 428, 1515 Holcombe Blvd., Houston, TX 77030; email:
sverstov@mdanderson.org.
1092-9118/10/1-10
SMALL-MOLECULE INHIBITORS FOR MPN

Table 1. Agents in Recent or Ongoing Clinical Studies for Myelofibrosis
Product
MOA
Company
Phase
Givinostat (ITF2357)
Panobinostat (LBH589)
Pracinostat (SB939)
Saridegib (IPI-926)
GS6624 (AB0024)
Pomalidomide (CC-4047)
AZD1480
BMS911543
CYT387
LY2784544
Ruxolitinib (Jakafi U.S. trade name;
INCB018424; INC424)
SAR302503 (TG101348)
Pacritinib (SB1518)
NS-018
Plitidepsin (Aplidin)
Everolimus (Afinitor; RAD-001)
HDAC inhibitor
HDAC inhibitor
HDAC inhibitor
Hedgehog inhibitor (inhibits smoothened)
LOXL2 humanized monoclonal antibody (mAb)
Immunomodulating agent
JAK2 inhibitor
JAK2 Inhibitor
JAK1/2, TYK2 inhibitor
JAK2 inhibitor
JAK1/2 inhibitor
Italfarmaco (Milan, Italy)

Novartis Pharmaceuticals (Basel, Switzerland)
S*Bio (Singapore, Singapore)
Infinity Pharmaceuticals (Cambridge, MA)
Gilead Sciences (Foster City, CA)
Celgene (Summit, NJ)
AstraZeneca (Wilmington, DE)
Bristol-Myers Squibb (Princeton, NJ)
YM Biosciences (Cytopia; Ontario, Canada)
Eli Lilly (Indianapolis, IN)
Incyte Corporation (Wilmington, DE) and
Novartis Pharmaceuticals
Sanofi (TargeGen; Paris, France)
S*Bio
Nippon Shinyaku (Kyoto, Japan)
PharmaMar (Madrid, Spain)
Novartis Pharmaceuticals
IIA
I/II
II
II
II
III
I/II
I/II
I/II
I
Approved in
United States
III
I/II
I/II
II
I/II
JAK2 inhibitor
JAK2 inhibitor
JAK2 inhibitor
Marine cyclic depsipeptide
mTOR inhibitor
Abbreviations: MOA, mode of action; HDAC, histone deacetylase; JAK, janus kinase; mTOR, mammalian target of rapamycin.
States as therapy for patients with intermediate- and highrisk MF.

COMFORT-I is a double-blind, placebo-controlled study
that enrolled 309 adults in the United States, Canada, and
Australia with palpable splenomegaly and an International
Prognostic Scoring System (IPSS) classification of intermediate-2 or high risk myelofibrosis (both primary or secondary MF).7 Patients were randomly assigned (1:1) to
receive ruxolitinib or placebo. Starting doses of ruxolitinib
depended on platelet count at baseline: 15 mg twice daily for
patients with at least 100 to 200 109/L and 20 mg twice
daily for patients with more than 200 109/L. The primary
end point was the proportion of patients achieving a reduction in spleen volume of at least 35% from baseline to week
24 as measured by MRI or computed tomography (CT). Key
KEY POINTS
Ruxolitinib, an oral Janus kinase (JAK) 1 and JAK2

inhibitor, has recently been approved in the USA as
the first medication for patients with intermediate- or
high-risk myelofibrosis (MF).
JAK inhibitor therapy in MF results in a rapid and
sustained reduction in organomegaly (spleen and
liver) and improvement in debilitating MF-related
constitutional symptoms. Quality-of-life improvement is of a significant clinical benefit for patients.
With better control of the signs and symptoms of the
disease with JAK inhibitors, patients with advanced
MF may live longer.
JAK2 inhibitors are not specific for JAK2V617F mutation, and therefore all patients with MF benefit
from therapy regardless whether they have JAK2
mutation.
JAK2 inhibitors do not eliminate the disease, and
new therapies and combination strategies are being
explored, including histone deacetylase inhibitors,
mammalian target of rapamycin inhibitors, and
others.
secondary end points included the durability of spleen volume reduction (loss of response defined as a reduction of less
than 35% from baseline and an increase of at least 25% from
nadir), the proportion of patients with at least 50% improvement in symptoms (as measured by a Total Symptom Score
[TSS] from the modified MFSAF version 2.0 electronic
diary), and overall survival. At week 24, 41.9% of patients
receiving ruxolitinib and 0.7% of patients receiving placebo
achieved a spleen volume reduction of at least 35% from
baseline (p 0.0001). In patients treated with ruxolitinib
who achieved at least 35% reduction in spleen volume,
67.0% maintained the reduction for 48 weeks or more. At
week 24, the proportion of patients who experienced a
50% or greater improvement in TSS was 45.9% with ruxolitinib and 5.3% with placebo (p 0.0001). In addition,
mean TSS improved by 46.1% with ruxolitinib and worsened
by 41.8% with placebo (p 0.0001). Additional analyses of
COMFORT-I showed similar trends in spleen volume reductions and TSS improvements with ruxolitinib treatment
regardless of patient subgroup evaluated, including myelofibrosis subtype (PMF, post-PV MF [PPV-MF], or post-ET
MF [PET-MF]), age ( 65 or 65 years), IPSS risk category
(high-risk or intermediate-2), baseline hemoglobin level
( 10 or 10 g/dL), baseline palpable spleen length ( 10
or 10 cm), and JAK2V617F mutation status (positive or
negative). At the time of the prospectively defined data
cutoff (median follow-up of 32 weeks), there were 10 deaths
with ruxolitinib and 14 deaths with placebo (6.5% vs. 9.1%;
hazard ratio [HR] 0.67; 95% CI, 0.30 to 1.50; p 0.33). In
a subsequent survival analysis on the basis of a planned
data cutoff with an additional 4 months of follow-up (median
follow-up, 51 weeks), there were 13 deaths with ruxolitinib
and 24 deaths with placebo (8.4% vs. 15.6%; HR 0.50; 95%
CI, 0.25 to 0.98; p 0.04).
COMFORT-II is an open-label phase III study that enrolled patients in Europe with palpable splenomegaly with
IPSS intermediate-2 and high-risk PMF, PPV-MF, or
PET-MF (N 219).12 Patients were randomly assigned (2:1)
to receive ruxolitinib or investigator-determined best available therapy (BAT). The dosing regimen for ruxolitinib
was similar to that in COMFORT-I, and BAT included any
commercially available monotherapy or combination ther-
407
SRDAN VERSTOVSEK
apy or no therapy. The primary end point was the proportion

of patients achieving a reduction in spleen volume of at least
35% from baseline to week 48, as measured by MRI or CT.
Key secondary end points included spleen volume reduction at 24 weeks, duration of spleen volume reduction,
progression-free survival, leukemia-free survival, and overall survival. Measurement of patient-reported quality of life
and reduction in MF-related symptoms by using the European Organisation for Research and Treatment of Cancer
(EORTC) QLQ-C30 and the Functional Assessment of Cancer Therapy Lymphoma (FACT-Lym) questionnaires was
an exploratory end point. At week 48, 28.5% of patients who
received ruxolitinib achieved at least 35% reduction from
baseline in spleen volume. In contrast, no patients who
received BAT achieved this end point (p 0.0001). Similarly, only patients in the ruxolitinib group achieved a
reduction from baseline of at least 35% at 24 weeks (31.9%,
ruxolitinib; 0%, BAT; p 0.0001). The median duration of
response with ruxolitinib was not reached; 80% of patients
continued to have a response after a median follow-up of
12 months. Subgroup analyses demonstrated that ruxolitinib was more effective than BAT at reducing spleen
volume regardless of gender, age, JAK2V617F mutation
status, IPSS risk category, baseline spleen size, myelofibrosis subtype, or ruxolitinib starting dose (15 or 20 mg). There
was no significant difference in progression-free survival,
leukemia-free survival, or overall survival between treatment groups. However, the study was not powered for
statistical analyses of these end points. Compared with
patients treated with BAT, ruxolitinib-treated patients experienced improvements in quality of life and marked reductions in MF-related symptoms, as measured by the EORTC
QLQ-C30 and FACT-Lym subscales.
The most common hematologic adverse events in both
COMFORT-I and COMFORT-II were thrombocytopenia and
anemia. These events were manageable with dose reduction or temporary interruption of therapy, and rarely led to
discontinuation of ruxolitinib (one patient for each event
in COMFORT-I; one patient for thrombocytopenia in
COMFORT-II). The most common nonhematologic adverse
events with ruxolitinib in COMFORT-I were ecchymosis,
dizziness, and headache; these events were primarily grade
1 or 2. The percentage of patients that stopped therapy
as a result of ruxolitinib-related adverse effects in the
COMFORT-I study was similar to the percentage of patients
that stopped therapy as a result of placebo-related adverse
effects. After temporary interruption or discontinuation of
ruxolitinib, MF-related symptoms generally returned to
baseline levels within approximately 1 week. Analysis of
adverse events that occurred after interruption or discontinuation of ruxolitinib therapy showed no clear pattern to
indicate a specific withdrawal effect. However, gradual tapering of the ruxolitinib may be considered when discontinuing therapy for reasons other than thrombocytopenia.
Importantly, there are no contraindications to prescribing
ruxolitinib to patients with intermediate- or high-risk MF.5
SAR302503
A phase I/II clinical trial with SAR302503 has been

published, and updated follow-up data were recently presented.13,14 Fifty-nine patients with MF were enrolled and
received SAR302503 at doses ranging from 30 to 800 mg
once daily. The maximum tolerated dose (MTD) was 680 mg
408
daily, and dose-limiting toxicities (DLTs) were asymptomatic grade 3 to 4 hyperamylasemia and hyperlipasemia.
After six cycles of therapy, the spleen response in the MTD
cohort was 45% by IWG criteria. No MRI testing was
performed to assess volumetric reduction of the spleen;
measurements were done by palpation. Most patients who
presented with leukocytosis and thrombocytosis experienced
normalized counts. There was a significant improvement in
systemic symptoms after two to six cycles of therapy: early
satiety (56% complete resolution), fatigue (63% improvement), night sweats (64% complete resolution), cough (67%
complete resolution), and pruritus (50% complete resolution). A decrease in the JAK2V617F allele burden was
observed in some patients: patients who presented with high
( 20%) allele burden at diagnosis had a significant decrease
after 24 months. There was considerable hematologic toxicity: new-onset transfusion dependent anemia (grade 3 to 4,
35.1%) and thrombocytopenia (grade 3 to 4, 23.7%). Other
adverse effects included diarrhea, nausea, and vomiting.
Current new clinical trials with this JAK2 inhibitor include a phase II randomized trial exploring alternative
dose schedules to improve tolerability and maintain efficacy,
as well as phase III placebo controlled blinded study for
possible approval of this medication as therapy for MF
(patients are randomly assigned to placebo or SAR302503 at
initial doses of 400 and 500 mg once daily; JAKARTA;
NCT01437787).
CYT387
Results of a phase I/II clinical trial with CYT387 have

been presented in abstract form.15,16 MTD was determined
to be 300 mg/day, and DLTs included grade 3 headache and
grade 3 hyperlipasemia at 400 mg/day. Most patients experienced improvements in pruritus, night sweats, bone pain,
and fever. In the last update, an experience in 166 patients
was summarized, after a median follow-up of 10.4 months;
32 patients (19%) have discontinued therapy. A reduction
in splenomegaly of at least 50% to qualify as a response
by IWG-MRT criteria occurred in 31% of patients (by palpation). More strikingly was the response observed in anemia.
Sixty-eight patients (41%) were transfusion dependent at
baseline. The rate of transfusion independence for 12 weeks
and hemoglobin of at least 8 g/dL was 46%, and was 62%
among patients receiving treatment at MTD. Median time
to transfusion independence was 84 days, and median duration has not been reached. The mechanism of anemia
response remains a focus of active investigation. One complicating factor is that the IWG response criteria definitions for transfusion dependence and independence have
been shown inadequate for proper assessment of clinical
benefit: In COMFORT-1 study (described earlier, evaluating
ruxolitinib therapy vs. placebo in blinded fashion), those
patients receiving placebo experienced transfusion independence by IWG criteria in 47% of cases.7 Further maturation
of data and results of this study are awaited in the near
future. The most common hematologic adverse effect was
thrombocytopenia, which occurred in 33% of cases and was
grade 3 to 4 in 17%. The most common nonhematologic
toxicity (20%; grade 1 only) was called the first dose effect:
transient lightheadedness and/or dizziness which may be
accompanied by hypotension. Other less common and low
grade nonhematologic adverse effects included peripheral
neuropathy, diarrhea, nausea, and headache.
SMALL-MOLECULE INHIBITORS FOR MPN

LY2784544
Immunomodulatory Inhibitory Drugs
In an open-label phase I trial, 19 patients received treatment with LY2784544.17 The MTD was 120 mg. Similar to
results with other JAK2 inhibitors, a reduction in spleen
size and systemic symptoms was observed within the first
two to three cycles of therapy. No reduction of JAK2V617F
allele burden had been noted; however, preliminary results
suggest that the compound could improve the fibrosis associated with this disorder. Tumor lysis syndrome was observed at the lower end of the dose range associated with
efficacy, and therefore dosing has been amended to include a
lower dose lead-in period. This study is ongoing.
The first published study of pomalidomide (an analog of

thalidomide, formerly CC-4047) evaluated 84 patients with
MF who were randomly assigned among four arms: pomalidomide 2 mg daily (n 22), pomalidomide 2 mg daily plus
prednisone (n 19), pomalidomide 0.5 mg daily plus prednisone (n 22) and prednisone alone (n 21).25 Observed
benefit was limited to improvement in anemia, according to
IWG criteria: Response rates were 23% (pomalidomide 2 mg
plus placebo), 16% (pomalidomide 2 mg plus prednisone),
36% (pomalidomide 0.5 mg plus prednisone), and 19% (prednisone alone). The drug was very well tolerated.25 In a phase
I/II trial, the MTD of pomalidomide was determined to be
3 mg/day, and DLT was myelosuppression.26 In accordance
with the results of the randomized study, better response
rates were observed in patients who received low-dose pomalidomide (63% response in anemia).26 Begna and colleagues recently reported the results of a use of low-dose
pomalidomide alone (0.5 mg/day) in 58 patients with MF.27
The anemia response was 17% by IWG-MRT criteria. An
increase in platelet count in patients with thrombocytopenia
was observed in 58% of cases. No patient had an improvement in splenomegaly. A phase III randomized study of
pomalidomide compared with placebo in patients with MF
who are transfusion dependent is underway. Patients are
randomly assigned to either pomalidomide (0.5 mg/day) or
placebo; primary end point is rate of transfusion independence after 6 months of therapy.
HDACIs
HDACIs are compounds that inhibit activity of HDAC and

can lead to increased histone acetylation and gene expression. Some HDACIs in clinical trials for MPNs include
givinostat (formerly known as ITF2357) and panobinostat
(formerly known as LBH589). A pilot study with givinostat
in MF was recently published.18 Twenty-nine patients with
JAK2V617F-positive MPNs (PV, n 12; ET, n 1; MF, n
16) received treatment with givinostat 50 mg twice daily.
Responses were seen in 54% of PV/ET patients (complete,
n 1; partial, n 6) according to European LeukemiaNet
criteria. Responses in MF were more modest: Only three of
16 patients had a major response. Adverse effects included
diarrhea, anemia, thrombocytopenia, fatigue, and QTc elongation. Panobinostat was evaluated in a phase I trial for
patients with hematologic malignancies, including 13 patients with MF.19 Most patients received panobinostat
thrice weekly, and the MTD was 60 mg/dose. Grade 3 to 4
adverse effects included thrombocytopenia (33%), fatigue
(28%, DLT), neutropenia (28%), and anemia (12%). Four
patients with MF had response by IWG-MRT criteria. These
results were followed by two studies evaluating panobinostat solely in patients with MF. Mascarenhas and colleagues
reported on a phase I study in 15 patients with MF;
thrombocytopenia was the DLT.20 Five patients received
more than 6 months of therapy, and all achieved clinical
improvement in spleen by IWG response criteria.21 In another trial, 31 patients with MF received treatment with
panobinostat (initial dose was 60 mg thrice weekly) but
the toxicity precluded delivery of the medication beyond 1
month of therapy in almost all patients.22 A combination
study of ruxolitinib and panobinostat at low dose is underway on the basis of exciting preclinical results.23
mTOR Inhibitor
Activated mTOR is a serine/threonine kinase which regulates cell growth, proliferation and metabolism. Results for
30 patients from a phase I/II study of everolimus (RAD-001)
in high- or intermediate-risk primary or secondary MF were
published during 2011.24 No DLT was observed up to 10
mg/day. At this dose, toxicities were infrequent; the most
common toxicity was grade 1 to 2 stomatitis. A splenomegaly
reduction of more than 50% from baseline occurred in 20% of
patients, whereas a more than 30% reduction occurred in
44% of patients. A total of 69% and 80% of patients experienced complete resolution of systemic symptoms and pruritus, respectively. Response in leukocytosis, anemia, and
thrombocytosis occurred in 15% to 25% of patients.
Conclusion
In recent years, the outlook for patients with chronic

Ph-negative MPNs has changed with new discoveries on
molecular biology of these diseases and the subsequent
development of new compounds directed against those molecular defects. JAK2 inhibitors and other compounds, such
as pomalidomide, are on the verge of making the transition
from clinical trials to routine clinical use. Indeed, ruxolitinib, an oral JAK1 and JAK2 inhibitor, has recently been
approved in the United States as the first medication ever
for treatment of patients with intermediate- and high-risk
MF.13 The most striking benefits observed with JAK2 inhibitors are a reduction in spleen size and improvement in
constitutional symptoms. As suggested by COMFORT-I trial
results, with the better control of the symptoms and signs
of the disease, thus potentially delaying a progression of the
disease, patients with advanced MF may live longer. However, with JAK2 inhibitors a reduction in bone marrow
fibrosis is mostly anecdotal. Similarly, with the majority of
compounds there is no significant reduction in the JAK2
allele burden. JAK2 inhibitors are not able to clonally
eradicate the disease. Therefore, JAK inhibitors are just
the first building block in our efforts to effectively treat MF,
and there is much room for improvement. We need to better
understand how these drugs work, and through which
mechanism(s) they are producing benefits for individual
patients. We need biomarkers to determine which patients
will respond to a specific agent, as well as to rationally
combine active agents for additional benefits. With coordinated efforts and appropriate design of clinical trials, we can
hope to overcome these challenges and improve outcomes for
patients with MF, not just in controlling disease signs and
symptoms, but potentially eliminating the disease.
409
SRDAN VERSTOVSEK
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Srdan Verstovsek
Research
Funding
Expert
Testimony
Other
Remuneration
AstraZeneca;
Bristol-Myers
Squibb; Celgene;
Geron; Gilead
Sciences; Incyte;
Lilly; NS Pharma;
Roche; S*Bio
REFERENCES
1. Cervantes F, Pereira A. Advances in the understanding and management of primary myelofibrosis. Curr Opin Oncol. 2011;23:665-671.
2. Mesa RA. Assessing new therapies and their overall impact in myelofibrosis. Hematology Am Soc Hematol Educ Program. 2010;2010:115-121.
3. James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation
leading to constitutive signalling causes polycythaemia vera. Nature. 2005;
434:1144-1148.
4. Quintas-Cardama A, Kantarjian H, Cortes J, et al. Janus kinase inhibitors for the treatment of myeloproliferative neoplasias and beyond. Nat Rev
Drug Discov. 2011;10:127-140.
5. JAKAFI [Prescribing Information]. Wilmington, DE: Incyte Corporation;
2011.
6. Verstovsek S, Kantarjian H, Mesa RA, et al. Safety and efficacy of
INCB018424, a JAK1 and JAK2 inhibitor, in myelofibrosis. N Engl J Med.
2010;363:1117-1127.
7. Verstovsek S, Mesa RA, Gotlib J, et al. A double-blind, placebocontrolled trial of ruxolitinib for myelofibrosis. N Engl J Med. 2012;366:799807.
8. Tefferi A. Novel mutations and their functional and clinical relevance
in myeloproliferative neoplasms: JAK2, MPL, TET2, ASXL1, CBL, IDH and
IKZF1. Leukemia. 2010;24:1128-1138.
9. Cervantes F, Dupriez B, Pereira A, et al. New prognostic scoring system
for primary myelofibrosis based on a study of the International Working
Group for Myelofibrosis Research and Treatment. Blood. 2009;113:2895-2901.
10. Gale RP, Barosi G, Barbui T, et al. What are RBC-transfusiondependence and -independence? Leuk Res. 2011;35:8-11.
11. Barosi G, Tefferi A, Barbui T. Do current response criteria in classical
Ph-negative myeloproliferative neoplasms capture benefit for patients? Leukemia. Epub 2011 Nov 22.
12. Harrison C, Kiladjian JJ, Al-Ali HK, et al. JAK inhibition with
ruxolitinib versus best available therapy for myelofibrosis. N Engl J Med.
2012; 366:787-798.
13. Pardanani A, Gotlib J, Jamieson C, et al. SAR302503: interim safety,
efficacy and long-term impact on JAK2 V617F allele burden in a phase I/II
study in patients with myelofibrosis. Blood. 2011;118 (abstr 3838).
14. Pardanani A, Gotlib JR, Jamieson C, et al. Safety and efficacy of
TG101348, a selective JAK2 inhibitor, in myelofibrosis. J Clin Oncol. 2011;
29:789-796.
15. Pardanani AD, Caramazza D, George G, et al. Safety and efficacy of
410
CYT387, a JAK-1/2 inhibitor, for the treatment of myelofibrosis. J Clin Oncol.

2011;29 (suppl; abstr 6514).
16. Pardanani A, Gotlib J, Gupta V, et al. An expanded multicenter phase
I/II study of CYT387, a JAK- 1/2 inhibitor for the treatment of myelofibrosis.
Blood. 2011;118 (abstr 3849).
17. Verstovsek S, Mesa RA, Rhoades SK, et al. Phase I study of the JAK2
V617F inhibitor, LY2784544, in patients with myelofibrosis (MF), polycythemia vera (PV), and essential thrombocythemia (ET). Blood. 2011;118 (abstr
2814).
18. Rambaldi A, Dellacasa CM, Finazzi G, et al. A pilot study of the
histone-deacetylase inhibitor givinostat in patients with JAK2V617F positive
chronic myeloproliferative neoplasms. Br J Haematol. 2010;150:446-455.
19. DeAngelo DJ, Spencer A, Fischer T, et al. Activity of oral panobinostat
(LBH589) in patients with myelofibrosis. Blood, 2009;114 (abstr 2898).
20. Mascarenhas J, Wang X, Rodriguez A, et al. A phase I study of LBH589,
a novel histone deacetylase inhibitor in patients with primary myelofibrosis
(PMF) and post-polycythemia/essential thrombocythemia myelofibrosis (PostPV/ET MF). Blood. 2009;114 (abstr 308).
21. Mascarenhas J, Mercado A, Rodriguez A, et al. Prolonged low dose
therapy with a pan-deacetylase inhibtor, panobinostat (LBH589), in patients
with myelofibrosis. Blood. 2011;118 (abstr 794).
22. DeAngelo DJ, Tefferi A, Fiskus W, et al. A phase II trial of panobinostat, an orally available deacetylase inhibitor (DACi), in patients with primary
myelofibrosis (PMF), post essential thrombocythemia (ET), and post polycythemia vera (PV) myelofibrosis. Blood. 2010;116 (abstr 630).
23. Baffert F, Evrot E, Ebel N, et al. Improved efficacy upon combined
JAK1/2 and pan-deacetylase inhibition using ruxolitinib (INC424) and panobinostat (LBH589) in preclinical mouse models of JAK2V617F-driven disease.
Blood. 2011;118 (abstr 798).
24. Guglielmelli P, Barosi G, Rambaldi A, et al. Safety and efficacy of
everolimus, a mTOR inhibitor, as single agent in a phase 1/2 study in patients
with myelofibrosis. Blood. 2011;118:2069-2076.
25. Tefferi A, Verstovsek S, Barosi G, et al. Pomalidomide is active in the
treatment of anemia associated with myelofibrosis. J Clin Oncol. 2009;27:
4563-4569.
26. Mesa RA, Pardanani AD, Hussein K, et al. Phase1/-2 study of pomalidomide in myelofibrosis. Am J Hematol. 2009;85:129-130.
27. Begna KH, Mesa RA, Pardanani A, et al. A phase-2 trial of low-dose
pomalidomide in myelofibrosis. Leukemia. 2011;25:301-304.
Insights into the Molecular Genetics of

Myeloproliferative Neoplasms
By Huong (Marie) Nguyen, MD, and Jason Gotlib, MD, MS
Overview: The molecular biology of the BCR-ABL1-negative

chronic myeloproliferative neoplasms (MPNs) has witnessed
unprecedented advances since the discovery of the acquired
JAK2 V617F mutation in 2005. Despite the high prevalence of
JAK2 V617F in polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis (PMF), and the common
finding of dysregulated JAK-STAT signaling in these disorders, it is now appreciated that MPN pathogenesis can reflect
the acquisition of multiple genetic mutations that alter several
biologic pathways, including epigenetic control of gene expression. Although certain gene mutations are identified at
higher frequencies with disease evolution to the blast phase,
PNS ARE clonal hematopoietic disorders that result

in overproduction of one or more terminally differentiated blood cell types arising from the myeloid lineage. In
the 2008 World Health Organization (WHO) classification,
MPNs are divided into eight subtypes: chronic myeloid
leukemia (CML); polycythemia vera (PV); essential thrombocythemia (ET); primary myelofibrosis (PMF); systemic
mastocytosis (SM); chronic neutrophilic leukemia; chronic
eosinophilic leukemia, not otherwise specified (CEL, NOS);
and MPN-unclassifiable (MPN-U).1 Myeloid (and lymphoid)
neoplasms associated with eosinophilia and rearrangement
of platelet-derived growth factor receptor alpha or beta
(PDGFRA or PDGFRB) and fibroblast growth factor receptor 1 (FGFR1) are distinguished by their own major WHO
disease category, but share numerous clinicopathologic features with MPNs.1 A pathogenetic hallmark of MPNs is
dysregulation of tyrosine kinases (TKs), which in turn results in aberrant downstream signaling and increased cellular proliferation and/or decreased apoptosis. Table 1
summarizes the molecular lesions (e.g., reciprocal chromosomal translocations, point mutations, interstitial chromosomal deletions) that generate oncogenic TKs in MPNs and
myeloid neoplasms associated with eosinophilia. The notion
that PV, ET, or MF is driven by a single TK lesion such as
JAK2 V617F has now been abandoned given the genetic and
epigenetic complexity observed in most patients (Fig. 1). The
cytogenetics of MPNs is important to understand disease
pathogenesis and prognosis; however, this topic is not addressed in this monograph, and readers are directed elsewhere for reviews on the subject.
Before JAK2 V617F: Clonality, Cytokine Independence,
and Aberrant JAK-STAT Signaling
In a 1951 Blood editorial, Dr. William Dameshek first

conceptualized the inter-relatedness of PV, ET, and MF and
postulated a hitherto undiscovered stimulus as the biologic
basis of their shared myeloproliferative features.2 In the
1970s, Adamson and colleagues used restriction fragment
length polymorphism analysis of the X-linked glucose-6phosphate dehydrogenase (G6PD) gene in a female patient
to confirm the clonal basis of PV, with ensuing studies
demonstrating clonality in ET and MF.3 These investigations were followed by two seminal observations: 1) hematopoietic progenitors from patients with PV (and in some
MPN initiation and progression are not explained by a single,

temporal pattern of clonal changes. A complex interplay
between acquired molecular abnormalities and host genetic
background, in addition to the type and allelic burden of
mutations, contributes to the phenotypic heterogeneity of
MPNs. At the population level, an inherited predisposition to
developing MPNs is linked to a relatively common JAK2associated haplotype (referred to as 46/1), but it exhibits a
relatively low penetrance. This review details the current state
of knowledge of the molecular genetics of the classic MPNs
PV, ET, and PMF and discusses the clinical implications of
these findings.
cases ET and MF) proliferate in the absence of exogenous

cytokines such as erythropoietin (Epo) (e.g., endogenous
erythroid colony growth [EEC]), and 2) such cells are hypersensitive to growth factors such as Epo, thrombopoietin
(Tpo), and interleukin-3 (IL-3).4
The endogenous, self-stimulatory property of MPN cells
and cytokine hypersensitivity led to increasing interest in
JAK2 as a contributor to MPN pathogenesis. JAK2 belongs
to a family of four janus kinases, which also includes JAK1,
JAK3, and TYK2. Each JAK protein has an active tyrosine
kinase domain (JAK homology 1[JH1]), an inactive pseudokinase domain (JAK homology 2 [JH2]), an SRC homology 2
domain (SH2), and an amino terminal FERM (4-point-1,
Erzin, Radixin, Moesin) homology domain, which binds to
the cytoplasmic tail of cytokine receptors. JAK2 facilitates
normal myelopoiesis by transmitting signals from type I
receptors for Epo (EpoR), thrombopoietin (TpoR or MPL),
granulocyte-colony-stimulating factor (G-CSFR), and IL-3
(IL-3R). In the normal state, binding of ligand to receptor
causes JAK2 to change from a receptor-bound, inactive
conformation to an active catalytic enzyme because of escape
from the inhibitory effects of the pseudokinase domain on
the kinase domain.5 Auto-phosphorylation of JAK2 and
phosphorylation of downstream signaling intermediates results in recruitment of SH2-domain containing proteins
such as STAT3 and STAT5. After phosphorylation by JAK2,
the STAT proteins homodimerize and translocate to the
nucleus, where they activate transcription of target genes
involved in regulating a variety of cellular processes, including proliferation, differentiation, and apoptosis. Dampening
of JAK-STAT activation occurs via different negative feedback mechanisms, including the suppressor of cytokine
signaling (SOCS) family of proteins, LNK, CBL, and various
tyrosine phosphatases.
From the Division of Hematology, Department of Medicine, Stanford University School of

Medicine/Stanford Cancer Institute, Stanford, CA.
Address reprint requests to Jason Gotlib, MD, MS, Associate Professor of Medicine,
Stanford Cancer Institute, 875 Blake Wilbur Drive, Room 2324, Stanford, CA 94305-5821;
email: jason.gotlib@stanford.edu.
1092-9118/10/1-10
411
NGUYEN AND GOTLIB

Table 1. Tyrosine Kinase Mutations in Myeloproliferative Neoplasms
Prototypic Tyrosine
Kinase Mutation
Chronic Myeloid Leukemia
Polycythemia Vera
Essential Thrombocythemia
Primary Myelofibrosis
Systemic Mastocytosis
Chronic Eosinophilic Leukemia, Not Otherwise Specified
Chronic Neutrophilic Leukemia
Myeloproliferative Neoplasm, Unclassifiable
BCR-ABL1
JAK2 V617F
JAK2 V617F
JAK2 V617F
KIT D816V
None
None
None
Myeloid and Lymphoid Neoplasms with Eosinophilia and

Abnormalities of PDGFRA, PDGFRB, or FGFR1
PDGFRA-Rearranged
FIP1L1-PDGFRA
PDGFRB-Rearranged
FGFR1-Rearranged
Frequency
Chromosome
Comments
100%
95%
5060%
5060%
80%
N/A
N/A
N/A
t(9;22)(q34;q11.2)
9p24
9p24
9p24
4q12
N/A
N/A
N/A
10%*
4q12 cryptic interstitial 5 additional variant fusion partners with

chromosome deletion
PDGFRA identified
t(5;12)(q3133;p13)
20 additional variant fusion partners with
PDGFRB identified
t(8;13)(p1112;q12)
10 additional variant fusion partners with
FGFR1 identified
ETV6-PDGFRB
Rare
ZNF198-FGFR1
Rare
BCR-ABL1 defines CML

Post-MPN AML frequency 50%
Juxtamembrane KIT mutations 5%

No characteristic TK abnormality
Abbreviations: TK, tyrosine kinase; MPN, myeloproliferative neoplasm; AML, acute myeloid leukemia; N/A, not applicable.
* Estimated frequency of 10% among patients with idiopathic hypereosinophilia in developed countries.
The JAK2 V617F Mutation
Aberrant JAK-STAT signaling had been demonstrated in

various MPN patient subgroups, laying the groundwork for
identification of the molecular alterations in this pathway.
In 2005, the somatic activating mutation JAK2 V617F
(V617F) was discovered by several groups using either
high-throughput tyrosine kinase genome sequencing, or
functional methods such as microsatellite mapping to further define a segment of chromosome 9p containing the
JAK2 gene in which loss of heterozygosity (LOH) in that
region had previously been identified in some PV and ET
patients.6-9 The French groups focus on JAK2 derived from
investigations in which a JAK2 inhibitor and siRNA against
KEY POINTS
412
JAK2 V617F is a common pathogenetic mutation in

myeloproliferative neoplasms (MPNs) and is sufficient to produce a myeloproliferative phenotype in
murine models.
Molecular abnormalities in positive and negative
regulators of the JAK-STAT axis (e.g., MPL, CBL,
LNK) can recapitulate activation of this signaling
pathway in patients negative for JAK2 V167F.
The acquisition of multiple pre- and post-JAK2 mutational events is common in MPNs and likely contributes to phenotypic diversity and progression to
acute myeloid leukemia.
Epigenetic dysregulation, including changes in pathways that control DNA methylation and modification
of chromatin, has emerged as an important paradigm
in MPN biology.
Inherited predisposition to MPNs is partly explained
by a JAK2-associated haplotype. Genome-wide association studies are in progress to identify additional
susceptibility loci.
JAK2 blocked terminal differentiation of PV erythroid progenitors and/or blocked EEC growth.9 Sequencing of the
JAK2 gene revealed a somatic point mutation at base pair
1849 (G3 T), causing substitution of the normal valine to
phenylalanine in codon 617 (V617F) of exon 14. This V617F
mutation is found in approximately 95% of PV patients and
50% to 60% of ET and PMF patients. The mutation frequency is considerably less (e.g., 5%) in patients with other
MPNs, such as systemic mastocytosis and chronic eosinophilic leukemias or acute myeloid leukemia. However, in
patients with overlap MDS/MPNs, such as proliferative-type
chronic myelomonocytic leukemia (CMML) and refractory
anemia with ring sideroblasts with thrombocytosis (RARST), the frequency increases to approximately 10% and 50%,
respectively.
The V617F mutation resides in the autoinhibitory pseudokinase domain of JAK2. Modeling studies suggest that
V617F causes a structural change in the pseudokinase
domain, relieving inhibition of JAK2 kinase activity.5 Overexpression of the JAK2 V617F allele in cell lines results
in phosphorylation of JAK2 and STAT5 in the absence of
cytokine stimulation.6,8,9 The human erythroleukemia
(HEL) cell line, which carries a homozygous V617F mutation, exhibits constitutive phosphorylation of JAK2 and
STAT5; treatment of HEL cells with a JAK2 inhibitor led to
reduced phosphorylation of JAK2 and STAT5 and inhibition of proliferation.7 In a V617F knock-in murine model of
PV, STAT5 was an absolute requirement for the pathogenesis of PV.10 Deletion of STAT5 normalized all the clinical,
blood, and histopathologic features of PV, including EEC
formation.
One JAK2 V617F Mutation, Multiple MPN Phenotypes
The high prevalence of the V617F mutation in three

clinically distinct MPNs begs the question of what additional factors contribute to phenotypic diversity between PV,
ET, and PMF. First, mutant allele burden of V617F may
modulate phenotype. Murine retroviral transplant models
resulting in high levels of V617F expression produced a
PV-like phenotype with marked erythrocytosis.11 Con-
GENETICS OF MYELOPROLIFERATIVE NEOPLASMS
Fig. 1. Genetic and Epigenetic Dysregulation in Myeloproliferative Neoplasms. Adapted from Expert Review of Hematology, JAK2 V617F
and beyond: role of genetics and aberrant signaling in the pathogenesis of myeloproliferative neoplasms, Vol. 3, No. 3, June 2010, pages
323-327. Stephen Oh and Jason Gotlib, Figure 1, with permission of Expert Reviews Ltd.; and adapted with kind permission from Springer
Science and Business Media: Current Hematologic Malignancy Reports, Disordered Epigenetic Regulation in the Pathophysiology of Myeloproliferative Neoplasms, vol. 7, no. 1, March 2012, pages 34-42, Su-Jiang Zhang and Omar Abdel-Wahab, Figure 1.
versely, transgenic models with more physiologic levels of

V617F expression resulted in phenotypes resembling ET
and PMF.12 These animal models corroborate the findings
in patients in which V617F allele burden tends to be the
highest in PV and PMF, with lower levels in ET patients.
For ET patients, positivity for V617F tends to confer a
PV-like phenotype, with higher hemoglobin and lower platelet counts than in V617F-negative ET patients. Differences
in intracellular signaling arising from V617F may also
explain the development of PV compared with ET: preferential activation of STAT1 constrains erythroid differentiation
and promotes megakaryocytic development, leading to an
ET phenotype.13 In contrast, reduced STAT1 phosphorylation promotes erythroid development as observed in PV.
The type of hematopoietic progenitor(s) targeted by the
V617F mutation may contribute to differences in MPN
subtype. The V617F mutation is found in myeloid, or less
commonly, lymphoid lineage cells from MPN patients, suggesting that it arises in a hematopoietic stem cell (HSC) or
early progenitor cell.14 Jamieson and colleagues identified
the V617F mutation in cells from PV patients with an HSC
immunophenotype and demonstrated that these cells were
skewed toward the erythroid lineage.15 These findings suggest that either the V617F mutation induces erythroid
differentiation or that the mutation preferentially targets an
HSC subset already committed to an erythroid fate.
Host genetic background may also influence disease presentation. In retroviral transplant models, disparate phenotypes were observed depending on the mouse strain. In
C57Bl/6 mice, transplantation with JAK2 V617Ftransduced cells resulted in a PV-like disease predominantly
characterized by erythrocytosis.16 However, in Balb/C mice,
similar experiments yielded mice with erythrocytosis, but
also leukocytosis and the subsequent development of myelofibrosis.17

Perhaps the most compelling basis for MPN diversity
comes from the additional molecular abnormalities that
either precede or follow the acquisition of V617F (Table 2).
The aggregate data suggest that there is no strict temporal
order of mutation occurrence that defines the development
or natural history of specific MPNs. However, several lines
of evidence support that V617F may arise on a pre-existing
abnormal clonal substrate: 1) in some patients, the V617F
burden is relatively small compared with the proportion of
cells with a coexistent clonal karyotypic abnormality; and
2) in AML arising from a V617F-positive MPN, the mutant
V617F allele can frequently no longer be detected.18 This
suggests that the MPN and AML share a clonal origin that
likely preceded the acquisition of V617F. Furthermore, in
V617F-positive PV and ET patients, both JAK2 wild-type
and V617F-positive EECs have been detected within the
same patient, suggesting that a separate event may confer
clonal, erythropoietin-independent growth, before V617F.
One study in ET patients demonstrated that V617F was
acquired on separate alleles as multiple, independent
events.19 Therefore, at least for ET, the V617F mutation
does not necessarily confer clonal dominance, permitting
JAK2 V617F-negative cells to continue to proliferate and
subsequently acquire the V617F mutation independently.
Activated JAK-STAT Signaling without JAK2 V617F
In the absence of V617F, activation of JAK-STAT signaling can be demonstrated in some MPN patients, suggesting
that molecular alterations that interdigitate with this axis
may contribute to MPN pathogenesis. This paradigm is
413
NGUYEN AND GOTLIB

Table 2. Estimated Frequencies of Non-JAK2 V617F Gene Mutations in PV, ET, MF, and Post-MPN AML
Gene
Exon 12 JAK2
MPL
CBL
LNK
TET2
DNMT3A
IDH1/IDH2
EZH2
ASXL1
IKZF1
TP53
Chromosome Location
9p24
1p34
11q23
12q24
4q24
2p23
2q33.3/15q26.1
7q35
20q11.21
7p12
17p13.1
PV
13%
Rare
Rare
Rare/Reported
716%
7%
2%
35%
25%
Rare
3% in chronic phase MPN
ET
PMF
Post-MPN AML
Rare
15%
Rare
5%
411%
3%
1%
Rare
5%
Rare
Rare
510%
510%
5%
817%
715%
4%
613%
1323%
Rare
Not Reported
Not Reported
Rare/Reported
10%
20%
17%
22%
Not Reported
20%
21%
27%
Abbreviations: PV, polycythemia vera; ET, essential thrombocythemia; MF, myelofibrosis; MPN, myeloproliferative neoplasms; AML, acute myeloid leukemia.
exemplified by mutations in exon 12 of the JAK2 gene, the

receptor for thrombopoietin (MPL), CBL, and LNK.
Exon 12 JAK2
A combination of missense, insertion, or deletion mutations in exon 12 of JAK2 was first described in 10 patients
initially diagnosed with idiopathic erythrocytosis.20 The
identification of these mutations operationally redefine
these patients as having PV. These mutations affect a region
5 of the pseudokinase domain, spanning residues 536 to
547. In contrast to the singular V617F mutation in exon 14,
a review published in 2011 catalogued 37 different exon 12
mutations, including 11 nonsynonymous substitutions, 20
deletion variants, and six duplications.21 Similar to V617F,
the residues affected by exon 12 mutations are located at the
interface of the pseudokinase and kinase domains and are
postulated to cause a structural change resulting in JAK2
activation. This notion is supported by cell line experiments,
in which each of the mutant alleles conferred IL-3 independent growth and constitutive phosphorylation of JAK2.29
In addition, in a murine retroviral transplant assay, one of
the mutant alleles (K539L) caused a pronounced erythrocytosis and growth of Epo-independent erythroid colonies.20
In contrast to the V617F mutation, exon 12 mutations
appear to be restricted to PV and account for an aggregate
3% mutation frequency among 11 published PV cohorts.21
CBL
Casitas B-lineage lymphoma (CBL) proteins have dual

roles as multifunctional adaptor proteins, which recruit
components to downstream signaling pathways, and as E3
ubiquitin ligases, which are involved in the trafficking and
degradation of activated tyrosine kinases. Three mammalian CBL homologs exist: c-CBL, CBL-b, and CBL-c. c-CBL
consists of a tyrosine kinase binding (TKB) domain, a linker
domain, a RING finger domain (RFD), a proline-rich region,
and a ubiquitin-associated (UBA) domain overlapping with
a leucine zipper (LZ) motif. Several groups identified recurrent acquired uniparental disomy (UPD) at 11q (which
encompasses c-CBL) in a wide spectrum of myeloid malignancies, and the majority of these cases were found to
harbor mutations in c-CBL.24,25 The highest frequency of
c-CBL mutations have been identified in juvenile myelomonocytic leukemia (JMML) (approximately 20%),26 but are
also found in a low proportion of patients with MPN (e.g.,
10% of MF), MDS/MPN overlap conditions besides JMML,
and secondary AML transformed from MDS or MPN.27 The
majority of mutations in c-CBL are missense mutations
localizing to the linker or RFD domains that result in loss of
ubiquitin ligase activity. The high frequency of homozygous
mutations findings are consistent with the notion that
c-CBL is a tumor suppressor and that loss of c-CBL function
leads to hypersensitivity to cytokines via dysregulation of
cytokine receptor-mediated signaling.
MPL
LNK
Mutations of MPL, the gene encoding the thrombopoetin

receptor, have been identified in approximately 5% to 10% of
PMF patients, approximately1% to 5% of patients with ET,
and rarely in PV.22,23 The two most common amino acid
substitutions at codon 515 result in a tryptophan to leucine
(W515L) or lysine (W515K) change, with rare asparagine
and alanine variants having been reported. Although originally described in familial ET cohorts, acquired MPL S505N
mutations have also been described. All MPL gene mutations reside in exon 10, which comprises the juxtamembrane, intracytosolic portion of the protein important for
preventing spontaneous receptor activation. Overexpression
of the mutant MPL W515L allele in cell lines results in
cytokine-independent growth, TPO hypersensitivity, and
activated JAK-STAT signaling.22 In a murine retroviral
transplant model, the W515L allele produced a phenotype of
marked thrombocytosis, splenomegaly, and reticulin fibrosis, but not erythrocytosis.
LNK (SH2B3) belongs to a family of adaptor proteins (e.g.,

also SH2-B, APS) that share several structural motifs,
including a proline-rich N-terminus, a pleckstrin homology
(PH) domain, an SH2 domain, and a conserved tyrosine
residue near the C-terminus. LNK binds to MPL via its SH2
domain and colocalizes to the plasma membrane via its PH
domain. On cytokine stimulation with Tpo, LNK binds
strongly to JAK2 and inhibits downstream STAT activation,
thereby providing critical negative feedback regulation.
LNK/ mice exhibit a phenotype similar to human MF,
including leukocytosis and thrombocytosis, as well as
splenomegaly with marked fibrosis and extramedullary hematopoiesis.28 An initial study of 33 V617F-negative ET and
PMF patients identified two individuals (6%) with mutations in exon 2 of LNK.29 One patient with PMF exhibited a
5 base-pair deletion and missense mutation leading to a
premature stop codon and loss of the pleckstrin homology
(PH) and SH2 domains. A second patient with ET had a
414
missense mutation (E208Q) in the PH domain. BaF3-MPL

cells transduced with these LNK mutants displayed augmented and sustained thrombopoietin-dependent growth
and signaling and primary samples from these patients
exhibited aberrant JAK-STAT activation. Follow-up studies
indicate a low frequency of LNK mutations in the chronic
phase of MPN (5%), increasing to approximately 10% with
transformation to AML.30 Mutations have also been found
in a few patients with idiopathic erythrocytosis/PV.31 Most
mutations cluster in an exon 2 hot spot, which are expected
to disrupt the PH domain, but frameshift, missense, and
nonsense mutations in exons 5, 7, and 8 have now been
reported, and may coexist with other MPN mutations in the
same patient. Taken together, these findings indicate that
JAK-STAT activation caused by loss of LNK negative feedback regulation can phenocopy MPN disease.
Mutations in Components of the Epigenetic Machinery
TET2
SNP and CGH arrays identified acquired LOH on chromosome 4q in patients with myeloid neoplasms, with further
mapping defining the TET oncogene family member 2
(TET2) gene as a mutated locus. Somatic mutations in TET2
are a mixture of deletions, frameshifts, stop codons, or
conserved amino acid substitutions and occur at a patient
frequency of 7% to 16% in PV, 4% to 11% in ET, and 8% to
17% in PMF.32,33 The biologic and biochemical roles of TET
family proteins and the functional consequences of TET2
mutations are becoming better clarified. TET proteins are
-ketoglutarate-dependent enzymes that catalyze the conversion of 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC).34 As DNA methylation is known to be an
important modulator of gene expression, it has been speculated that the conversion of 5mC to 5hmC may alter chromatin structure and restrict access to DNA methyltransferases,
which mediate repression of gene transcription. It follows
that loss of TET enzyme activity results in increased methylation, and in a sample of patients with TET2 mutations,
a reduction in levels of 5hmc levels is demonstrated.35 In
various TET2 knockout murine models, loss of TET2 results
in a myeloproliferative phenotype (e.g., splenomegaly, extramedullary hematopoiesis) with expansion of the hematopoietic stem cell compartment.36 Expression of mutant
TET2 in early hematopoietic precursors skews hematopoiesis in favor of myeloid over lymphoid progenitors.37
DNMT3A
Whole-genome sequencing of patients with AML revealed

recurrent mutations in 22% of AML patients in the gene for
DNA methyltransferase 3A (DNMT3A), another epigenetic
regulator.38 In this study, DNMT3A mutations were associated with poor outcome. In one analysis of patients with
MPN, 12 DNMT3A variants were found in 115 patients
(10%).39 Mutations were most frequently detected in patients during blastic phase transformation (6/35, 17%) and
MF (3/20, 15%), followed by PV (2/30, 7%) and ET (1/30, 3%).
In another study of 94 patients (46 PMF, 22 post-PV/ET MF,
11 blast-phase MPN, and 15 CMML), only three DNMT3A
mutations were found among 46 PMF patients (7%).40 In
both studies, mutations were always heterozygous, and in
some cases, coexisting JAK2 V617F, TET2, ASXL1, and IDH
1/2 mutations were found. Although DNMT3A mutations
are ubiquitous among myeloid neoplasms, their mechanistic

consequences are less clear. Using a conditional ablation
model in mice, it was found that loss of DNMT3A results in
impairment of hematopoietic stem cell (HSC) differentiation
and expansion of HSC with serial transplantation.41 However, analysis of HSC from DNMT3A-null mice showed both
increased and decreased methylation at specific loci. In the
study of AML patients, global methylation patterns and
5-methylcytosine content in genomes were not significantly
altered in patients with DNMT3A mutations. The potential
cooperative effects of DNMT3A with other epigenetic modifiers, and the genes role in disease transformation requires
further study.
IDH 1/2
The genes IDH1 and IDH2 encode enzymes that catalyze

oxidative decarboxylation of isocitrate to -ketoglutarate.
Mutant IDH has decreased affinity to isocitrate, but displays neomorphic catalytic activity toward -ketoglutarate,
resulting in accumulation of 2-hydroxyglutarate. IDH1 and
IDH2 mutations were first reported in gliomas/glioblastomas and AML. In a multi-institutional cohort of 1,473
MPN patients, 38 IDH mutations were found, and ranged in
frequency from 0.8% to 4% among chronic phase PV, ET, and
PMF patients (highest in the latter), and significantly increased to 21% in blast phase MPN.42 This study corroborates the results of smaller studies, which documented a
higher incidence of IDH 1/2 mutations in the leukemic phase
of MPNs. In a multivariate analysis, the presence of IDH
mutations predicted a worse survival in a subgroup of 43
blastic phase MPN patients from the Mayo Clinic.42
It is noteworthy that TET2 activity is impaired in cells
with mutated IDH 1/2 because its activity depends on
a-ketoglutarate.43 Mutual exclusivity of IDH 1/2 and TET2
mutations has been observed in AML and was speculated
to extend to MPNs as well. However, in the aforementioned
study, IDH mutational frequencies were similar among
patients with JAK2 (3.6%), MPL (4.3%) and TET2 (3.2%)
mutations.42 Given the coexistence of these molecular derangements, the specific contribution of IDH 1/2 mutations
to epigenetic dysregulation, and MPN biology in general,
becomes more difficult to dissect.
Mutations of the Polycomb Repressive Complex (PRC2)
The polycomb repressive complex (PRC2) is the major

methyltransferase for histone H3K27 methylation, one type
of post-translational histone modification involved in regulation of gene transcription. A low frequency of mutations
in the PRC2 enzymatic component EZH2, noncatalytic components EED and SUZ12, and PRC2-associated protein
JARID2 have been described in either MDS, MPNs, or
MDS/MPNs overlap disorders.44-46 Acquired uniparental
disomy on chromosome 7q led to identification of EZH2
mutations, which are either heterozygous or homozygous
and result in loss-of-function.44 To date, a well-defined
picture of the effect of PRC2 complex-related mutations on
myeloid pathobiology has not yet emerged. The prognostic
relevance of EZH2 mutations was assessed in a survey of
370 patients with PMF and 148 patients with post PV/ETMF. Twenty-five different EZH2 mutations were found in
5.9% of PMF, 1.2% of post PV-MF, and 9.4% of post-ET MF
patients. Leukemia-free and overall survival was significantly reduced in patients with EZH2-mutated PMF.47
415
NGUYEN AND GOTLIB

ASXL1 and L3MBTL1
Genetic Instability and Apoptosis Resistance
ASXL1 belongs to the Enhancer of trithorax and Polycomb

gene family. Its biologic roles are not well understood;
however, one epigenetic function ascribed to the Drosophilia
homolog is modification of chromatin complexes, including
ubiquitination of histone H2A lysine 119.48 Similar to other
epigenetic modifiers, mutation of ASXL1 (primarily exon 12)
is found in a wide spectrum of myeloid malignancies, but
among the MPNs, seems to be preferentially associated with
MF at a rate up to 13% to 23%.45
Another polycomb family member, L3MBTL1, is the human homolog of the Drosophila lethal,3 malignant brain
tumor (L3MBT) that functions as a tumor suppressor. It is
located on the long arm of chromosome 20 (q12), within a
region commonly deleted in several myeloid malignancies
and is a candidate tumor suppressor gene. Depletion of
L3MBTL1 from human cells causes replicative stress, DNA
breaks, activation of the DNA damage response, and
genomic instability.49 In addition to being involved in histone H4 methylation,50 it has been found that haploinsufficiency for L3MBTL1 promotes erythroid differentiation,
suggesting a role in PV development.51
Increased DNA damage and evasion of apoptosis may

also play a role in MPN pathogenesis. Expression of the
antiapoptotic protein Bcl-xL is increased in PV erythroid
progenitors, suggesting that this is one possible mechanism
of resistance to apoptosis that normally occurs on withdrawal of Epo.56 In cell line experiments, overexpression of
V617F induced an increase in homologous recombination
and genetic instability.57 Furthermore, a marked increase in
homologous recombination was found in CD34 positive progenitors from patients with PV and PMF when compared
with controls. Nonenzymatic deamidation of Bcl-xL is a key
step in the induction of apoptosis in response to DNA
damage. This pathway is inhibited in both CML and PV.58
JAK2 inhibitors restored the Bcl-xL deamidation pathway
in primary samples from PV patients, suggesting that activated signaling mediated by V617F may lead to resistance to
apoptosis.
Extra-TK Functions of JAK2 V617F
Recently, the unexpected finding was made that normal

JAK2 can translocate to the nucleus. Among its nuclear
roles, JAK2 can phosphorylate histone H3Y41, releasing the
transcriptional repressor HP1alfa from chromatin.52 Exclusion of HP1alfa resulted in increased gene transcription and
elevated expression of the oncogene LMO2 in leukemia
cells, which was reversed with inhibition of JAK2. It still
remains to be determined how JAK2s effects on chromatin
structure translate into expression of specific genes and
promotion of oncogenesis in certain cellular or disease contexts. A nucleus-associated gain-of-function of JAK2 V617F,
separate from its role in activated JAK-STAT signaling, is
its capacity to phosphorylate the protein arginine methyltransferase 5 (PRMT5).53 This modification reduces PRMT5s
ability to methylate histones H2A and H4. These chromatin
changes were modeled with knockdown of PRMT5 by a short
hairpin RNA in CD34-positive cells, resulting in erythroid
differentiation and increased colony formation. These data
suggest that the inhibitory effects of JAK2 V617F on PRMT5
activity may be relevant to MPN pathobiology.
Other Genetic Mutations and Pathway
Alterations in MPNs
Hemizygous deletions encompassing all or part of the gene

encoding the Ikaros (IKZF1) transcription factor on chromosome 7 are associated with MPN transformation.54 Modeling
of IKZF1 haploinsufficiency in mice progenitors with SiRNA
knockdown of IKZF1 led to an increase in cytokinedependent growth and STAT5 activation. In another study
of post-MPN AML, P53 mutations were found at a frequency
of 27%, including independent bi-allelic abnormalities and
homozygous mutations caused by acquired uniparental disomy of chromosome 17p.55 Other genes mutated in the
leukemic phase of MPNs include NRAS and RUNX1. The
role of all of these genes in disease initiation compared to
MPN transformation remains to be clarified.
416
Hypermethylation and Phosphorylation of SOCS Family Members
The SOCS proteins are SH2-domain containing proteins

that function as key negative regulators of JAK-STAT signaling. SOCS1 and SOCS3 can bind to the catalytic domain
of JAK2 and inhibit its kinase activity, as well as target
JAK2 for ubiquitination and subsequent degradation. Recent studies have suggested that hypermethylation of SOCS
genes may also play a role in MPN pathogenesis. Hypermethylation of SOCS3, and to a lesser extent, SOCS1, has
been observed in either V617F or wild type JAK2 chronic
and blast phase MPN patient samples.59
Phosphorylation of SOCS proteins leads to enhanced
SOCS protein degradation, and this process may be altered
in MPNs. In addition to its previously highlighted gain-offunctions, activated JAK2 V617F has also been shown to
overcome normal SOCS regulation by hyper-phosphorylating SOCS3, thus effectively blocking its inhibitory
activity and perhaps even potentiating the proliferative
capacity of the mutant V617F allele.60 To date, somatic
mutations in SOCS genes in MPNs have not been demonstrated.
Inherited Susceptibility to MPNs
A five- to seven-fold increased risk of MPN development is

found in first-degree relatives of patients with MPN. Three
studies demonstrated that a germline haplotype (GGCC,
referred to as 46/1) encompassing the 3 region of JAK2
gene is associated with a three- to four-fold risk of developing a V617F-positive (as well as MPL-mutated) MPN.61-63
Patients who were heterozygous for this haplotype preferentially acquired the V617F mutation in cis with the predisposition allele, suggesting that the haplotype may lead to
hypermutability at the JAK2 locus. However, the haplotype
was also weakly associated with JAK2 V617F-negative
MPNs, suggesting that it may confer a more generalized
propensity for MPN development independent of JAK2
V617F. One possibility is that the germline haplotype may
result in a functional difference such that cells with the risk
haplotype gain a selective advantage. However, evidence to
support this notion has been lacking. No differences in JAK2
expression level or nonsynonymous JAK2 coding polymorphisms associated with this haplotype have been identified.
Given the low penetrance of the haplotype, and a lack of
correlation with specific-disease related features, testing for
the 46/1 haplotype in first-degree relatives is not recommended. Recently, Mendelian inheritance of non-V617F
germline activating JAK2 variants, V617I and R564Q, was
identified in two families with thrombocytosis.64,65
Clinical Implications of the Molecular
Genetics of MPNs
Despite the burgeoning genetic data in MPNs, their clinical utility has thus far been limited. The evaluation of an
erythrocytosis, leukocytosis, or thrombocytosis begins with
a differential diagnosis between a reactive condition and
primary bone marrow disorder. The identification of an
activating mutation such as V617F, exon 12 JAK2 or MPL,
or clonal cytogenetic abnormality, identifies a myeloid neoplasm. However, the WHO classification requires a combination of clinical, laboratory, and histopathologic data in
order to diagnose a specific MPN. Despite individual reports
finding statistical significance between the presence (or
allelic burden) of certain molecular abnormalities in myelofibrosis, such data have not been prognostically validated
for overall and leukemia-free survival in the current scoring
systems (e.g., DIPSS-Plus). In addition, risk stratification in
PV and ET is still guided primarily by age and prior history
of thrombosis. At this time, this new molecular information
also has minimal influence on treatment decisions. In MF,
JAK inhibitors demonstrate activity in patients with either
wild-type or mutant JAK2, and trial participation is not
dependent on JAK2 mutation status.
Conclusion and Future Directions
Although mutations such as V617F and MPL are alone

sufficient to induce myeloproliferative disease in mice, MPN
initiation and progression in humans is genetically more
complex. Current whole genome and exome sequencing
approaches will be useful for unearthing novel molecular
determinants of MPN and to further evaluate specific genetic loci derived from genome-wide association studies. In
the current genomic era, evaluation of panels of acquired
somatic variants, which are linked to patient outcome data,
will help establish whether particular combinations of mutations generate additional prognostic risk information. The
recognition that both genetic and epigenetic dysregulation
underlie MPN disease justifies the use of not only JAK
inhibitors, but also DNA/chromatin-modifying agents such
as hypomethylating drugs and histone deacetylase inhibitors. An enhanced understanding of the molecular genetic
profile of the leukemic stem cells from which MPNs originate is critical to developing therapies that eradicate minimal residual disease and effect cure.
Acknowledgments
The authors wish to thank members of the Stanford Division
of Hematology and the International Working Group for Myeloproliferative Neoplasms Research and Treatment (IWGMRT) for their support. Because of space restrictions, we
apologize to investigators whose work was not included in this
review.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Huong (Marie) Nguyen*

Jason Gotlib
Gilead Sciences;
Incyte; Novartis;
YM BioSciences
Gilead Sciences;
Incyte; Infinity;
Novartis; Sanofi;
YM BioSciences
REFERENCES
1. Bain BJ, Gilliland DG, Horny H-P, et al. Myeloproliferative Neoplasms
and Myeloid and lymphoid neoplasms with eosinophilia and abnormalities of
PDGFRA, PDGFRB, or FGFR1. In: Swerdlow S, Harris NL, Stein H, Jaffe
ES, Theile J, Vardiman JW (eds). World Health Organization Classification of
Tumours. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France: IARC Press, 2008:30-73.
2. Dameshek W. Some speculations on the myeloproliferative syndromes.
Blood. 1951;6:372-375.
3. Adamson JW, Fialkow PJ, Murphy S, et al. Polycythemia vera: Stem-cell
and probable clonal origin of the disease. N Engl J Med. 1976;295:913-916.
4. Prchal JF, Axelrad AA. Bone-marrow responses in polycythemia vera.
N Engl J Med. 1974;290:1382.
5. Saharinen P, Silvennoinen O. The pseudokinase domain is required for
suppression of basal activity of Jak2 and Jak3 tyrosine kinases and for
cytokine-inducible activation of signal transduction. J Biol Chem. 2002;277:
47954-47963.
6. Baxter EJ, Scott LM, Campbell PJ, et al. Acquired mutation of the
tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet. 2005;
365:1054-1061.
7. Levine R, Wadleigh M, Cools J, et al. Activating mutation in the tyrosine
kinase JAK2 in polycythemia vera, essential thrombocythemia, and myeloid
metaplasia with myelofibrosis. Cancer Cell. 2005;7:387-397.
8. Kralovics R, Passamonti F, Buser AS, et al. A gain-of-function mutation
of JAK2 in myeloproliferative disorders. N Engl J Med. 2005;352:1779-1790.
9. James C, Ugo V, Le Couedic JP, et al. A unique clonal JAK2 mutation
leading to constitutive signalling causes polycythaemia vera. Nature. 2005;

434:1144-1148.
10. Yan D, and Mohi G. Critical Role for Stat5 in the Initiation and
Maintenance of Polycythemia Vera in a Jak2V617F Knock-In Mouse Model.
Blood (ASH Annual Meeting Abstracts). 2011;118:121.
11. Lacout C, Pisani DF, Tulliez M, et al. JAK2V617F expression in murine
hematopoietic cells leads to MPD mimicking human PV with secondary
myelofibrosis. Blood. 2006;108:1652-1660.
12. Xing S, Wanting T, Zhao W, et al. Transgenic expression of JAK2V617F
causes myeloproliferative disorders in mice. Blood. 2008;111:5109-5117.
13. Chen E, Beer PA, Godfrey AL, et al. Distinct clinical phenotypes
associated with JAK2V617F reflect differential STAT1 signaling. Cancer Cell.
2010;18:524-535.
14. Delhommeau F, Dupont S, Tonetti C, et al. Evidence that the JAK2
G1849T (V617F) mutation occurs in a lymphomyeloid progenitor in polycythemia vera and idiopathic myelofibrosis. Blood. 2007;109:71-77.
15. Jamieson CH, Gotlib J, Durocher JA, et al. The JAK2 V617F mutation
occurs in hematopoietic stem cells in polycythemia vera and predisposes
toward erythroid differentiation. Proc Natl Acad Sci U S A. 2006;103:62246229.
16. Wernig G, Mercher T, Okabe R, et al. Expression of Jak2V617F causes
a polycythemia vera-like disease with associated myelofibrosis in a murine
bone marrow transplant model. Blood. 2006;107:4274-4281.
17. Bumm T, Elsea C, Corbin A, et al. Characterization of murine
JAK2V617F-positive myeloproliferative disease. Cancer Res. 2006;66:1115611165.
417
NGUYEN AND GOTLIB

18. Theocharides A, Boissinot M, Girodon F, et al. Leukemic blasts in
transformed JAK2-V617F-positive myeloproliferative disorders are frequently negative for the JAK2-V617F mutation. Blood. 2007;110:375-379.
19. Lambert JR, Everington T, Linch DC, et al. In essential thrombocythemia, multiple JAK2 V617F clones are present in most mutant-positive
patients: A new disease paradigm. Blood. 2009;114:3018-3023.
20. Scott LM, Tong W, Levine R, et al. JAK2 exon 12 mutations in
polycythemia vera and idiopathic erythrocytosis. N Engl J Med. 2007;356:
459-468.
21. Scott LM. The JAK2 exon 12 mutations: A comprehensive review. Am J
Hematol. 2011;86:668-676.
22. Pikman Y, Lee BH, Mercher T, et al. MPLW515L is a novel somatic
activating mutation in myelofibrosis with myeloid metaplasia. PLoS Med.
2006;3:e270.
23. Pardanani A, Levine R, Lasho T, et al. MPL515 mutations in myeloproliferative and other myeloid disorders: A study of 1182 patients. Blood.
2006;108:3472-3476.
24. Dunbar AJ, Gondek LP, OKeefe CL, et al. 250K single nucleotide
polymorphism array karyotyping identifies acquired uniparental disomy and
homozygous mutations, including novel missense substitutions of c-Cbl, in
myeloid malignancies. Cancer Res. 2008;68:10349-10357.
25. Grand FH, Hidalgo-Curtis CE, Ernst T, et al. Frequent CBL mutations
associated with 11q acquired uniparental disomy in myeloproliferative neoplasms. Blood. 2009;113:6182-6192.
26. Loh ML, Sakai DS, Flotho C, et al. Mutations in CBL occur frequently
in juvenile myelomonocytic leukemia. Blood. 2009;114:1859-1863.
27. Sanada M, Suzuki T, Shih LY, et al. Gain-of-function of mutated C-CBL
tumour suppressor in myeloid neoplasms. Nature. 2009;460:904-908.
28. Velazquez L, Cheng AM, Fleming HE, et al. Cytokine signaling and
hematopoietic homeostasis are disrupted in Lnk-deficient mice. J Exp Med.
2002;195:1599-1611.
29. Oh ST, Simonds EF, Jones C, et al. Novel mutations in the inhibitory
adaptor protein LNK drive JAK-STAT signaling in patients with myeloproliferative neoplasms. Blood. 2010;116:988-992.
30. Pardanani A, Lasho T, Finke C, et al. LNK mutation studies in
blast-phase myeloproliferative neoplasms, and in chronic-phase disease with
TET2, IDH, JAK2 or MPL mutations. Leukemia. 2010;24:1713-1718.
31. Lasho TL, Pardanani A, Tefferi A. LNK mutations in JAK2 mutationnegative erythrocytosis. N Engl J Med. 2010;363:1189-1190.
32. Delhommeau F, Dupont S, Della Valle V, et al. Mutation in TET2 in
myeloid cancers. N Engl J Med. 2009;360:2289-2301.
33. Abdel-Wahab O, Mullally A, Hedvat C, et al. Genetic characterization
of TET1, TET2, and TET3 alterations in myeloid malignancies. Blood.
2009;114:144-147.
34. Tahiliani M, Koh KP, Shen Y, et al. Conversion of 5-methylcytosine to
5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science. 2009;324:930-935.
35. Figueroa ME, Abdel-Wahab O, Lu C, et al. Leukemic IDH1 and IDH2
mutations result in a hypermethylation phenotype, disrupt TET2 function,
and impair hematopoietic differentiation. Cancer Cell. 2010;18:553-567.
36. Li Z, Cai X, Cai C, et al. Deletion of Tet2 in mice leads to dysregulated
hematopoietic stem cells and subsequent development of myeloid malignancies. Blood. 2011;118:4509-4518.
37. Moran-Crusio K, Reavie L, Shih A, et al. Tet2 loss leads to increased
hematopoietic stem cell self-renewal and myeloid transformation. Cancer
Cell. 2011;20:11-24.
38. Ley TJ, Ding L, Walter MJ, et al. DNMT3A mutations in acute myeloid
leukemia. N Engl J Med. 2010;363:2424-2433.
39. Stegelmann F, Bullinger L, Schlenk RF, et al. DNMT3A mutations in
myeloproliferative neoplasms. Leukemia. 2011;25:1217-1219.
40. Abdel-Wahab O, Pardanani A, Rampal R, et al. DNMT3A mutational
analysis in primary myelofibrosis, chronic myelomonocytic leukemia and
advanced phases of myeloproliferative neoplasms. Leukemia. 2011;25:12191220.
41. Challen GA, Sun D, Jeong M, et al. Dnmt3a is essential for hematopoietic stem cell differentiation. Nat Genet. 2011;44:23-31.
42. Tefferi A, Lasho TL, Abdel-Wahab O, et al. IDH1 and IDH2 mutation
studies in 1473 patients with chronic-, fibrotic- or blast-phase essential
418
thrombocythemia, polycythemia vera or myelofibrosis. Leukemia. 2010;24:

1302-1309.
43. Xu W, Yang H, Liu Y, et al. Oncometabolite 2-hydroxyglutarate is a
competitive inhibitor of alpha-ketoglutarate-dependent dioxygenases. Cancer
Cell. 2011;19:17-30.
44. Ernst T, Chase AJ, Score J, et al. Inactivating mutations of the histone
methyltransferase gene EZH2 in myeloid disorders. Nat Genet. 2010;42:722726.
45. Abdel-Wahab O, Pardanani A, Patel J, et al. Concomitant analysis of
EZH2 and ASXL1 mutations in myelofibrosis, chronic myelomonocytic leukemia and blast-phase myeloproliferative neoplasms. Leukemia. 2011;25:12001202.
46. Score J, Hidalgo-Curtis C, Jones AV, et al. Inactivation of polycomb
repressive complex 2 components in myeloproliferative and myelodysplastic/
myeloproliferative neoplasms. Blood. 2012;119:1208-1213.
47. Guglielmelli P, Biamonte F, Score J, et al. EZH2 mutational status
predicts poor survival in myelofibrosis. Blood. 2011;118:5227-5234.
48. Scheuermann JC, de Ayala Alonso AG, Oktaba K, et al. Histone H2A
deubiquitinase activity of the Polycomb repressive complex PR-DUB. Nature.
2010;465:243-247.
49. Gurvich N, Perna F, Farina A, et al. L3MBTL1 polycomb protein, a
candidate tumor suppressor in del(20q12) myeloid disorders, is essential for
genome stability. Proc Natl Acad Sci U S A. 2010;107:22552-22557.
50. Min J, Allali-Hassani A, Nady N, et al. L3MBTL1 recognition of monoand dimethylated histones. Nat Struct Mol Biol. 2007;14:1229-1230.
51. Perna F, Gurvich N, Hoya-Arias R, et al. Depletion of L3MBTL1
promotes the erythroid differentiation of human hematopoietic progenitor
cells: Possible role in 20q- polycythemia vera. Blood. 2010;116:2812-2821.
52. Dawson MA, Bannister AJ, Gottgens B, et al. JAK2 phosphorylates
histone H3Y41 and excludes HP1alpha from chromatin. Nature. 2009;461:
819-822.
53. Liu F, Zhao X, Perna F, et al. JAK2V617F-mediated phosphorylation
of PRMT5 downregulates its methyltransferase activity and promotes myeloproliferation. Cancer Cell. 2011;19:283-294.
54. Jager R, Gisslinger H, Passamonti F, et al. Deletions of the transcription factor Ikaros in myeloproliferative neoplasms. Leukemia. 2010;24:12901298.
55. Harutyunyan A, Klampfl T, Cazzola M, et al. p53 lesions in leukemic
transformation. N Engl J Med. 2011;364:488-490.
56. Silva M, Richard C, Benito A, et al. Expression of Bcl-x in erythroid
precursors from patients with polycythemia vera. N Engl J Med. 1998;338:
564-571.
57. Plo I, Nakatake M, Malivert L, et al. JAK2 stimulates homologous
recombination and genetic instability: Potential implication in the heterogeneity of myeloproliferative disorders. Blood. 2008;112:1402-1412.
58. Zhao R, Follows GA, Beer PA, et al. Inhibition of the Bcl-xL deamidation pathway in myeloproliferative disorders. N Engl J Med. 2008;359:27782789.
59. Capello D, Deambrogi C, Rossi D, et al. Epigenetic inactivation of
suppressors of cytokine signalling in Philadelphia-negative chronic myeloproliferative disorders. Br J Haematol. 2008;141:504-511.
60. Hookham MB, Elliott J, Suessmuth Y, et al. The myeloproliferative
disorder-associated JAK2 V617F mutant escapes negative regulation by
suppressor of cytokine signaling 3. Blood. 2007;109:4924-4929.
61. Jones AV, Chase A, Silver RT, et al. JAK2 haplotype is a major risk
factor for the development of myeloproliferative neoplasms. Nat Genet.
2009;41:446-449.
62. Kilpivaara O, Mukherjee S, Schram AM, et al. A germline JAK2 SNP is
associated with predisposition to the development of JAK2(V617F)-positive
myeloproliferative neoplasms. Nat Genet. 2009;41:455-459.
63. Olcaydu D, Harutyunyan A, Jager R, et al. A common JAK2 haplotype
confers susceptibility to myeloproliferative neoplasms. Nat Genet. 2009;41:
450-454.
64. Mead AJ, Rugless MJ, Jacobsen SE, et al. Germline JAK2 mutation in
a family with hereditary thrombocytosis. N Engl J Med. 2012;366:967-969.
65. Etheridge L, Corbo LM, Kaushansky K, et al. A novel activating JAK2
mutation, JAK2R564Q, causes familial essential thrombocytosis (fET) via
mechanisms distinct from JAK2V617F. Blood (ASH Annual Meeting Abstracts). 2011;118:123.
C l a s s i fi c a t i o n o f M y e l o p r o l i f e r a t i v e
Neoplasms and Prognostic Factors
By Francesco Passamonti, MD
Overview: Myeloproliferative neoplasms (MPNs) are currently diagnosed according to the World Health Organization
(WHO) criteria. Molecular profiling should include the analysis
of JAK2 V617F (first, exon 12 only in V617F-negative polycythemia vera [PV]) and MPL mutations (in V617F-negative
essential thrombocythemia [ET] and myelofibrosis [MF]). For
patients with PV and ET, the risk stratification of low- and
high-risk disease requires only two parameters: older than
age 60 and prior history of thrombosis. Additionally, it might
be important to monitor leukocyte count and know the mutational profile.
PNs INCLUDE different entities summarized in

Table 1.1 This article will focus on classical BCRABL1-negative MPNs, including ET, PV, and primary myelofibrosis (PMF).
Classification of MPNs
MPNs are clonal stem cell neoplasms found to share some

relevant biologic features. Recent molecular advances have
demonstrated that the abnormal myeloproliferation arises
from constitutively active signal transduction pathways,
caused by specific mutations affecting protein tyrosine kinases or related molecules. Current classification of MPNs is
based on the WHO criteria established in the 2008 revision
(Tables 2-4), which, besides simple clinical parameters,
convey two novelties: molecular genetics and histopathology.
Complete Blood Count (CBC) for MPN Diagnosis
PV is suspected in men with hemoglobin levels greater

than 18.5 g/dL or 16.5 g/dL in women or hemoglobin levels
greater than 17 g/dL in men or 15 g/dL in women if
associated with a documented and sustained increase of at
least 2 g/dL from an individuals baseline value. It is no
longer necessary to use red cell mass measurement to
exclude secondary polycythemia. In approximately 20% to
40% of PV cases, leukocytosis and/or thrombocytosis might
be present. Thrombocytosis remains a criterion for the
diagnosis of ET. In the revised WHO criteria, the platelet
threshold for the diagnosis of ET was lowered to 450
109/L. This level of thrombocytosis is not specific for ET and
can be secondary to other conditions (e.g., iron deficiency,
trauma, infection, inflammation, bleeding), which must be
excluded first. Thrombocytosis can also be present in chronic
myeloid leukemia (CML; test for BCR-ABL1 fusion gene), in
refractory anemia with ringed sideroblasts associated with
marked thrombocytosis (RARS-T; signs of dyserythropoiesis
at morphological examination), and also in PV cases in
which erythrocytosis is not evident because of relative iron
deficiency. Concerning MF, anemia might be accompanied
with leukopenia/leukocytosis and/or thrombocytopenia/
thrombocytosis. Peripheral blood smear should be reviewed
in all MPN cases as microcytic red blood cells in PV/ET
might be a sign of iron deficiency, leukoerythroblastosis is
present in almost all PMF cases, and myeloid progenitors
are typical of CML.
Survival of patients with MF is defined by the International

Prognostic Scoring System (IPSS) model at diagnosis and the
Dynamic IPSS (DIPSS) anytime during the course of the
disease. The IPSS and the DIPSS are based on patient age
older than age 65, presence of constitutional symptoms,
hemoglobin level less than 10 g/dL, leukocyte count greater
than 25 10 9 /L, and circulating blast cells 1% or greater. The
DIPSS-plus adds critical prognostic information and suggests
also considering cytogenetic categories, platelet count, and
red blood cell transfusion need.
Molecular Genetics for MPN Diagnosis
Screening assays for MPN genetics are not standardized,

and the possibility of false-positives or false-negatives can be
an issue when using highly sensitive allele-specific assays
and in cases of low mutant allele burden. The JAK2 V617F
mutation is found in more than 95% of patients with PV and
in nearly 50% to 60% of those with ET and PMF. It has also
been found in other MPNs but not in nonmyeloid malignancies or in cases of secondary polycythemia. Therefore, JAK2
V617F is a sensitive diagnostic marker of PV. Although low
JAK2 V617F allele burden is more distinctive of ET than of
PV and MF, mutational load measurement is not useful for
diagnostic purposes. Less than 3% of patients with PV carry
exon 12 mutations of JAK2, and, as different mutations do
not confer different phenotypes, a screening test highresolution melting is adequate.2 Exon 12 mutations of JAK2
should be screened in case of JAK2 V617F-negative erythrocytosis with low erythropoietin level. A small portion of
patients with ET and PMF who lack a mutated JAK2 may
have activating mutations in MPL (mainly W515K/L). For
the time being, the study of JAK2 and MPL mutations is to
be included in the diagnostic workup of MPN, while all other
less prevalent mutations (LNK, NF1, cCBL, SOCS1, TET2,
EZH2, ASXL1, IDH1/2, DNMT3A) cannot enter into the
diagnostic process.2
Serum Erythropoietin
A simple and timeless PV test is serum erythropoietin

dosage, which can discriminate PV (low level) from secondary erythrocytosis (high level). In the absence of any JAK2
mutation or in case of unavailability of the JAK2 test, low
erythropoietin levelsif accompanied by MPN-consistent
bone marrow featuresis of diagnostic value for PV.
Bone Marrow Histopathology
In the WHO classification, bone marrow histopathology

has assumed a critical diagnostic role, since the distinction
From the Division of Hematology, Department of Internal Medicine, University Hospital

Ospedale di Circolo e Fondazione Macchi, Varese, Italy.
Address reprint requests to Francesco Passamonti, MD, Division of Hematology, Department of Internal Medicine, University Hospital Ospedale di Circolo e Fondazione Macchi,
Viale L. Borri 57, 21100 Varese, Italy; email: francesco.passamonti@ospedale.varese.it.
1092-9118/10/1-10
419
FRANCESCO PASSAMONTI
between ET, prefibrotic PMF, and PMF requires bone marrow evaluation. In PV, marrow is usually hypercellular for
age with trilineage growth and prominent erythroid, granulocytic, and megakaryocytic proliferation, and bone marrow fibrosis is generally absent, even though fibrosis may be
found in approximately 15% of cases. In ET, bone marrow
examination reveals no or only a slight increase in agematched cellularity, no significant increase in granulo- and
erythropoiesis or prominent large to giant mature megakaryocytes with hyperlobulated or deeply folded nuclei,
dispersed or loosely clustered in the marrow space. In PMF,
the bone marrow demonstrates marked megakaryocyte
proliferation with atypia described as small to large megakaryocytes with aberrant nuclear/cytoplasmic ratio and
hyperchromatic and irregularly folded nuclei with dense
clustering, which is accompanied by reticulin and/or collagen fibrosis. When reticulin fibrosis is absent, a diagnosis of
prefibrotic PMF needs to be considered. Prefibrotic PMF
presents with marked increase in age-matched cellularity,
pronounced granulopoiesis precursors, reduced erythroid
precursors, and dense or loose clustering of medium-sized to
giant megakaryocytes. The histopathologic distinction between ET and prefibrotic PMF is of prognostic value, as the
latter disease is associated with a higher risk of evolution
into MF and acute myeloid leukemia (AML) and worse
survival.3
Thus, as thrombosis is the most frequent complication in PV

and ET, there is a comprehensible rationale for stratifying
these patients according to the risk of thrombosis.
Risk Factors for Thrombosis
KEY POINTS
There is a general agreement among investigators to

consider patient age older than 60 at diagnosis and presence
of vascular events in the patients history as the two prognostic factors for thrombosis in PV and ET (Table 5).
Therefore, patients with one or two of these risk factors at
diagnosis are considered at high risk, while those with none
of them are considered at low risk. This risk classification
has an effect on the therapeutic approach for patients with
PV and ET.
Cardiovascular risk factors (arterial hypertension, smoking, hypercholesterolemia, diabetes) do not enter in the risk
stratification of ET and PV, but an appropriate strategy of
prevention and management is recommended.5
Concerning the CBC, hematocrit levels up to 50% and
high platelet count are not associated with thrombosis.5 On
the other hand, extreme thrombocytosis might provoke
excess bleeding from acquired von Willebrand disease, and
platelet counts over 1,500 109/L should be considered as a
criterion to start cytoreduction in ET and PV. The relationship between leukocytosis and thrombosis was largely studied in ET. A retrospective analysis seems to indicate that
patients with low-risk ETmost frequently not treated
have a higher risk of thrombosis in the presence of leukocytosis at diagnosis or developed during the disease course.6
Concerning patients with PV, a subsequent analysis of the
ECLAP trial showed that patients with a leukocyte count
exceeding 15 109/L had a significant increase of myocardial infarction compared with those patients with leukocyte
counts below 10 109/L (p 0.017).5
Many studies recently evaluated the relationship between
the presence of the JAK2 V617F mutation and thrombosis.7
Meta-analyses suggest a correlation between the JAK2
V617F mutation and thrombosis in ET. Among patients with
JAK2 V617F-positive ET, those with higher allele burden
seem to have a higher risk of thrombosis, while a prospective
observation of PV did not disclose any relationship.8,9
Regarding patients with exon 12 mutations of JAK2, the
proposed risk stratification of PV may be applied.10 Rare
mutations involving MPL and TET2 genes seem not to effect
events in patients with ET.2
Prediction of PV and ET Disease Evolution
Prognostication in PV and ET
Life expectancy of patients with PV is reduced when

compared with that of the general population, while the
survival of patients with ET is not significantly shortened
(831 patients with PV and ET followed for a median of 10
years).4
Disease-related complications affecting survival are
mainly vascular events (thrombosis and hemorrhage) and
transformation to MF or AML. In PV, the incidence of
thrombosis was estimated at 18 1,000 person years and
for MF and AML at 5 1,000 person years.4 In ET, the
incidence of thrombosis, MF, and AML was 12, 1.6, and
1.2 1,000 person years.4 The ECLAP study also indicated
that cardiovascular mortality in PV accounted for 45% of all
deaths and hematologic transformation accounted for 13%.5
420
Myeloproliferative neoplasms (MPNs) are diagnosed

according to the 2008 World Health Organization
criteria.
Molecular analysis in MPNs should include JAK2
V617F first, exon 12 only in V617F-negative polycythemia vera (PV) and MPL mutations in V617Fnegative essential thrombocythemia (ET) and
myelofibrosis (MF).
Bone marrow biopsy is mandatory to distinguish ET
from primary myelofibrosis.
Risk stratification of PV and ET is based on patient
age older than age 60 and prior history of thrombosis.
Risk stratification of MF is based on the IPSS prognostic model at diagnosis and the DIPSS model
during follow-up.
Approximately 10% of patients with PV evolve in post-PV

MF with progressive splenomegaly, MF-related symptoms,
anemia, and leukocytosis.11 MF evolution is difficult to
predict, but leukocyte count greater than 15 109/L has
been documented as a risk factor for disease evolution.12 In
a prospective study that included 338 patients with PV, an
allele burden greater than 50% implied a higher risk of MF
evolution.9 When PV progresses to post-PV MF, survival
worsens and might be predicted using a dynamic prognostic
score based on three risk factors: hemoglobin level less than
10 g/dL, leukocyte count greater than 30 109/L, and
platelet count less than 100 109/L.12 The low-risk group
includes patients without risk factors, while higher-risk
categories include patients with one, two, or three risk
factors. When a single patient acquires one risk factor, his or
her survival worsens 4.2-fold. Despite the availability of this
CLASSIFICATION AND PROGNOSIS OF MPNs

Table 1. Classification of Chronic Myeloid Neoplasms According to the World Health Organization in 2008
Diseases
Main criteria
Chronic myelogenous leukemia, BCRABL1-positive
Positivity for the BCR-ABL1 fusion gene
Chronic neutrophilic leukemia
PB leukocytosis, WBC 25 109/L

- Segmented PMNs and band forms are 80% of WBCs
- Immature granulocytes 10% of WBCs
- Myeloblasts 1% of WBCs
Hypercellular BM biopsy (myeloblasts 5% of NMCs)
Hepatosplenomegaly
No evidence of reactive neutrophila
No Philadelphia chromosome or BCR-ABL1 fusion gene; no rearrangement of PDGFRA, PDGFRB, or FGFR1; no evidence of PV, ET, or PMF;
no evidence of MDS or MDS/MPN
Polycythemia vera
See Table 2
Primary myelofibrosis
See Table 4
Essential thrombocythemia
See Table 5
Chronic eosinophilic leukemia, NOS
Eosinophils 1.5 109/L

No Philadelphia chromosome or BCR-ABL1 fusion gene or other MPN (PV, ET, or PMF) or MDS/MPN (CMML or aCML)
No t(5;12)(q31-35;p13) or other rearrangement of PDGFRB
No FIP1L1-PDGFRA fusion gene or other rearrangement of PDGFRA
No rearrangement of FGFR1
PB and BM blasts 20%, no inv(16)(p13.1;q22) or t(16;16)(p13.1;q22) or other feature diagnostic of AML
Presence of a clonal cytogenetic or molecular genetic abnormality, or blast cells 2% in the PB or 5% in the BM
Cutaneous mastocytosis: mast cell infiltrate in the skin (criteria for SM not met)
SM: mast cell infiltrate in BM and/or other extracutaneous organs (minor diagnostic criteria include mast cell atypia in 25% of cells,
detection of a mutation at codon 816 of KIT, an aberrant mast cell immunophenotype and elevated serum triptase levels)
Clinical, laboratory, and morphologic features of an MPN without meeting the diagnostic criteria for one specific MPN
Mastocytosis
MPN, unclassifiable
Myeloid and lymphoid neoplasms associated with eosinophilia and abnormalities of PDGFRA, PDGFRB, or FGFR1
Myeloid and lymphoid neoplasms
associated with eosinophilia and
abnormalities of PDGFRA, PDGFRB,
or FGFR1
An MPN with prominent eosinophilia (cases presenting with AML or ALL with eosinophilia are also included) AND presence of a FIP1L1PDGFRA fusion gene or presence of t(5;12)(q31q33;p12) or a variant translocation or demonstration of PDGFRB fusion gene or of
rearrangement of PDGFRB or presence of t(8;13)(p11;q12) or a variant translocation leading to FGFR1 rearrangement
MDS/MPN
9
Chronic myelomonocytic leukemia
Persistent PB monocytosis 1 10 /L
No Philadelphia chromosome or BCR-ABL1 fusion gene
No rearrangement of PDGFRA or PDGFRB
20% blasts (including myeloblasts, monoblasts, and promonocytes) in PB and BM
Dysplasia in 1 myeloid lineageif no/minimal myelodysplasia:
- Presence of an acquired clonal cytogenetic or molecular genetic abnormality OR
- Persistent monocytosis (3 mo) and exclusion of all other causes of monocytosis
Juvenile myelomonocytic leukemia
Persistent PB monocytosis 1 109/L

20% blasts (including promonocytes) in PB and BM
Two of the following: hemoglobin F increased for age; immature granulocytes in PB; WBC count 10 109/L, clonal chromosomal
abnormality (may be monosomy 7); GM-CSF hypersensitivity of myeloid progenitors in vitro
Atypical chronic myeloid leukemia, BCRABL1-negative
PB leukocytosis (WBCs 13 109/L) because of increased numbers of neutrophils and their precursors (which constitute 10% of leukocytes)
with prominent dysgranulopoiesis
No rearrangement of PDGFRA or PDGFRB
Minimal absolute basophilia; basophils usually 2% of leukocytes
No or minimal absolute monocytosis; monocytes 10% of WBCs
Hypercellular BM with granulocytic proliferation and dysplasia, dysplasia in the other lineages
20% blasts in PB and BM
MDS/MPN, unclassifiable
- Provisional entity: refractory anemia
with ring sideroblasts associated with
marked thrombocytosis
Clinical, laboratory and morphologic features of an MDS AND

Prominent myeloproliferative features, e.g., platelet count 450 109/L associated with megakaryocytic proliferation, or WBC count 13
109/L, splenomegaly AND
No preceding MDS or MPN, no history of cytotoxic or growth factor therapy, no Philadelphia chromosome or BCR-ABL1 fusion gene, no
rearrangement of PDGFRA, PDGFRB or FGFR1, and no isolated del(5q), t(3;3)(q21;q26) or inv(3)(q21;q26) OR
De novo disease with mixed myeloproliferative and myelodysplastic features that cannot be assigned to any of the MDS, MPN, or MDS/MPN
categories
RARS-T: anemia, ring sideroblasts 15% of erythroid precursors, platelet count 450 109/L, large and atypical BM megakaryocytes, no PB
blasts, 5% BM blasts AND NO BCR-ABL1 fusion gene, isolated del(5q), t(3;3)(q21;q26), inv(3)(q21;q26)
Myelodysplastic syndromes
Myelodysplastic syndrome
PB cytopenia(s), dysplasia in 1 myeloid cell line, PB and BM blasts 20%, BM ring sideroblastsseveral MDS categories exist: refractory
cytopenias with unilineage dysplasia (refractory anemia, refractory neutropenia, refractory thrombocytopenia), refractory anemia with ring
sideroblasts, refractory cytopenia with multilineage dysplasia, refractory anemia with excess blasts type 1 or 2, myelodysplastic syndrome
unclassified, MDS associated with isolated del(5q)
Abbreviations: PB, peripheral blood; WBCs, white blood cells; PMNs, polymorphonucleated cells; BM, bone marrow; NMCs, nucleated marrow cells; PV: polycythemia
vera; ET, essential thrombocythemia; PMF, primary myelofibrosis; MDS, myelodysplastic syndrome; MDS/MPN: myelodysplastic/myeloproliferative neoplasm; NOS, not
otherwise specified; MPN, myeloproliferative neoplasms; CMML, chronic myelomonocytic leukemia; aCML, atypical chronic myeloid leukemia; AML, acute myeloid
leukemia; SM, systematic mastocytosis; ALL, acute lymphoblastic leukemia; GM-CSF, granulocyte macrophage colony-stimulating factor; RARS-T, refractory anemia
with ring sideroblasts associated with marked thrombocytosis.
421
Table 2. WHO Criteria for Diagnosis of Polycythemia Vera
Major Criteria
Hemoglobin 18.5 g/dL in men, 16.5 g/dL in women, or other evidence of
increased red cell volume
Presence of JAK2 V617F or other functionally similar mutation (JAK2 exon 12
mutation)
Minor Criteria
Bone marrow biopsy showing hypercellularity for age with trilineage
myeloproliferation
Serum erythropoietin level below the normal reference range
Endogenous erythroid colony formation in vitro
Abbreviations: WHO, World Health Organization; PV, polycythemia vera.
a
Diagnosis of PV requires meeting either both major criteria and one minor
criterion or the first major criterion and two minor criteria.
disease-specific model, many investigators are more confident applying prognostic systems developed in PMF in this
patient subset.13,14
Evolution to AML is a rare event, and the predictors
include advanced age (sign of genomic instability) and a high
leukocyte count (sign of myeloproliferation). Concerning the
role of chemotherapy, a recent population-based study on
11,039 MPNs proved that 25% of patients with post-MPN
AML were never exposed to cytotoxic drugs and that hydroxyurea at any dose is not associated with an increased
risk of AML, whereas an increasing cumulative dose of
alkylators is.15
Specific Clinical Situations in PV and ET
A retrospective study on the outcome of 311 surgical

interventions for patients with PV and ET showed that 7.7%
of them were complicated by fatal arterial or venous thromboses and 7.3% by fatal major hemorrhage within 3 months
from the procedure.16 Although no prognostic factors could
be identified to predict postsurgery outcome, these data
should mandate watchfulness in the surgical management
of these patients.
When treating patients with ET (generally younger than
those with PV and PMF), a potential issue is the approach to
pregnancy. A study on 103 pregnancies in 62 women with
ET reported a 64% live birth rate, with 51% of pregnancies
being uneventful.17 Maternal complications occurred in 9%
of cases, while fetal complications occurred in 40% of them.
Fetal loss was 3.4-fold higher than that of the general
population. The JAK2 V617F mutation was an independent
predictor of pregnancy complications.
Prognostication in MF
Among MPNs, PMF has the most heterogeneous clinical

presentation, which may encompass anemia, splenomegaly,
Table 3. WHO Criteria for Diagnosis of Essential
Thrombocythemia a
Sustained platelet count over 450 109/L
Bone marrow biopsy specimen showing proliferation mainly of the megakaryocytic
lineage with increased numbers of enlarged, mature megakaryocytes; no
significant increase or left-shift of neutrophil granulopoiesis or erythropoiesis
Not meeting WHO criteria for PV or PMF, BCRABL1-positive CML, MDS, or other
myeloid neoplasm
Demonstration of JAK2 V617F or other clonal marker or, in the absence of JAK2
V617F, no evidence of reactive thrombocytosis
Abbreviations: WHO, World Health Organization; PV, polycythemia vera; PMF,
primary myelofibrosis; CML, chronic myelogenous leukemia; MDS, myelodysplastic syndrome.
a
The diagnosis of ET requires meeting all four criteria.
422
Table 4. WHO criteria for Diagnosis of Primary Myelofibrosis a

Major Criteria
Presence of megakaryocyte proliferation and atypia, accompanied by either
reticulin or collagen fibrosis orin the absence of significant reticulin
fibrosisthe megakaryocyte changes must be accompanied by an increased
marrow cellularity characterized by granulocytic proliferation and often
decreased erythropoiesis (i.e., prefibrotic cellular-phase disease)
Not meeting WHO criteria for PV, CML, MDS, or other myeloid disorders
Demonstration of JAK2 V617F or other clonal marker (e.g., MPLW515K/L) or
in the absence of the above clonal markersno evidence of secondary bone
marrow fibrosis
Minor Criteria
Leukoerythroblastosis
Increased serum lactate dehydrogenase level
Anemia
Splenomegaly
Abbreviations: WHO, World Health Organization; PV, polycythemia vera; CML,
chronic myelogenous leukemia; MDS, myelodysplastic syndrome.
a
Diagnosis of PMF requires all three major criteria and two minor criteria.
leukocytosis or leukopenia, thrombocytosis or thrombocytopenia, and constitutional symptoms. Median survival in

PMF is estimated at 6 years, but it can range from a few
months to many years.13,14,18,19 Causes of death in MF
include bone marrow failure (severe anemia, bleeding from
thrombocytopenia, and infections from leukopenia) in 25%
to 30% of patients, AML transformation in 10% to 20% of
patients, cardiovascular complications in 15% to 20%, and
portal hypertension in 10%.
Risk Factors for Survival
The following were shown to be associated with poor

outcome in patients with PMF: advanced age, anemia, red
blood cell transfusion need, leukopenia, leukocytosis, thrombocytopenia, peripheral blast count, constitutional symptoms, hepatic myeloid metaplasia, decreased marrow
cellularity with higher degree of fibrosis, higher degree of
microvessel density, high number of circulating CD34positive cells, cytogenetic abnormalities, mutational profile,
and high level of proinflammatory cytokines (IL-8 and IL2R). Among these, some parameters require a more detailed
discussion.
Molecular Abnormalities. Within a large international
database, 345 patients had an available JAK2 mutational
status and no association was observed between the JAK2
status and survival.14 This result parallels other studies
including 199, 186, and 174 patients each that showed
no significant correlation between the presence of the
JAK2 V617F mutation and survival or leukemic transformation.20-22 However, there is no a general agreement on
this matter. Data seem to indicate that having a lower JAK2
V617F allele burden implies a worse survival.21,22 The
explanation for this association, however, differs as patients
mostly died from infections secondary to myelodepletion in
one study and from AML evolution in the other.21,22 A study
on 139 patients with MF receiving allogenic stem cell transplantation (ASCT) reported very intriguing results on the
association between allele burden reduction and post-ASCT
Table 5. Risk Categories in Essential Thrombocythemia
and Polycythemia Vera
Low risk
High risk
Age 60 and no history of thrombosis

Age 60 or history of thrombosis
CLASSIFICATION AND PROGNOSIS OF MPNs

Table 6. Score Values for the International Prognostic Scoring
System (IPSS) 14 and Dynamic International Prognostic
Scoring System (DIPSS) 13
Scores
Parameter
IPSS
DIPSS
Age 65
Hemoglobin 10 g/dL
Leukocyte count 25 109/L
Blast cells 1%
Constitutional symptoms
1
1
1
1
1
1
2
1
1
1
IPSS: score 0 for low risk, score 1 for intermediate-1 risk, score 2 for
intermediate-2 risk, score 3 for high risk; DIPSS: score 0 for low risk, score 1-2
for intermediate-1 risk, score 3-4 for intermediate-2 risk, score 5-6 for high risk.
outcome.23 Patients who abrogated the JAK2 mutation 6

months after ASCT had a significantly lower relapse rate,
shedding light on the likely clinical benefit of reduced V617F
allele burden (p 0.04).
Concerning less frequent mutations described so far in
MF, studies reported that mutations involving TET2 (approximately 15% to 30% of patients with MF) and MPL
(approximately 7% to 9% of patients with MF) genes seem
not to affect survival, while IDH1 (approximately 4% of
patients with MF) implies a worse survival and higher risk
of AML, as does EZH2 mutations (approximately 6% of
patients with MF).24-27 Nullizygosity for the JAK2 46/1
haplotype was associated with shortened survival, raising
the possibility that non-46/1 haplotypes are associated with
a biologically more aggressive phenotype.28
Chromosomal Abnormalities. An abnormal karyotype is
present in approximately 30% to 40% of MF and implies a
shorter survival.14,29 A study of 433 patients with MF
provided a two-tiered cytogenetic-risk stratification with a
respective 5-year survival of 8% (high risk) and 51% (low
risk).20 Cytogenetic profiles are discussed in Table 7.
Red Blood Cell Transfusion Dependency. Patients with
the JAK2 V617F mutation seem not as prone to anemia as
those carrying the MPL or TET2 mutations.24,25 The prognostic effect of red blood cell transfusion need was examined
in 254 consecutive patients.30 Median survival was 35
months for patients receiving red blood cell transfusions and
117 months for patients who were nontransfused.
Prognostic Models at Diagnosis
The first prognostic model used to stratify PMF was the

Lille score, which included hemoglobin level less than10
g/dL and leukocyte count less than 4 109/L or greater than
30 109/L.29 Patients were grouped into three categories
with median survivals of 93, 26, and 13 months.
More recently, the International Working Group on MPN
Research and Treatment (IWG-MRT) developed the IPSS to
predict survival in patients with PMF at diagnosis.14 The
IPSS was developed using a dataset of 1,054 patients with
PMF from seven international centers. Currently, the IPSS
Table 7. Score Values for the Dynamic International Prognostic

Scoring System-Plus (DIPSS-Plus)
Parameter
Score value
DIPSS intermediate-1
DIPSS intermediate-2
DIPSS high risk
Unfavorable karyotype*
Red blood cell need
Platelet 100 109/L
1
2
3
1
1
1
* Unfavorable karyotype: complex karyotype or sole or two abnormalities that

include 8, 7/7q-, i(17q), 5/5q-, 12p-, inv(3), or 11q23 rearrangement;
favorable karyotype: normal and diploid karyotype, sole or two abnormalities not
included in the unfavorable karyotype category.
DIPSS-plus: score 0 for low risk, score 1 for intermediate-1 risk, score 2-3 for
intermediate-2 risk, score 4-6 for high risk.
is the prognostic scoring system used for diagnosis in clinical

practice. Table 6 illustrates risk factors and the score system. According to the four risk categories, median survivals
were 135, 95, 48, and 27 months.
A composite prognostic model was developed by the University of Texas M. D. Anderson Cancer Center involving
256 patients with PMF at diagnosis.19 The risk factors were
unfavorable karyotype, hemoglobin level less than 10 g/dL,
platelet count less than 100 109/L, and performance
status greater than 1. Overall survival ranged from 7 to 69
months.
Dynamic Prognostic Models
Dynamic prognostic models are based on the knowledge

that the acquisition of additional risk factors during the
disease course may substantially modify patients outcome.
The DIPSS Model. Among 1,054 patients evaluated to
develop the IPSS model, 525 had adequate follow-up data for
a time-dependent analysis aimed at the definition of the
DIPSS.13 The DIPSS is based on the same factors as the
IPSS and the score system is illustrated in Table 6. Median
survival was not reached in patients at low risk; it was 14.2
years in intermediate-1, 4 years in intermediate-2, and 1.5
years in patients at high risk.
The DIPPS-Plus Model. The DIPSS-plus is a refinement
of the DIPSS, which includes additional variables as illustrated in Table 7. DIPSS-plus includes four categories with
median survivals of 185, 78, 35, and 16 months, respectively.18
Conclusion
MPN diagnosis must be done according to the 2008 WHO

criteria. Molecular analysis should include JAK2 and MPL
mutations. The risk stratification for patients with PV and
ET requires only two parameters: age older than 60 and
prior history of thrombosis. As for PMF, IPSS at diagnosis
and DIPSS anytime during disease course easily define
survival of these patients very adequately.
Author
Francesco Passamonti*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
423
REFERENCES
1. Tefferi A, Thiele J, Orazi A, et al. Proposals and rationale for revision of
the World Health Organization diagnostic criteria for polycythemia vera,
essential thrombocythemia, and primary myelofibrosis: Recommendations
from an ad hoc international expert panel. Blood. 2007;110:1092-1097.
2. Tefferi A. How I treat myelofibrosis. Blood. 2011;117:3494-3504.
3. Barbui T, Thiele J, Passamonti F, et al. Survival and disease progression
in essential thrombocythemia are significantly influenced by accurate morphologic diagnosis: An international study. J Clin Oncol. 2011;29:3179-3184.
4. Passamonti F, Rumi E, Pungolino E, et al. Life expectancy and prognostic factors for survival in patients with polycythemia vera and essential
thrombocythemia. Am J Med. 2004;117:755-761.
5. Landolfi R, Marchioli R, Kutti J, et al. Efficacy and safety of low-dose
aspirin in polycythemia vera. N Engl J Med. 2004;350:114-124.
6. Carobbio A, Antonioli E, Guglielmelli P, et al. Leukocytosis and risk
stratification assessment in essential thrombocythemia. J Clin Oncol. 2008;
26:2732-2736.
7. Passamonti F. Prognostic factors and models in polycythemia vera,
essential thrombocythemia, and primary myelofibrosis. Clin Lymphoma Myeloma Leuk. 2011;suppl 1:S25-S27.
8. Vannucchi AM, Antonioli E, Guglielmelli P, et al. Clinical profile of
homozygous JAK2 617VF mutation in patients with polycythemia vera or
essential thrombocythemia. Blood. 2007;110:840-846.
9. Passamonti F, Rumi E, Pietra D, et al. A prospective study of 338
patients with polycythemia vera: The impact of JAK2 (V617F) allele burden
and leukocytosis on fibrotic or leukemic disease transformation and vascular
complications. Leukemia. 2010;24:1574-1579.
10. Passamonti F, Elena C, Schnittger S, et al. Molecular and clinical
features of the myeloproliferative neoplasm associated with JAK2 exon 12
mutations. Blood. 2011;117:2813-2816.
11. Barosi G, Mesa RA, Thiele J, et al. Proposed criteria for the diagnosis
of post-polycythemia vera and post-essential thrombocythemia myelofibrosis:
A consensus statement from the International Working Group for Myelofibrosis Research and Treatment. Leukemia. 2008;22:437-438.
12. Passamonti F, Rumi E, Caramella M, et al. A dynamic prognostic model
to predict survival in post-polycythemia vera myelofibrosis. Blood. 2008;111:
3383-3387.
13. Passamonti F, Cervantes F, Vannucchi AM, et al. A dynamic prognostic
model to predict survival in primary myelofibrosis: A study by the IWG-MRT
(International Working Group for Myeloproliferative Neoplasms Research
and Treatment). Blood. 2010;115:1703-1708.
14. Cervantes F, Dupriez B, Pereira A, et al. New prognostic scoring system
for primary myelofibrosis based on a study of the International Working
Group for Myelofibrosis Research and Treatment. Blood. 2009;113:2895-2901.
15. Bjorkholm M, Derolf AR, Hultcrantz M, et al. Treatment-related risk
factors for transformation to acute myeloid leukemia and myelodysplastic
syndromes in myeloproliferative neoplasms. J Clin Oncol. 2011;29:2410-2415.
16. Ruggeri M, Rodeghiero F, Tosetto A, et al. Postsurgery outcomes in
424
patients with polycythemia vera and essential thrombocythemia: A retrospective survey. Blood. 2008;111:666-671.
17. Passamonti F, Randi ML, Rumi E, et al. Increased risk of pregnancy
complications in patients with essential thrombocythemia carrying the JAK2
(617VF) mutation. Blood. 2007;110:485-489.
18. Gangat N, Caramazza D, Vaidya R, et al. DIPSS plus: A refined
Dynamic International Prognostic Scoring System for primary myelofibrosis
that incorporates prognostic information from karyotype, platelet count, and
transfusion status. J Clin Oncol. 2011;29:392-397.
19. Tam CS, Abruzzo LV, Lin KI, et al. The role of cytogenetic abnormalities as a prognostic marker in primary myelofibrosis: Applicability at the
time of diagnosis and later during disease course. Blood. 2009;113:4171-4178.
20. Caramazza D, Begna KH, Gangat N, et al. Refined cytogenetic-risk
categorization for overall and leukemia-free survival in primary myelofibrosis: A single center study of 433 patients. Leukemia. 2011;25:82-88.
21. Guglielmelli P, Barosi G, Specchia G, et al. Identification of patients
with poorer survival in primary myelofibrosis based on the burden of
JAK2V617F mutated allele. Blood. 2009;114:1477-1483.
22. Tefferi A, Lasho TL, Huang J, et al. Low JAK2V617F allele burden in
primary myelofibrosis, compared to either a higher allele burden or unmutated status, is associated with inferior overall and leukemia-free survival.
Leukemia. 2008;22:756-761.
23. Alchalby H, Badbaran A, Zabelina T, et al. Impact of JAK2V617Fmutation status, allele burden and clearance after allogeneic stem cell
transplantation for myelofibrosis. Blood. 2010;116:3572-3581.
24. Tefferi A, Pardanani A, Lim KH, et al. TET2 mutations and their
clinical correlates in polycythemia vera, essential thrombocythemia and
myelofibrosis. Leukemia. 2009; 23:905-911.
25. Pardanani A, Guglielmelli P, Lasho TL, et al. Primary myelofibrosis
with or without mutant MPL: Comparison of survival and clinical features
involving 603 patients. Leukemia. 2011;25:1834-1839.
26. Tefferi A, Jimma T, Sulai NH, et al. IDH mutations in primary
myelofibrosis predict leukemic transformation and shortened survival: Clinical evidence for leukemogenic collaboration with JAK2V617F. Leukemia.
Epub 2011 Sept 13.
27. Guglielmelli P, Biamonte F, Score J, et al. EZH2 mutational status
predicts poor survival in myelofibrosis. Blood. 2011;118:5227-5234.
28. Tefferi A, Lasho TL, Patnaik MM, et al. JAK2 germline genetic
variation affects disease susceptibility in primary myelofibrosis regardless of
V617F mutational status: Nullizygosity for the JAK2 46/1 haplotype is
associated with inferior survival. Leukemia. 2010;24:105-109.
29. Dupriez B, Morel P, Demory JL, et al. Prognostic factors in agnogenic
myeloid metaplasia: A report on 195 cases with a new scoring system. Blood.
1996;88:1013-1018.
30. Tefferi A, Siragusa S, Hussein K, et al. Transfusion-dependency at
presentation and its acquisition in the first year of diagnosis are both equally
detrimental for survival in primary myelofibrosis-prognostic relevance is
independent of IPSS or karyotype. Am J Hematol. 2010;85:14-17.
AROUND THE WORLD IN ALMOST 80 MINUTES:

LUNG CANCER CARE AND RESEARCH
CHAIR
Gilberto Schwartsmann, MD, PhD
Federal University of Rio Grande do Sul
Porto Alegre, Brazil
SPEAKERS
Tony Mok, MD
Prince of Wales Hospital
Shatin, Hong Kong
Tudor E. Ciuleanu, MD
Institute Ion Chiricuta
Cluj-Napoca, Romania
Lung Cancer in Brazil

By Gilberto Schwartsmann, MD, PhD
Overview: Cancer is now the second leading cause of death

in Brazil (after cardiovascular diseases) and a public health
problem, with around 500,000 new cases in 2012. Excluding
nonmelanoma skin cancer, lung cancer is the second most
incident cancer type in men, with 17,210 expected new cases.
In women, it is the fifth most incident cancer, with 10,110
expected new cases. The estimated age-adjusted lung cancer
mortality rate is about 13/100,000 for men and 5.4/100,000 for
women. Lung cancer rates in men increased until the early
1990s and decreased thereafter, especially in the younger
population. In contrast, a steady upward trend was observed
for women. The positive effects in men were probably due to
the successful anti-tobacco campaign conducted in Brazil
over the last decades, which led to a decrease in the adult
ORLDWIDE, LUNG cancer is the number one cause

of cancer death in men and the second in women,
with more than 1.6 million new cases and 1.4 million deaths
every year. Notably, the majority of lung cancer cases now
occur in developing countries (55%), a substantial increase
since the 1980s, when only one-third of cases occurred in
these regions.1
Considering that cigarette smoking is the main causative
factor for about four of five lung cancer deaths in men and for
one-half of deaths in women, the variation in incidence
largely reflects the patterns of tobacco consumption. Environmental exposure to other causative factorssuch as
asbestos, radon, arsenic, radiation, air pollution, coal smoke,
and indoor emission from unventilated coal-fueled stoves
and cookingis also relevant in specific regions.2
Epidemiology
Brazil is the largest country in South America, with an

estimated population of more than 190 million. Cancer is
now the second most common cause of death in most
geographic regions in the country (Fig. 1). In 2012, more
than 500,000 new cancer cases are estimated, equally distributed between genders. Lung cancer will be responsible
for around 17,210 new cancer cases in men and 10,110 new
cases in women (Fig. 2). Male lung cancer increased until the
early 1990s and decreased in the 2000s, especially in the
younger population. In women, there was a steady upward
trend for cancers of the lung during the same period.3
Recent data from the Brazilian National Cancer Institute
estimate that lung cancer is the second and fifth most
common type of cancer among men and women, respectively
(not including nonmelanoma skin cancers) (Fig. 2). Trends
in lung cancer mortality showed that the age-adjusted
mortality rate among men increased from 10.6 to 13.1/
100,000 between the years 1979 and 2004, with an increase
from 3.0 to 5.4/100,000 among women during the same
period. As a rule, lung cancer mortality rates in Brazil are
significantly higher among men. Specific rates for men over
the age of 64 and for women of all ages are increasing.
Notably, there is a lesser risk of mortality among men born
after 1950 and an increasing risk across all cohorts among
women. The results regarding younger generations indicate
that present trends are likely to continue. Interestingly, the
cohort effect in women points to an increasing trend in
426
smoking population, from 32% in the early 1980s to 17% in the

2000s. Although the Brazilian National Cancer Institute is
strongly committed to providing excellence in multimodality
care to cancer patients, limitations in availability and adequate geographic distribution of specialists and wellequipped cancer centers are evident. Major disparities in
patient access to proper staging and state-of-the-art treatment still exist. Considering that World Health Organization
(WHO) officials estimate that cancer will become the number
one cause of death in most developing countries, including
Brazil, in the next decades, it is highly recommended for
government authorities to implement firm actions to face this
tremendous challenge.
mortality rates, whereas the reduction in rates in men

younger than age 65 suggests that the above-mentioned
trend should continue, probably as a consequence of the
tobacco control measures adopted after the 1980s.4
Lung Cancer in Never Smokers
As previously reported, patients who have never smoked

tend to have a better survival rate than patients who did
smoke, regardless of gender and histologic type. This was
also illustrated by a report by Brazilian investigators from a
large academic hospital in Sao Paulo, Brazil, where 56 of
285 (19%) patients with non-small cell lung cancer (NSCLC)
were never smokers. They were more likely to be female
(68% vs. 32%) and have adenocarcinoma (70% vs. 51%).5
Prevention and Early Diagnosis
Brazil was one of the first countries to support the WHO

Framework Convention on Tobacco Control and, over the
last few decades, implemented several tobacco control measures, including banning smoking in public places, restricting tobacco advertising and promotion, counter-advertising,
raising the price of cigarettes, and providing counseling for
tobacco dependence. As a result, the rate of active smokers
dropped from approximately 32% in the late 1980s to 17% in
the 2000s. Brazilian authorities are projecting a further
0.3% annual reduction in smoking, with an expected smoking population of approximately 11% by 2020.6
Worldwide, late diagnosis is a critical component of the
dismal survival rates of patients with lung cancer. As
localized disease can be treated with curative intent, new
hope is derived from the identification of effective screening
efforts. Recent trials have demonstrated that lung cancer
From the Department of Medical Oncology, Hospital de Clnicas de Porto Alegre, Federal
University of Rio Grande do Sul, Porto Alegre, Brazil; and South-American Office for
Anticancer Drug Development, Porto Alegre, Brazil.
Address reprint requests to Gilberto Schwartsmann, MD, PhD, Department of Medical
Oncology, Hospital de Clinicas de Porto Alegre, Federal University of Rio Grande do Sul,
Rua Ramiro Barcelos 2350/s399, CP 90 035-903, Porto Alegre, RS, Brazil; email:
gilberto.ez@terra.com.br.
1092-9118/10/1-10
LUNG CANCER IN BRAZIL
Fig. 1. Relative distribution of causes of death in Brazil, 2010.31

Abbreviation: CNS, central nervous system.
mortality can be reduced by annual screening with low-dose

CT, especially in high-risk populations. However, this strategy can be associated with possible harm.7 Currently, there
is still no formal recommendation for the routine use of
screening CT as standard clinical practice in Brazil.
KEY POINTS
Cancer is now a public health problem in Brazil, with

more than 500,000 new cases expected for the year
2012.
Lung cancer will be responsible for 17,210 new cases
in men and 10,110 new cases in women.
Brazil reduced the percentage of smokers from
around 32% in the 1970s to 17% in 2010s, and a
further annual decrease of about 0.3% is projected by
government health authorities.
Diagnosis, staging, and treatment guidelines for patients with lung cancer follow the recommendations
of the Brazilian Society of Clinical Oncology and the
Brazilian Cancer Society, which are consistent with
the National Comprehensive Cancer Network and
European Society for Medical Oncology guidelines.
As a developing country with large cultural and
socioeconomic contrasts and an estimated population
of over 190 million inhabitants for the year 2012,
disparities in patient access to proper medical care
are still a major challenge for cancer control in Brazil.
Classification and Prognostic Factors
Lung cancer is grouped as NSCLC and small-cell lung

cancer (SCLC). NSCLC accounts for more than 85% of all
cases and is classified as squamous (SCC) and nonsquamous
cell carcinoma (NSCC). NSCC includes adenocarcinomas,
large-cell carcinomas, and other subtypes. In a retrospective
analysis of 240 consecutive patients treated at an academic
hospital in Sao Paulo, Brazil, between 2000 and 2006, the
most common NSCLC subtype was SCC.8 Similar results
were reported in a study performed in Manaus, Brazil,
between 1995 and 2002.9 In contrast, in a recent series of
patients with NSCLC from the Brazilian National Cancer
Institute in Rio de Janeiro, Brazil, the most common histologic subtype was adenocarcinoma.10 This pattern is being
also observed in an NSCLC series from developed countries.11 As a rule, most patients with NSCLC (80%) present
with locally advanced or advanced disease (stages III and
IV), while a minority (20%) of patients have early disease
(stages I and II) at presentation.8,10
Favorable prognostic factors for survival in patients with
NSCLC include early stage, good performance status, no
weight loss, and female gender. In contrast, p53 mutations
and KRAS activation may predict a less favorable prognosis.12 In another study, 114 patients with NSCLC from Porto
Alegre, Brazil, underwent tumor resection for stage I disease. The 5-year survival rate was 85.5% and 46.4% for
women and men, respectively (p 0.0001).13
Pathology
Tumor specimens can be obtained from fine-needle aspiration, core needle, endobronchial or transbronchial biopsy,
427
GILBERTO SCHWARTSMANN
Fig. 2. Total number and percent distribution of the ten most frequent types of cancers projected for the year 2012 in Brazil (excluding
nonmelanoma skin cancer).3
Abbreviation: CNS, central nervous system.
and bronchial brushings or washings. Mediastinal lymph

nodes are often sampled for staging. The pathologist provides information about the histologic type (or subtype), the
presence and extent of tumor invasion, and surgical margins. The presence of EGFR or EML4-ALK mutations is also
important, as it can direct treatment toward the use of
specific tyrosine kinase inhibitors (TKIs).12
In a study performed by investigators at the Brazilian
National Cancer Institute, 162 paraffin-embedded specimens obtained from patients with NSCLC were analyzed for
EGFR (exons 18 to 21), KRAS (exon 2), and BRAF (exons 11
and 15) mutations and for MET amplification and ALK
rearrangement (Fig. 3).10 EGFR analysis was successful
in 150 cases, and mutations were identified in 25.3% (38
patients) of patients (6, 19, 13, and 5 mutations in exons 18,
19, 20, and 21, respectively; 5 cases encompassing mutations
in 2 different exons). These abnormalities were observed
mainly in adenocarcinomas (52.8%) but were also frequent
in SCC (36.1%).
In the above study, KRAS mutations were observed in 30
of 148 analyzed cases (20.3%), in which 76.7% were adenocarcinomas. BRAF mutations were found in 13 of 145
patients (9.0%), mainly in squamous histology (61.5%). Notably, mutations in exon 11 in men with SCC were identified, although no case of V600E mutation was identified.10
The latter observation differs from previous reports of a high
prevalence of this abnormality in patients with a BRAF
mutation.14 MET geneincreased copy number (mean 5
copies/cell) was observed in 21 of 152 evaluated cases
(13.8%), mostly in adenocarcinomas (73.7%). ALK rearrangement was present in four of 161 cases (2.5%), three of
them with adenocarcinoma and one with bronchioloalveolar
histology.
During the surgical procedure, the removed tissue should
428
provide information about the status of resection margins,

involvement of regional lymph nodes and the presence of
incidental nodules. The information on the WHO histologic
type (and subtype), disease staging, and prognostic factors
are essential for treatment planning. The recognition of a
bronchioloalveolar subtype of adenocarcinoma is also relevant, as EGFR-TKIs are beneficial for these patients.15
Furthermore, immunohistochemistry is a critical tool in
distinguishing between primary lung cancer and other neoplasms, such as mesothelioma, metastatic cancer of another
origin, or neuroendocrine tumors.12
Clinical Evaluation
The initial evaluation of a patient with NSCLC includes

medical history, physical examination, laboratory tests,
chest x-ray, abdominal ultrasound, and/or CT scans. Bronchoscopy is recommended for the diagnosis of central lesions, while percutaneous biopsy is usually performed in
peripheral lesions. It is recommended that the patient be
referred to a smoking cessation program. Staging includes a
chest CT with the inclusion of the upper abdomen and
adrenals.
Mediastinoscopy is the gold standard for the study of
mediastinal lymph nodes. The sampling of mediastinal
lymph nodes is essential, as CT scans have limitations to
rule out lymph node involvement. Endobronchial ultrasoundtransbronchial needle aspiration (EBUS-TBNA) is also being considered for the mediastinal staging of patients with
NSCLC worldwide. Unfortunately, only a small number of
cancer centers in Brazil are currently applying this technique in the routine treatment of patients with lung cancer.
The use of PET imaging is also a valuable tool in the
current staging of patients with NSCLC. However, because
of its high cost and limited availability through the National
Fig. 3. Characteristics of groups mutated

for EGFR, KRAS, BRAF, and TP53.10
Abbreviations: EGFR, epidermal growth
factor receptor; SCC, squamous cell carcinoma.
Health System in Brazil, its use is usually restricted to

patients with early disease who are candidates for curative
surgery. In the absence of specific symptoms, there is no
indication for the use of brain MRI or bone scans in patients
with stage IV disease. An MRI of the brain, however, is
considered by many oncologists in Brazil in the case of lung
adenocarcinoma, because of the higher risk of brain metastases. Finally, surgical resection without prior invasive
testing may be an option in patients with a highly suspicious
solitary pulmonary nodule.
In a series of 291 patients with NSCLC treated in an
academic hospital in Sao Paulo, Brazil, clinical staging was
based on clinical and imaging studies. PET scan was not
routinely performed. Pathologic staging differed from clinical staging in 33% of cases (15% were upstaged and 18%
downstaged). Sensitivity, specificity, positive and negative
predictive values, and accuracy for clinical staging were
78%, 69%, 82%, 64%, and 67%, respectively.16
In another study performed in Porto Alegre, Brazil, the
authors looked at the indications for MRI studies in patients
with NSCLC. MRI was used mainly for patients with superior sulcus tumors, suspected spinal cord canal invasion, and
suspected brain metastases.17
Biomarkers
EGFR exon 19 deletion or exon 21 L858R mutation result

in activation of the tyrosine kinase domain and are predictive of better sensitivity to EGFR-TKIs. These mutations are
found in approximately 10% to 15% of white and 30% to 40%
of Asian patients with NSCLC.10,18 As described below,
these results were not confirmed in a recent study performed
by investigators of the Brazilian National Cancer Institute,
in which 25.3% (38 of 150 cases) of patients with NSCLC
showed these mutations. EGFR mutations were more common in women (60.6%), smokers (51.4%), and in the adeno-
carcinoma subtype (54.5%). These mutations were also

detected in never (15.2%) and previous (27.3%) smokers, as
well as in a significant percentage of patients with the
squamous cell subtype (33.4%).10
Patients with NSCLC with high ERCC1 expression in the
tumor were shown to have a better overall survival than
those with low ERCC1 expression. High ERCC1 levels were
also associated with resistance to platinum-based chemotherapies.12,19 KRAS mutation is observed in approximately
25% of adenocarcinomas. It is associated with worse survival
and resistance to platinum-based chemotherapy or EGFRTKIs.12,20 Furthermore, high RRM1 levels are associated
with better survival but poor response to gemcitabine or
carboplatin.12,21
Treatment
Surgical resection should be attempted in all patients with

stage I or II good performance status NSCLC, as it provides
the best chances of cure.12 In a study performed in a cancer
hospital in Sao Paulo, Brazil, including 737 patients with
NSCLC, complete tumor resection was performed in 24.6%
of patients. The overall 5-year survival rate was 28%, and
the median survival was 18.9 months.22
The survival advantage of adjuvant cisplatin-based chemotherapy for patients with completely resected stage I, II,
or III NSCLC was demonstrated in prospective randomized
trials (IALT, NCIC, and ANITA). Postoperative cisplatinbased chemotherapy produced an absolute benefit of 5% in
5-year survival, especially in patients with stage II-III
disease with good performance status. Adjuvant chemotherapy was also beneficial for patients with stage I disease
with 4-cm tumors.12,23
Several therapeutic options can be considered for patients
with stage IIIA disease. For unresectable cases of stage IIIA
or IIIB disease, chemoradiation was shown to be superior to
429
GILBERTO SCHWARTSMANN
radiation alone. Concurrent chemoradiation showed better

results as compared with sequential therapy. Most centers
in Brazil adopt a cisplatin/etoposide concurrent chemoradiation regimen in this context, although other regimens are
used, such as carboplatin/paclitaxel or cisplatin/vinblastine.
The question of adding additional courses of chemotherapy
after chemoradiation is unsolved.
In a trial performed at the Brazilian National Cancer
Institute, 30 patients with clinical stages IB to IIIA NSCLC
who were candidates for surgical resection received three
cycles of neoadjuvant chemotherapy. Patients without evidence of progression underwent mediastinoscopy. Those
with negative lymph nodes underwent resection, whereas
radiation was given to those with positive nodes. Twentythree patients (77%) responded to neoadjuvant chemotherapy, and complete resection rate was achieved in 21 patients
(70%).24
For patients with stage IV disease and a good performance
status, chemotherapy is the treatment of choice. With the
exclusion of patients with central nervous system metastasis, chemotherapy greatly increases survival compared with
best supportive care.12 This was corroborated by a study of
patients with stage IV NSCLC in a cancer hospital in Sao
Paulo, Brazil.25 In another trial in three cancer centers in
Brazil, 564 patients with stage IV disease were evaluated.
Of those, 335 (59.4%) received chemotherapy. There was a
great heterogeneity in drug regimens used in the patients.
Again, overall survival was better with chemotherapy compared with best supportive care.26
Several agents showed objective responses in advanced
stage NSCLC, including paclitaxel, docetaxel, vinorelbine,
etoposide, pemetrexed, irinotecan, and gemcitabine.
Platinum-containing doublets can lead to 30% to 40% oneyear survival rates and are superior to single agents.12
Carboplatin/paclitaxel, cisplatin/vinorelbine, cisplatin/gemcitabine, carboplatin/docetaxel, or carboplatin/pemetrexate
are drug regimens commonly used in the treatment of
patients with advanced NSCLC in Brazil.
It should be emphasized that in patients with good performance status and solitary brain metastasis, surgical
resection may improve survival. The results for solitary
metastatic lesions in other sites are controversial.12
New targeted therapies were shown to produce tumor
responses in patients with advanced lung cancer. Bevacizumab may be used in combination with paclitaxel and
carboplatin in patients with advanced stage NSCLC with
nonsquamous histology and no hemoptysis. In a metaanalysis published by investigators from Campinas, Brazil,
bevacizumab improved response rate and progression-free
survival in patients with NSCLC receiving chemotherapy,
while the effect on overall survival was uncertain.27 Erlotinib is approved as first-line therapy in patients with EGFR
mutationpositive stage IV NSCLC. It can also be used in
patients who progressed after one prior chemotherapy regimen. Furthermore, cetuximab showed an increase in overall
survival in patients with stage IV disease when given in
430
combination with cisplatin/vinorelbine.12 The continuation

of a biologic agent, such as bevacizumab or cetuximab, can
be an option after four to six chemotherapy courses for
responding patients or stable disease. At present, the abovementioned biological agents are not provided routinely as
part of the standard treatment of patients with NSCLC
through the Brazilian National Health System.
Pemetrexed is an alternative maintenance therapy. Two
recent studies have demonstrated progression-free and overall survival benefit with the administration of switch maintenance with pemetrexed (for nonsquamous histology only),
erlotinib, or docetaxel, following four to six courses of
platinum-containing chemotherapy. It should be emphasized, however, that there is no data to support the continuation of combination chemotherapy beyond four to six
treatment courses.28
Systemic Therapy for SCLC
SCLC represents approximately 10% to 15% of all lung

cancers. As SCLC has a high risk of early metastatic
dissemination, only a minority of cases (30%) have limited
disease at presentation. Tumor responses to both chemotherapy and radiotherapy are frequent (50% to 80%), but
relapses are the rule. Platinum plus etoposide chemotherapy is the cornerstone regimen in the treatment of these
patients. In most centers in Brazil, platinum/etoposide chemotherapy combined with thoracic radiotherapy is the
choice for patients with limited disease, while chemotherapy
alone is used in patients with extensive disease. Because of
the high risk of central nervous system involvement, prophylactic cranial irradiation is indicated for complete responders. Five-year survival of patients presenting with
limited disease varies between 10% and 25%, although it
does not exceed 10% at 2 years in patients with extensive
disease. Most patients relapse within the first two years,
and there are few second-line treatment options for these
patients.
Irinotecan was shown to produce objective responses in
patients with SCLC. In a phase II trial conducted in Japan,
irinotecan was associated with a median survival of 13
months in patients with extensive-stage disease. Subsequently, a phase III trial comparing cisplatin/irinotecan
with cisplatin/etoposide demonstrated superior median
1-year and 2-year survival rates for the irinotecan arm.29
This was corroborated by a meta-analysis conducted by
investigators from Sao Paulo, Brazil, that included eight
trials and 3,086 patients with SCLC.30 Presently, cisplatin/
etoposide, cisplatin/irinotecan, and carboplatin/irinotecan
are the most popular drug regimens used for the treatment
of SCLC in Brazil. Most academic hospital still favor the use
of cisplatin/etoposide with concomitant irradiation as firstline therapy for patients with limited disease and this
regimen alone for the initial treatment of patients with
extensive disease. The cyclophosphamide/doxorubicin/vincristine regimen is now rarely used as first-line therapy for
this disease in Brazil.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Gilberto Schwartsmann*
REFERENCES
1. Siegel R, Naishadham D, Jemal A. Cancer statistics, CA Cancer J Clin.
2012;62:10-29.
2. Ferlay J, Shin HR, Bray F, et al. Estimates of worldwide burden of
cancer in 2008: GLOBOCAN 2008. Int J Cancer. 2010;127:2893-2917.
3. Instituto Nacional de Cancer and Ministerio da Saude. Estimativa 2012:
Incidencia de cancer no Brasil. Rio de Janeiro: INCA. http://www.inca.gov.br.
Accessed January 2012.
4. Souza MC, Vasconcelos AGG, Cruz OG. Trends in lung cancer mortality
in Brazil from the 1980s into the early 21st century: age-period-cohort
analysis. Cad Saude Publica. 2012;28:21-30.
5. Santoro IL, Ramos RP, Franceschini J, et al. Non-small cell lung cancer
in never smokers: A clinical entity to be identified. Clinics (Sao Paulo).
2011;66:1873-1877.
6. Monteiro CA, Cavalcante TM, Moura EC, et al. Population-based evidence of a strong decline in the prevalence of smokers in Brazil (1989-2003).
Bull World Health Organ. 2007;85:527-534.
7. McLoud TC. Initial results of the National Lung Cancer Screening Trial.
Cancer Imaging. 2011;11:S85.
8. Novaes FT, Cataneo DC, Ruiz Junior RL, et al. Lung cancer: Histology,
staging, treatment and survival. J Bras Pneumol. 2008;34:595-600.
9. Westphal FL, Lima LC, Andrade EO, et al. Characteristics of patients
with lung cancer in the city of Manaus. Brazil J Bras Pneumol. 2009;35:157163.
10. Melo, ACD, Inada HKP, Barros M, et al. Non-small cell lung cancer
(NSCLC) genotyping in a Brazilian cohort. Poster Session P2.123 presented
at 14th World Conference on Lung Cancer; 2011; Amsterdam, NL.
11. Janssen-Heijnen ML, Coebergh JW. Trends in incidence and prognosis
of the histological subtypes of lung cancer in North America, Australia, New
Zealand and Europe. Lung Cancer. 2001;31:123-137.
12. Ettinger DS, Akerley W, Bepler G, et al. Non-small cell lung cancer.
13. Chatkin JM, Abreu CM, Fritscher CC, et al. Is there a gender difference
in non-small cell lung cancer survival? Gend Med. 2004;1:41-47.
14. Paik PK, Arcila ME, Fara M, et al. Clinical characteristics of patients
with lung adenocarcinomas harboring BRAF mutations. J Clin Oncol. 2011;
29:2046-2051.
15. Wislez M, Lavole A, Gounant V, et al. Bronchiolar-alveolar carcinoma:
From concept to innovative therapeutic strategies. Presse Med. 2011;40:389397.
16. Younes RN, Schutz FA, Gross JL. Preoperative and pathological
staging of NSCLC: Retrospective analysis of 291 cases. Rev Assoc Med Bras.
2010;56:237-241.
17. Hochhegger B, Marchiori E, Sedlaczek O, et al. MRI in lung cancer: A
pictorial essay. Br J Radiol. 2011;84:661-668.
18. Ren JH, He WS, Yan GL, et al. EGFR mutations in non-small-cell lung
cancer among smokers and non-smokers: A meta-analysis. Environ Mol
Mutagen. 2012;53:78-82.
19. Vilmar A, Srensen JB. Excision repair cross-complementation group 1

(ERCC1) in platinum-based treatment of non-small cell lung cancer with
special emphasis on carboplatin: A review of current literature. Lung Cancer.
2009;64:131-139.
20. Gaughan EM, Costa DB. Genotype-driven therapies for non-small cell
lung cancer: Focus on EGFR, KRAS and ALK gene abnormalities. Ther Adv
Med Oncol. 2011;3:113-125.
21. Ryu JS, Shin ES, Nam HS, et al. Differential effect of polymorphisms of
CMPK1 and RRM1 on survival in advanced non-small cell lung cancer
patients treated with gemcitabine or taxane/cisplatinum. J Thorac Oncol.
2011;6:1320-1329.
22. Younes RN, Deutsch F, Badra C, et al. Nonsmall cell lung cancer:
Evaluation of 737 consecutive patients in a single institution. Rev Hosp Clin
Fac Med Sao Paulo. 2004;59:119-127.
23. NSCLC Meta-analyses Collaborative Group, Arriagada R, Auperin A,
et al. Adjuvant chemotherapy, with or without postoperative radiotherapy, in
operable non-small-cell lung cancer: Two meta-analyses of individual patient
data. Lancet. 2010;375:1267-1277.
24. Martins RG, Dienstmann R, de Biasi P, et al. Phase II trial of
neoadjuvant chemotherapy using alternating doublets in non-small-cell lung
cancer. Clin Lung Cancer. 2007;8:257-263.
25. Anelli A, Lima CA, Younes RN, et al. Chemotherapy versus best
supportive care in stage IV non-small cell lung cancer, non metastatic to the
brain. Rev Hosp Clin Fac Med Sao Paulo. 2001;56:53-58.
26. Naime FF, Younes RN, Kersten BG, et al. Metastatic non-small cell
lung cancer in Brazil: Treatment heterogeneity in routine clinical practice.
Clinics (Sao Paulo). 2007;62:397-404.
27. Botrel TE, Clark O, Clark L, et al. Efficacy of bevacizumab (Bev) plus
chemotherapy (CT) compared to CT alone in previously untreated locally
advanced or metastatic non-small cell lung cancer (NSCLC): systematic
review and meta-analysis Lung Cancer. 2011;74:89-97.
28. Azzoli CG, Temin S, Aliff T, et al. 2011 Focused Update of 2009
American Society of Clinical Oncology Clinical Practice Guideline Update on
Chemotherapy for Stage IV Non-Small-Cell Lung Cancer. J Clin Oncol.
2011;29:3825-3831.
29. Planchard D, Le Pechoux C. Small cell lung cancer: New clinical
recommendations and current status of biomarker assessment. Eur J Cancer.
2011;47 Suppl 3:S272-S283.
30. Lima JP, dos Santos LV, Sasse EC, et al. Camptothecins compared with
etoposide in combination with platinum analog in extensive stage small cell
lung cancer: Systematic review with meta-analysis. J Thorac Oncol. 2010;5:
1986-1993.
31. Brazilian Institute of Geography and Statistics - Instituto Brasileiro de
Geografia e Estatstica (IBGE), Brazil, 2011. www.ibge.gov.br. Accessed
January 2012.
431
Research and Standard Care: Lung Cancer

in China
By Tony S. Mok, MD, Qing Zhou, MD, and Yi-Long Wu, MD
Overview: China has an enormous burden from rising tobacco consumption and lung cancer incidence. Governmental
intervention on lung cancer prevention is insufficient, and both
incidence and mortality related to lung cancer are still on the
rise. Treatment guidelines are available, but heterogeneity in
the quality of care between centers, especially the disparity
between urban and rural areas, have resulted in inconsistent
care to patients with lung cancer. Despite knowledge on
molecular-targeted therapy, only a small fraction of patients
HINA, A COUNTRY in rapid development, is confronted with a major health hazard. Success of her
development comes with the price of pollution, change in
dietary habits, and increase in tobacco consumption. Cancer
has become the leading cause of death in China. According to
the Ministry of Health in China, mortality related to cancer
has risen by 80% during the last 30 years to an alarming
number of 1.8 million cancer deaths in 2010. Cancer incidence rose from 74.2/100,000 in 1970 to 135.9/100,000 in
2004.1 Among the death tolls, a substantial proportion is
attributed to lung cancer. Similar to other developed countries, majority of patients with lung cancer are diagnosed at
advanced stage, and only a small proportion of patients have
curable disease at presentation. During the last three decades, deaths related to lung cancer have increased by 465%.
But different from other developed countries, the quality of
health care in this most populated country is highly heterogeneous. The tremendous success in economic development
brings wealth to a small sector of society, whereas a large
proportion of residents are still living close to poverty line.
With the lack of universal health care system, many suffer
from inadequate treatment as well as from disease itself. In
this review, we summarize the disease burden of lung cancer
in China, the current treatment standard across the country, and the impressive development in clinical/translational
research on this dreadful illness.
The Burden
The burden starts with tobacco consumption. Being the

world largest tobacco-producing country, China also consumes majority of her own production. The World Health
Organization (WHO) estimated China to have 320 million
smokers in 2008, and this may indirectly cause environmental tobacco exposure to an estimate of more than 500 million
nonsmokers. With the lung cancer risk being six to 10 times
higher in smokers and 30% to 60% higher in nonsmokers
with exposure to second-hand smoke, the incidence of lung
cancer in China is expected to increase. To date, there is
still an absence in reversing the trend in the number of
smokers in China. This country generates over US$21 billion in tobacco tax annually, and some provinces such as
Yunnan are basically dependent on this tax revenue.2 It is
unlikely that there will be dramatic control of tobacco
production or consumption.
China has signed the WHO Framework Convention on
Tobacco Control in 2003. The scale of the health care burden
is being recognized, but the total budget spent on tobacco
432
have access to routine EGFR mutation analysis. Platinumbased doublet chemotherapy remains the most commonly
used regimen irrespective of mutation status. On a positive
note, both clinical and translational research on lung cancer
are in rapid progress. The Chinese Thoracic Oncology Group
(CTONG) has already contributed substantially to the care of
patients with lung cancer and is expected to continue in the
trend.
control was estimated to be less than US$1 million, which is

a very small fraction compared with the revenue generated
by tobacco. In 2002, it was estimated in a national survey in
30 provinces in China that smoking prevalence rate was
35.8% (66% in men and 3.1% in women).3 The greatest
concern is the increasing number of young smokers. It was
estimated that about 15 million young people between the
ages of 13 and 18 years are regular smokers, and another 40
million in this age group are occasional smokers. The median age of starting smoking is 19 years.4 The smoking rate
in men (60%) is much higher than that in women ( 5%), but
the incidence of women smokers is rapidly rising.
Given the grim situation in tobacco consumption, it is not
surprising that lung cancer incidence has increased. More
than 400,000 cancer deaths were attributed to tobacco
consumption in 2005, among which lung cancer accounted
for more than 240,000.5 In 2004, the mortality rate of lung
cancer was 41/100,000 in urban areas and 26/100,000 in
rural areas. WHO estimated that total number of new cases
of lung cancer may reach 1 million every year by 2025; thus
the death toll will increase proportionally.6
The only logical way to reduce lung cancer incidence and
related death is tobacco control. The China National Office
for Cancer Prevention is responsible for reduction of cancer
incidence and mortality, and their primary targets are lung,
liver, and gastric cancers. Tobacco control is on the top of
their agenda, but investment in the program was relatively
small comparing to the magnitude of the problem. Joining
other countries, the WHO Framework Convention on Tobacco Control was signed in 2003 and ratified in 2005. The
government had invested 10 million RMB on the project.
The objective is to promote a smoke-free environment, build
an antismoking network, and provide assistance in smoking
cessation. But for the 320 million current smokers in China,
this amount is highly insufficient. To date there is legislation against smoking in public places but lack of effort in
execution of the law. Cigarette cost and taxes are still
relatively low. Although there is a law against sales of
From the Department of Clinical Oncology, State Key Laboratory of South China, Hong
Kong Cancer Institute, The Chinese University of Hong Kong, Hong Kong; Guangdong Lung
Cancer Institute, Guangdong General Hospital and Guangdong Academy of Medical
Sciences, Guangzhou, China.
Address reprint requests to Tony S. Mok, MD; email: tony@clo.cuhk.edu.hk.
1092-9118/10/1-10
LUNG CANCER IN CHINA
Fig 1. Chinese Guideline for Diagnosis

and Treatment of Lung Cancer.
Abbreviation: TNM, tumor mode metastasis.
tobacco to minors, there is again no specific action on

execution of this law. Unless the government has substantial change in attitude toward tobacco control, the burden
related to lung cancer is unlikely to cease.
The Treatment
Treatment for lung cancer is quite diverse in China.

Economic and technologic advantages in urban areas provide much better access to high-quality care, whereas treat-
KEY POINTS
China has a large burden of lung cancer. Being the

largest tobacco-producing country and a major consumer, its incidence of lung cancer is expected to
increase unless major action on tobacco control is
taken.
Treatment guidelines are available, but clinical practice may not always follow the guidelines.
With the exception of bevacizumab, most of the active
agents for lung cancer are available in China. Drug
cost is either partially reimbursed or self-financed.
Despite the high incidence of lung cancer, routine
EGFR mutation analysis is available to only a small
fraction of patients and is limited to major academic
centers.
The Chinese Thoracic Oncology Group (CTONG) is an
extremely active and well-organized study group.
Within a few short years since establishment, they
have already published major studies such as OPTIMAL and INFORM.
ment in rural area could be quite primitive. Unfortunately,

limited information is available from rural area, and for the
selected patients from rural areas who can afford the medical expenses, they would travel to urban areas for cancer
care. Thus, in this review, we will focus on only the current
status of lung cancer treatment in major cities.
Treatment guidelines are available, although there is no
legal or financial obligation for doctors to follow the guidelines. These are being used as references because the reimbursement system does not always follow the guidelines.
The most commonly used guidelines include the Chinese
Guideline for Diagnosis and Treatment of Lung Cancer7 and
the Chinese translation of the National Comprehensive
Cancer Network (NCCN) Guideline. The former was developed by local experts, whereas the latter is a direct translation of a U.S. national guideline with minor variations.
Contents of the two guidelines are very similar. The basic
framework of the Chinese guideline is summarized in Figure
1. Use of a multimodality approach is well recognized. The
major difference is limited to the availability of specific
medication in China. For example, bevacizumab has not
been approved for use in lung cancer in China. The Chinese
Guideline stated that antiangiogenesis drug may be added
if applicable, whereas the Chinese translation of NCCN
guidelines include endostar (a Chinese made angiotensin)
and ginseng extract as an option for first-line treatment in
combination with chemotherapy. Pemetrexed was not approved for first-line use until recently. Apart from the minor
variation, the use of surgery for early-stage disease, chemoradiotherapy for local advanced disease, epidermal growth
factor receptor tyrosine kinase inhibitor (EGFR TKI) for
patients with EGFR mutation, and systemic chemotherapy
for stage IV disease are well accepted. However, the guide-
433
MOK, ZHOU, AND WU
Fig 2. Growth in Number of Clinical Trials in

China.
SEER, Surveillance, Epidemiology and End
Results.
lines are not necessarily routinely translated into clinical

practice.
Shanghai is among the few cities that have detailed
epidemiologic data on their patients with lung cancer. The
Shanghai Municipal Center for Disease Control and Prevention reported a total of 25,927 male lung cancer deaths and
10,468 female lung cancer deaths between 2003 and 2007.
The pathologic distribution for adenocarcinoma, squamous
cell carcinoma and small cell lung cancer were 58.9%, 35.4%,
and 5.7%, respectively. Interestingly, the death rate has a
slight trend of decline since 2001, and the 5-year survival
rate has increased in the last decade. In Beijing, a total of
32,845 cases of lung cancer were diagnosed between 1998
and 2007.8 The crude incidence in women has increased.
However, only approximately half of the patients (54.9%)
had histologic confirmation of diagnosis. Histologic subtypes
of squamous cell carcinoma and adenocarcinoma was 30.4%
and 42.8%, respectively, in 1998, and changed to 24.1% and
46.8% in 2007, respectively.
The quality of thoracic surgery is generally good in urban
major hospitals. Yang and colleagues9 performed 621 anatomic lobectomies between 1996 and 2003, of which 113 were
by video-assisted thorascopic surgery (VATS). Perioperative
morbidity and mortality rates was 0% and 0.9%, respectively. The postoperative complication rate was 10.6%. The
survival rates at 5 years for stage I, II, and III non-small cell
lung cancer (NSCLC) were 79.1%, 45.2%, and 22.2%, respectively for VATS; and 81.6%, 47.2%, and 24.1%, respectively,
for open lobectomy. Chen and colleagues from Peking University School of Oncology (personal communication, November 2011) performed surgical resection in 660 patients
and reported 5-year survival for stage I, II and III disease to
be 78.3%, 52.3%, and 32.8%, respectively. These results are
not different from the international standard.
There is only limited information on the clinical practice
in management of advanced-stage disease in China. A national survey on medical treatment of NSCLC was performed by the Sun Yat-Sen University Cancer Center in
Guangzhou (personal communication, December 2011).
Questionnaires were sent to 202 practicing doctors in 135
centers in 12 major cities (Beijing, Shanghai, Guangzhou,
Chengdu, Hangzhou, Xian, Jinan, Wuhan, Tianjin, Nanjing, Chongqing, and Shenyang). A total of 987 cases were
recorded (381 early-stage lung cancer, 606 advanced-stage
lung cancer). The majority of patients (525 of 606; 86.7%)
with advanced-stage lung cancer received first-line chemo-
434
therapy. The most commonly used regimens included gemcitabine/platinum (27.4%), docetaxel/platinum (16.2%), and
paclitaxel/platinum (13.5%). For patients with adenocarcinoma, approximately 16% would use pemetrexed/platinum.
Only 4.9% used first-line EGFR TKI. The main reason is
that the number of patients who underwent EGFR mutation
analysis was low. Only 47 patients underwent EGFR mutation analysis, of whom 22 (47%) were positive for the
mutation. Demographics of the tested population are not
available, but the high mutation rate suggested that patients were clinically selected (female, nonsmoker, adenocarcinoma) for testing. Only 54 patients (9%) received secondline treatment. Interestingly, gemcitabine/platinum was
still the most popular regimen, whereas single-agent docetaxel, gefitinib, and pemetrexed account for 13%, 11%, and
9.3%, respectively. According to the authors, this is the first
national survey in China. This reflects the true clinical
practice across vast geographic areas. However, the study
falls short because of the relatively small sample size for a
large number of participating centers. On average, there
were fewer than eight patients from each center, and these
patients may not be representative of the clinical practice of
hosting hospital. This survey provides only a rough view of
current practice in management of advanced NSCLC in
China.
Molecular-targeted therapy has become an important part
of disease management. China contributed a substantial
proportion of patients to the IPASS study that established
the role of gefitinib in patients with lung cancer with EGFR
mutation.10 However, the afore-described survey suggested
that molecular testing is not widely practiced in China. A
total of 26 hospitals in China have in-house capacity to test
for EGFR mutation as a standard service, whereas another
50 hospitals routinely send their specimens to other hospitals or private laboratories. In 2011, an estimated 12,000
EGFR mutation analyses were performed, which is only a
small fraction of patients with lung cancer in China. The
cost of testing and availability of tumor sample are the
major reasons for not testing; furthermore, the majority of
patients were not able to afford the expensive EGFR TKI.
The Research
China has become an important partner and contributor

in laboratory, translational, and clinical research for lung
cancer. The nations case number represents the largest
patient resource in the world. Educated investigators are
LUNG CANCER IN CHINA

Table 1. Ongoing Chinese Thoracic Oncology Group Studies
Study No.
NCT No.
Investigational Drug(s)
CTONG 0801
NCT00765687
Bisphosphonates
CTONG 0802
NCT00874419
Erlotinib
CTONG 0803
NCT00663689
Erlotinib
CTONG 0804
NCT00770588
Gefitinib
CTONG 0805
NCT00922584
Sorafenib
CTONG 0806
NCT00891579
Pemetrexed/gefitinib
CTONG 0807
NCT00816868
Erlotinib/capecitabine
CTONG 0901
NCT01024413
Erlotinib/gefitinib
CTONG 0902
NCT00883779
Erlotinib
CTONG 0904
NCT01038661
Docetaxel
CTONG 1001
NCT01319669
rhTPO
CTONG 1002
NCT01236716
Nab-paclitaxel/gemcitabine
CTONG 1003
NCT01175096
Rad001 (everolimus)
CTONG 1101
NCT01297101
Erlotinib
CTONG 1102
CTONG 1103
NCT 01407822
Gefitinib
Erlotinib
CTONG 1104
NCT01405079
Gefitinib
Title
Screening Non Small Cell Lung Cancer With Bone Metastasis and
Efficacy and Safety Research of Receiving Bisphosphonates
(BLEST)
Erlotinib Versus Gemcitabine/Carboplatin in Chemo-naive Stage
IIIB/IV Non-Small Cell Lung Cancer Patients With Epidermal
Growth Factor Receptor (EGFR) Exon 19 or 21 Mutation(Optimal)
Efficacy of Erlotinib for Brain Metastasis of Non-Small Cell Lung
Cancer
Assess the Efficacy, Safety and Tolerability of Gefitinib (Iressa 250
mg) as Maintenance Therapy in Locally Advanced or Metastatic
(Stage IIIB/IV) Non Small Cell Lung Cancer (NSCLC) (INFORM)
Sorafenib Treatment in Non-Small Cell Lung Cancer After Failure of
Epidermal Growth Factor Receptor-Tyrosine Kinase Inhibitor
Study of Pemetrexed Versus Gefitinib in Patients With Locally
Advanced or Metastatic Non Small Cell Lung Cancer Who Have
Previously Received Platinum-Based Chemotherapy Without
Epidermal Growth Factor Receptor (EGFR) Mutations
A Study of TX Regimen as First-Line Treatment in Elderly Patients
With Stage IIIB/IV Adenocarcinoma Non-Small Cell Lung Cancer
Erlotinib Versus Gefitinib in Advanced Non Small Cell Lung Cancer
With exon21 MutationA,306:A Randomized Trial
A Study of Tarceva (Erlotinib) or Placebo in Combination With
Platinum-Based Therapy as First Line Treatment in Patients With
Advanced or Recurrent Non-Small Cell Lung Cancer
Tax First-line Chemotherapy With Different Doses and Then
Maintenance Therapy (TFINE)
Clinical Trial on the Prevention of Thrombocytopenia After First-line
Chemotherapy
Nab-Paclitaxel Treatment in Advanced Squamous Cell Carcinoma of
Lung
Safety and Tolerability Profile of RAD001 Daily in Chinese Patients
With Advanced Pulmonary Neuroendocrine Tumor
A Single Arm, One Center, Phase II Study of Sequential
Administration of Erlotinib in Combination with
Gemcitabine/Cisplatin As Neoadjuvant Treatment in Patients with
Stage IIIA NSCLC
Iressa Versus Chemo As Intermittent Treatment in Advanced NSCLC
Erlotinib Versus Chemo As Neoadjuvant in IIIA-N2 NSCLC with
EGFR Mutation in Exon 19 or 21
Gefitinib Versus Vinorelbine/Platinum As Adjuvant Treatment in
Stage II - IIIA(N1-N2) NSCLC with EGFR Mutation
Status
Ongoing
Ongoing
Ongoing
Completed
Recruiting
Recruiting
Ongoing
Recruiting
Ongoing
Recruiting
Recruiting
Not yet opening
Ongoing
Recruiting
Not yet opening

Ongoing
Ongoing
Abbreviations: EGFR, epidermal growth factor receptor; NCT, National Clinical Trial; NSCLC, non-small cell lung cancer; rhTPO, recombinant human thrombopoietin;
TX, treatment.
well equipped and organized for clinical trials. Impressed

by the large market size and potential growth, multiple
pharmaceutical companies have set up headquarters and
research facilities in China. The combination of pharmaceutical sponsorship, experienced researchers, and huge patient
resource, positions China to be a leading country in lung
cancer research.
Clinical research in China took off in 2004 on the discovery of EGFR mutation (Fig. 2). Incidence of EGFR mutation
was markedly higher in the Asian population; thus, clinical
research on the mutation would be most feasible in countries
such as Japan and China. IPASS is the first major international study that compares gefitinib with standard chemotherapy in a clinical selected population for EGFR mutation,
and the translational study has confirmed the superiority of
gefitinib in the mutated population. Sixteen centers from
China contributed 372 patients (31%), and the study was
completed in record speed and quality. On the basis of this
collaboration, Professor Yi Long Wu founded the Chinese
Thoracic Oncology Group (CTONG) in 2007. CTONGs mission is to design and develop multicenter clinical trials and
provide a high level of evidence for clinical practice. They
have now 23 active centers that participate either in
CTONG studies or in global studies on behalf of CTONG.
The number of lung cancer studies has been consistently
increasing over the years, and CTONG is the leading group.
Table 1 lists the ongoing CTONG studies. CTONG 0802
(OPTIMAL study) is the first randomized study that established the role of erlotinib in patients with EGFR mutation,
and its results have helped to register erlotinib as first-line
therapy in China and Europe.11 The INFORM study
(CTONG 0804) is the first maintenance study on gefitinib,
and the positive results were orally presented at the 2011
ASCO Annual Meeting. With her relatively short history,
CTONG has completed two important studies that influenced clinical management. Other ongoing studies such as a
comparative study of erlotinib with gefitinib in patients with
exon 21 mutation (CTONG 0901) and a comparative study of
435
MOK, ZHOU, AND WU
gefitinib and chemotherapy as adjuvant therapy with patients with EGFRpositive, resectable lung cancer (CTONG
1104) are extremely promising.
There are multiple tumor banks in China, most of which
are located in major academic centers. Investigators are
very active in molecular genomic researches and have provided important translational data for novel drug development. Wu and colleagues are among the first to report in a
meta-analysis from six centers in China on the relationship
between EGFR mutation and clinical parameters.12 There
have been extensive translational works on EGFR mutations including EGFR heterogenecity and detection of EGFR
mutation from plasma DNA. More importantly, these tumor
banks started to become centralized and there is plan for
unification of clinical data. More high-quality translational
research is expected in the future.
Conclusion
China is a developing country with an enormous burden

from rising tobacco consumption and lung cancer incidence.
Governmental intervention on lung cancer prevention is
insufficient. Treatment guidelines are available, but the
heterogeneity in quality of care between centers, especially
the disparity between urban and rural areas, has resulted in
inconsistent care for patients with lung cancer. Despite the
knowledge of molecular-targeted therapy, only a small fraction of patients have access to routine EGFR mutation
analysis. Platinum-based doublet chemotherapy remains
the most commonly used regimen irrespective of mutation
status. On a positive note, clinical and translational research on lung cancer are rapidly increasing. CTONG has
already contributed substantially to the care of patients
with lung cancer.
Author
Tony S. Mok
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
AstraZeneca;
AVEO;
Boehringer
Ingelheim;
Bristol-Myers
Squibb; Lilly;
Merck Serono;
Pfizer; Roche;
Taiho
Pharmaceutical
Honoraria
AstraZeneca;
AVEO;
Boehringer
Ingelheim; Lilly;
Pfizer; Roche
Research
Funding
Expert
Testimony
Other
Remuneration
AstraZeneca
Qing Zhou*
Yi-Long Wu
AstraZeneca;
Eli Lilly; Novartis;
Pfizer; Roche;
Sanofi
REFERENCES
1. Ministry of Health of the Peoples Republic of China. Report on the Third
National Sampling Survey of Causes of Death. Beijing: The Peoples Health
Press; 2008.
2. Lee AH, Liang Y. Tobacco Control in China. In Hu TW (ed). Tobacco
Control Policy Analysis in China: Economics and Health. Berkeley, CA: World
Scientific Publishing Company; 2007.
3. Yang G, Fan L, Tan J, et al. Smoking in China: findings of the 1996
National Prevalence Survey. JAMA. 2002;282:1247-1253.
4. Ministry of Health of the Peoples Republic of China. 2007 China Tobacco
Control Report. Beijing: Ministry of Health of the Peoples Republic of China;
2007.
5. Wang JB, Jiang Y, Wei WQ, et al. Estimation of cancer incidence and
mortality attributable to smoking in China. Cancer Causes Control. 2010;21:
959-965.
6. Parkin DM, Bray F, Ferlay J, et al. Global cancer statistics, 2002. CA
Cancer J Clin. 2005;55:74-108.
7. Zhi XY, Wu YL, Bu H, et al. Chinese guideline on diagnosis and
treatment of primary lung cancer. J Thorac Dis. 2011;4:88-101.
436
8. Wang N, Chen WQ, Zhu WX, et al. [Incidence trends and pathological
characteristics of lung cancer in urban Beijing during period of 1998 2007].
Zhonghua Yu Fang Yi Xue Za Zhi. 2011;45:249-254.
9. Yang X, Wang S, Qu J. Video-assisted thoracic surgery (VATS) compares
favorable with thoracotomy for the treatment of lung cancer: a five year
outcome comparison. World J Surg. 2009;33:1857-1861.
10. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatinpaclitaxel in pulmonary adenocarcinoma. N Engl J Med. 2009;361:947957.
11. Zhou C, Wu YL, Chen G, et al. Erlotinib versus chemotherapy as
firstline treatment for patients with advanced EGFR mutation-positive nonsmall-cell lung cancer (OPTIMAL, CTONG-0802): a multicenter, open label,
randomized phase III study Lancet Oncol. 2011;12:735-742.
12. Wu YL, Zhong WZ, Li LY, et al. Epidermal growth factor receptor
mutations and their correlation with gefitinib therapy in patients with
non-small cell lung cancer: a meta-analysis based on updated individual
patient data from six medical centers in mainland China. J Thorac Oncol.
2007;2:430-439.
Research and Standard of Care: Lung Cancer

in Romania
By Tudor E. Ciuleanu, MD, PhD
Overview: In Romania, lung cancer is the most frequent

cancer in men and fourth most frequent in women, and its
incidence and mortality continue to rise. Recently, firm antitobacco policies were implemented, in agreement with the
MPOWER strategies recommended by the World Health Organization (WHO). As of January 2012, the recognized official
standard of care in lung cancer is still represented by the 2009
edition of the European Society for Medical Oncology (ESMO)
guidelines. Cancer treatment is free, as the National Program
of Oncology covers the budget for all cytotoxic agents and
targeted therapy. However, reimbursement for several expen-
OMANIA BECAME a member of the European Union

in 2007. With 21.3 million inhabitants and a size
roughly the same as Great Britain, the country inherited a
predominantly state-owned health system, with welltrained health professionals but limited resources and a
centralized organization.
Cancer is the second cause of death in Romania, following
cardiovascular diseases, and the drugs used for its management are funded by a dedicated National Program of
Oncology.1
Patterns of Lung Cancer Epidemiology
According to Globocan 2008, 70,300 new patients with

cancer and 46,300 cancer deaths were encountered in Romania.2 Lung cancer was the most frequently diagnosed cancer
in both genders. Incidence of lung cancer was 10,384 new
cases (14.8% of all cancers), with an age standardized rate
(ASR) of 30%. This compares favorably with nearby Hungary, with the highest rate of lung cancer in the world and
an ASR of 52%. Mortality in Romania was 9,427 cases
(20.4%) with an ASR of 26.8%. The 5-year prevalence was
59.3%.
According to the Romanian National Center for Health
Statistics, the crude incidence of lung cancer rose continuously in a 25-year period, rising from 32.14% in men in 1982
to 56.14% in 2007 and from 5.93% in women to 15.5%.
Within 50 years, (from 1959 to 2010), the ASR for lung
cancer mortality more than doubled both in men and women
(men, 17.65% to 48.04%; women, 4.43% to 8.94%). Not all of
the eight regional Cancer Registries are yet providing accurate data because of a lack of trained personnel, insufficient
funds, or methodological gaps. According to the Cluj County
Registry, the cumulative lifetime risk of developing lung
cancer is 7.46% for men and 1.21% for women.
As compared with Western Europeans, more Romanian
patients have squamous cell carcinoma and less adenocarcinoma. A study at the Pneumology Institute of Bucharest
found 48% squamous cell, 29% adenocarcinomas, 7% large
cell carcinomas, and 16% small cell lung cancers (SCLC)
among 7,792 patients with confirmed lung carcinomas.
Sixty-five percent of the patients were aged 55 to 74 years at
diagnosis.3
Standards of Lung Cancer Care
Until recently, the management of lung cancer was rather

heterogenous among centers, although previous collabora-
sive drugs such as pemetrexed, erlotinib, and bevacizumab is

individually approved by a centralized commission. All new
drugs registered in Europe by the European Medicines Agency
are concomitantly registered in Romania. However, no new
drugs (such as gefitinib) or new indications (such as first-line
tyrosine-kinase inhibitors or maintenance treatment) have
been accepted for reimbursement since 2008. Clinical research is rapidly growing, and Romanian centers demonstrate
a high recruitment rate in pivotal trials, despite initial delays
because of a slow approval of the studies by authorities.
tion efforts led to published Romanian therapeutic guidelines.4 In 2009, the National Commission of Oncology
officially adopted the European Society for Medical Oncology
(ESMO) guidelines, but no update followed.5 Recently, the
third Central European Cooperative Oncology Group
(CECOG) Consensus on the Treatment of non-small cell
lung cancer (NSCLC) was published and is expected to
influence the management of lung cancer in Romania.6
Briefly, the consensus included the following recommendations: (1) early (operable) NSCLC: surgery (followed by
adjuvant chemotherapy in stage II and III and in selected
patients with stage IB disease; (2) locoregionally advanced
NSCLC: combined chemoradiotherapy and; (3) advanced
NSCLC: four to six courses of cisplatin-based chemotherapy
with a third generation cytotoxic drug (pemetrexed in nonsquamous NSCLC) for first-line treatment, with bevacizumab providing modest benefit with added toxicity;
epidermal growth factor receptor (EGFR) TKIs depending
on EGFR-activating mutation status; pemetrexed as maintenance of response immediately following cisplatin-based
chemotherapy resulting in significantly improved survival
(particularly in patients with nonsquamous NSCLC). Docetaxel, pemetrexed (for nonsquamous NSCLC), or erlotinib
for second-line therapy.
First-line tryrosine kinase inhibitors (TKIs), as well as
maintenance treatment, are not reimbursed in Romania,
although the drugs are registered and available at the
patients expense.
The standard ESMO guidelines apply for SCLC, which
include: (1) limited disease: combined chemoradiotherapy;
(2) extended disease: chemotherapy alone (platinum and
etoposide); (3) prophylactic cranial radiotherapy: recommended for responders, in both limited and extended disease
and; (4) second-line chemotherapy: recommended in patients with good performance status, with drugs such as
topotecan (oral or IV), ifosfamide, taxanes, or the CAV
(cyclophosphamide, doxorubicin, vincristine) combination.
From the Medical Oncology Department, Institute of Oncology Ion Chiricuta, ClujNapoca, Romania.
Address reprint requests Tudor Ciuleanu, MD, PhD, Institute of Oncology Ion Chiricuta,
St. Republicii 34-36, Cluj-Napoca, 400015, Romania; email: tudor@iocn.ro.
1092-9118/10/1-10
437
TUDOR E. CIULEANU
Some unsolved problems and disparities still persist in the

management of patients with lung cancer because of a
generally underfinanced health care system. Although a highquality disease management program can be offered in many
of the Romanian specialized cancer units, problems persist
related to the access to a complete diagnostic work-up and an
immediate multidisciplinary treatment. Each of the 40 counties in Romania have at least one oncology department
(medical oncology with or without radiation oncology). The
largest are the Cancer Institutes Alexandru Trestioreanu
in Bucharest and Ion Chiricuta in Cluj-Napoca, with a
third one recently inaugurated in Iasi.
Fiberoptic diagnosis is performed by 150 pneumologists,
but bronchoscopy departments are lacking in nine counties.
The number of bronchoscopies per year raised continuously
between 1990 and 2010 (2,100 vs. 10,000 procedures). This
represent an overload of the main four centers (Bucuresti,
Cluj, Timisoara, and Iasi), which together perform 75% of all
the examinations. Lack of sufficient modern bronchoscopes
or biopsy devices, lack of specialized pathology and cytology
departments (decreased diagnostic accuracy), and lack of
sufficient intensive care and thoracic surgery departments
(avoidance of the procedures with a higher risk of complications) have a negative effect on the quality of diagnosis.3
Thoracic surgery has evolved, but radical surgical treatment is confined to few specialized university centers.
Twenty nine centers have radiotherapy facilities throughout
the country, served by 110 radiation oncology specialists.
However, a recent analysis revealed that overall, the equipment is outdated and insufficient only 16 centers have
high-energy radiotherapy machines (11 linear accelerators
and 15 telecobalt machines, 6 of them outdated). It is
estimated that only 27% of the patients that require radiotherapy have access to treatment, mostly because of a lack of
facilities. There is one megavoltage machine per 1 million
inhabitants, even though the European Union standard
recommends one machine per every 300,000 inhabitants.7
Therefore, concomitant chemoradiotherapy is rarely used,
Table 1. Smoking Prevalence during the Lifetime, Romania, 2004*

Gender
Male
Female
Age Group (years)
1524
2534
3544
4554
5564
%
75.4
48.7
61.8
70.5
65.0
62.7
44.2
Education
Primary school (8 yr)
Secondary school (12 yr)
High school, university
33.3
65.8
76.0
Income
Low
Medium
High
58.5
67.6
82.0
Place of residence
Rural
Small town
Big city
53.9
67.5
68.5
Geographical region
Moldova
Muntenia
Transilvania
Bucharest
62.9
54.6
68.2
73.0
Total
62.1
* Source: Center for Health Policies and Services.
with the sequential approach permitting the scheduling of

radiotherapy.
Chemotherapy and targeted therapy are delivered by the
270 specialists in medical oncology. Optimal chemotherapy
can be offered by medical oncology departments throughout
the country in a timely manner. The registration of new
drugs is simultaneous with the European Medicines Agency
registration. The access to some expensive new molecules is,
however, limited because of the reimbursement policies.
Strategies for Tobacco Control
KEY POINTS
438
The officially recommended standard of care in Romania is in agreement with the 2009 European Society for Medical Oncology guidelines.
Antitobacco policies are in place, following the
MPOWER strategies recommended by the WHO.
Lung cancer treatment in Romania is free and funded
by a National Program of Oncology, with the exception of pemetrexed and targeted treatments, which
are individually reimbursed following the decision of
a centralized National Commission.
Although new treatments are promptly registered in
Romania through a common European procedure, the
decision for reimbursement is a multistep bureaucratic process that may take several years.
Romanian centers consistently contributed to recent
research in lung cancer, and clinical research is
rapidly growing, despite delays related to the initial
approval by authorities.
A study carried out in Bucharest, between 1999 and 2001,

found 84% of 8,856 patients with lung cancer were smokers.3
Data from the Center for Health Policies and Services show
a prevalence of smoking in the general population of 62.1%
during the entire life.8 The prevalence was higher in men
age 25 to 34 years, with a high education level, high income,
and from urban areas (Table 1). A study published in 2006
ranked Romania 29 of 30 European countries in respect to
the total measures taken to control tobacco smoking. The
score was based on price, public place bans, public information campaign spending, advertising bans, health warnings,
and treatment.9
The WHO Framework Convention on Tobacco Control
(WHO FCTC) and its guidelines provide the foundation for
countries to implement and manage tobacco control. The
WHO introduced the MPOWER measures, intended to assist in the country-level implementation of effective interventions to reduce the demand for tobacco. Romanian
tobacco control strategy follows the MPOWER measures
recommended by the WHO:
Monitoring of tobacco use and prevention policies. A

periodical monitoring of smoking is done every 2 years.
LUNG CANCER IN ROMANIA
Protection of people from tobacco smoke by banning

smoking in public places, including bars and restaurants. The legislation has improved. From a total allowance of smoking to being permitted in public places only
in separated, ventilated rooms and totally banned in
medical care units and transportation.
Offering assistance for quitting tobacco use. Treatment
for tobacco addiction has been freely available in the
framework of a national program funded by the Ministry of Health since 2007. A toll-free quit-line is also
available.
Warning about the dangers of tobacco. The health warnings are covering 30% to 40% of the main surfaces of
tobacco products, according to the provisions of the
European Union Tobacco Directive. In 2008, Romania
became the second European Union country implementing the pictorial health warnings on all tobacco products.
Enforcement of bans on tobacco advertising, promotion,
and sponsorship on radio, TV, outdoor and indoor billboards, mass media, toys, nontobacco objects, and
minors-intended events.
Raising taxes on tobacco, according to the European
Union Tobacco Taxation requirements. A part of the
funds collected finances the tobacco control program
and the treatment of some smoking-related diseases.
The National Tobacco Control Program implemented in

2007 by the Ministry of Health includes prevention (coordinated by the National Center for Health Promotion), treatment, and monitoring activities (coordinated by the National
Institute of Pneumology). Educational programs and communication campaigns are organized throughout the year,
but special interventions are prepared on May 31 and
November 17 (No-Tobacco Day). A toll-free quit-line is
manned by psychologists. The medical treatment is free-ofcharge and prescribed by 60 doctors trained in smoking
cessation.
The State Sanitary Inspection and the National Authority
for Consumer Protection are mandated by law to control the
implementation of the regulations. Some public health campaigns conducted by nongovernmental organizations and
professional associations associate the stop-smoking component in their programs, addressing lung disease, cardiovascular, or cancer prevention programs.
Media campaigns, in collaboration with international
partners, such as the HELP campaign are periodically
organized. These integrated programs use television, Internet, and mobile phones to build capacity for a life without
tobacco for youngsters and young adults. Currently, funds
available do not allow for comprehensive mass-media campaigns. Community-based campaigns are difficult to organize because of limited resources and lack of coordination
between institutions.
Health Care Financing in Romania
The treatment of insured patients with cancer is free. A

National Program of Oncology was created in 2003. The
budget allocated for this program raised continuously from
2003 to 2009 and then reached a plateau. All approved
anticancer agents are included on a special list of drugs that
are 100% reimbursed, financed by the National House of
Insurance. A special budget is prospectively allocated for the
National Program of Oncology, which is covering all anticancer drugs. The acquisition is done following a national
auction. Each oncology department is periodically buying
the drugs needed according to a planned budget. Oral drugs
may also be obtained from the pharmacies outside hospitals.
However, for the expensive new drugs such as pemetrexed, MoAbs (i.e., bevacizumab), and TKIs (i.e., erlotinib),
the reimbursement is subject to an individual approval from
a National Commission. The number of patients who benefited of systemic treatments covered by the National Program of Oncology, raised from 75,000 in 2007 to 96,700 in
2010 for all cancers. Approximately 3,400 patients are
treated concomitantly with expensive drugs in Romania.
There is a mismatch between the new published guidelines (including CECOG)6 and the lack of reimbursement for
the newly registered drugs or indications. First-line TKIs for
patients with EGFR mutations are not yet reimbursed.
Maintenance chemotherapy is not yet reimbursed, even
though Romanian centers were among the most active in the
pivotal maintenance registration trials for pemetrexed and
erlotinib.10,11 The Romanian legislation foresees the yearly
update of the list of compensated drugs. However, this has
not happened since 2008, and about 30 oncology drugs still
wait for the reimbursement decision. The way from registration to reimbursement is a multistep bureaucratic
process. Additionally, a new drug may be proposed for
compensation only after a minimum of 1 year of use in at
least three European countries.
In an attempt to provide the new drugs to patients,
different risk-sharing strategies (such as cost-volume, costresult, headroom agreements, and discounts) were initiated
between authorities and the pharmaceutical companies with
variable success. At the moment of this writing, a radical
reform in the Romanian Health System is expected, which
will probably include the privatization of many hospitals
and the transfer of the financing of all medical activities
(including oncology) to several private Insurance Companies, besides the public National House of Insurance.
The Research Structure at the Cancer Institute
Ion Chiricuta
The oldest Cancer Institute in Romania was founded in

1929, located in Cluj, and is called Ion Chiricuta, after the
name of a famous Romanian cancer surgeon, who was
awarded by the Heidelberg University as the pioneer of the
use of the omentum in plastic surgery and who was integral
to the development of the Institute. It is a comprehensive
cancer center that offers cancer diagnosis and treatment
(with 550 hospital beds). The clinical departments are represented by surgery, radiation oncology (external beam
radiotherapy with two linear accelerators and one cobalt
unit and brachytherapy with low-dose rate and high-dose
rate), medical oncology (continuous hospitalization and day
hospital), and pediatric oncology. There are dedicated histology and cytology departments, a certified clinical laboratory, and an imagery plateau with computed tomography
(CT) scans, ultrasound, and scintigraphy. All categories of
personnel are represented, but there is a work overload
because of an increasing number of patients (from 727 new
patients in 1955 to 7,156 new patients in 2008). The Institute has basic science/translational research departments,
with cellular biology, immunology, and radiobiology platforms. It is an active member of OECI (Organization of the
439
TUDOR E. CIULEANU
Table 2. The Contribution of the Cancer Institute Ion Chiricuta Cluj-Napoca, Romania (CIIC), to Multicenter Trials in Lung Cancer
Setting/Study
NSCLC adjuvant/ IALT (platinum doublets versus observation)

NSCLC 1st line/
CTNR (platinum versus non platinum doublets)
Stellar 3 (paclitaxel poliglumex/carbo versus paclitaxel/carbo)
NSCLC maintenance/
CECOG (gemcitabine versus observation)
JMEN (pemetrexed versus placebo)
SATURN (erlotinib versus observation)
NSCLC rescue treatments/
BR 21 (erlotinib versus placebo)
ISEL (gefitinib versus placebo)
TITAN (erlotinib versus docetaxel or pemetrexed)
SCLC prophylactic cranial irradiation (PCI) PCI 9901 (2 doses of PCI)
SCLC 1st line JMHO (pemetrexed/carbo versus etoposide/carbo)
SCLC rescue treatment/406 (oral topotecan versus observation)
Biogenerics/XM 02-INT 03 (biogeneric versus brand name G-CSF)
Total
Phase
Number pts CIIC/all pts
% pts CIIC
Reference
III
55/1,867
2.9%
13
IIR
III
40/102
19/400
39.2
4.75
14
15
III
III
III
24/352
70/663
65/889
6.8
10.6
7.3
18
10
11
III
III
III
III
III
III
III
68/731
58/1,692
48/424
35/720
20/908
13/141
34/240
549/9,129
9.3
3.4
11.3
4.9%
2.2
9.2%
14.2
6.01
16
17
23
21
19
20
22
Abbreviations: NSCLC, non-small cell lung cancer; IALT, International Adjuvant Lung Cancer Trial; CECOG, Central European Cooperative Oncology Group; SCLC,
small cell lung cancer.
European Cancer Institutes) and part of the CECOG and

BUON (Balkan Union of Oncology) network.
Highlights of Romanian Participation in Lung
Cancer Research
Clinical Trials and the Specific National Environment in Oncology
Clinical research has been growing steadily in Romania.

Although heterogeneity exists in the level of care, an increasing number of state-owned and private centers qualify
for participation in clinical studies. According to the Romanian National Drug Agency, the number of phase II and
phase III studies to which Romanian centers are contributing patients doubled in the past 2 years.
The climate in cancer research is influenced by three
factors: economic (a negative trend in 2009 2012, affected
by the global crisis), political (wind shear, with the permanent need to reform the sanitary system carried out by each
new government, the close parliamentary/presidential elections), and scientific (good premises with an increasing
presence in international studies, high recruitment potential, and increasingly experienced investigators).
Local Versus International Research Priorities
There is a steep decrease in local academic research in

favor of the international research, mainly industrysponsored clinical trials. Based on the European legislation
(Directive 2001/20/EC), clinical trials must get ethical approval and approval from the competent authorities. However, the duration of these regulatory procedures to initiate
a clinical trial is a factor determining the competitive
position in clinical research.
An analysis of the time interval between final protocol
approval (FPA) and inclusion of the first patient into randomized clinical trials was performed for six multicenter
trials in 25 CECOG study centers, including Romania.12 The
average time interval from FPA to the inclusion of the first
patient was 18.4 months. Most of this time has been spent
for regulatory procedures, i.e., the approval by the Ethical
Review Boards (9.6 7.2 months) and CAs (10.0 6.6
months). The Letters of Agreement were signed 11.5 to 9.4
months after FPA. As the regulatory procedures accounted
440
for more than 50% of the duration of the whole paper to

patient process, optimization is necessary to make novel
therapies available to patients more quickly.
As of January 2012, a total of 1,107 studies involved
Romanian sites for all the medical specialties, according to
clinicaltrials.gov. Fifty-eight involved lung cancers (compared with 48 for breast and 46 for digestive tumors).
Romanian centers participated mainly in phase III trials,34
followed by phase II,19 phase IV,3 and phase I/II (two
studies). Among these, 16 are actively recruiting. The investigational products in these studies were as follows: bevacizumab, BIBF1120, BNP7787, CS7017, darbepoetin alfa,
erlotinib, ganetespib, gefitinib, IMC-11F8, ipilimumab, LY2523355, necitumumab, obatoclax, onartuzumab, OSI-906, pemetrexed, picoplatin, ramucirumab, stimuvax, and sunitinib.
Since 2004, the Cancer Institute Ion Chiricuta recruited
549 of the total of 9,129 patients in 13 finalized multicentric
trials, representing 6% of the overall population included.
The participation ranged from 2.2% to 39.2% between studies (Table 2).13-23
The results lead to the United States Food and Drug
Administration (FDA) and/or European Medicines Agency
registration of several new molecules (i.e., erlotinib, oral
topotecan), new biogenerics (filgrastim), or new indications
(i.e., maintenance for pemetrexed and erlotinib). Some of
these trials helped to define the role of different therapeutic
modalities in lung cancer such as adjuvant chemotherapy
(IALT), rescue treatments (BR21, ISEL, TITAN), maintenance therapy (JMEN, SATURN), and prophylactic cranial
irradiation (PCI99-01). The first ASCO International Clinical Trials Workshop was held in Cluj in 2011, focusing on the
best practice in the implementation of a protocol and an
overview of clinical trial design.
Conclusion
Lung cancer constitutes an important health problem in

Romania, with a rising incidence and mortality. The implemented antitobacco strategies are expected to slow down
this trend in the future. All new drugs are simultaneously
registered in Europe and in Romania. However, a generally
underfinanced health system does not cover an expedited
LUNG CANCER IN ROMANIA
reimbursement of the new drugs for standard care. Clinical

research is constantly growing, and Romanian centers al-
ready showed a high recruitment rate in several pivotal

trials.
Author
Tudor E. Ciuleanu
Employment or
Leadership
Positions
Consultant or
Advisory Role
GlaxoSmithKline;
Lilly; Merck;
OSIP; Pfizer;
Roche
Stock
Ownership
Honoraria
GlaxoSmithKline;
Lilly; Merck;
Novartis; OSIP;
Pfizer; Roche
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Eniu A. International Insight: Cancer in Romania. ASCO News &
Forum. July 2009:42-43.
2. GLOBOCAN 2008 database (version 1.2). http://globocan.iarc.fr/. Accessed online January 2012.
3. Ulmeanu R. Bronchology in Romania: Where to? Pneumologia. 2006;55:
147-150.
4. Ciuleanu TE, Dediu M, Rusu P, et al. Lung cancer: Diagnostic and
Treatment Guideline. J Radiother & Med Oncol. 2007;13:5-18.
5. DAddario G, Felip E. On behalf of the ESMO Guidelines Working
Group. Non-small-cell lung cancer: ESMO Clinical Recommendations for
diagnosis, treatment and follow-up. Ann Oncol. 2009;20(suppl 4):68-70.
6. Brodowicz T, Ciuleanu T, Crawford J, et al. Third CECOG consensus on
the systemic treatment of non-small-cell lung cancer. Ann Oncol. Epub 2011
Sept 22.
7. Cernea V, Nagy V, Irimie A, et al. Report Regarding the Present State of
Radiotherapy Laboratories in Romania, National Program for Radiotherapy
2008. J Radiother & Med Oncol. 2008;15:7-24.
8. Vladescu C, Mihalan F, Sanda L. On behalf of the Center for Health
Policies and Services. Smoking and Public Health in Romania, 2004. http://
www.stopfumat.eu/Materiale/Studiu_CPSS_04. Accessed online January
2012.
9. Joosens L, Raw M. The Tobacco Control Scale: A new scale to measure
country activity. Tob Control. 2006;15:247-253.
10. Ciuleanu T, Brodowicz T, Zielinski C, et al. Maintenance pemetrexed
plus best supportive care versus placebo plus best supportive care for
non-small-cell lung cancer: A randomised, double-blind, phase 3 study.
Lancet. 2009;374:1432-1440.
11. Cappuzzo F, Ciuleanu T, Stelmakh L, et al. Erlotinib as maintenance
treatment in advanced non-small-cell lung cancer: A multicentre, randomised, placebo-controlled phase 3 study. Lancet Oncol. 2010;11:521-529.
12. Brodowicz, I. Steiner, S. Beslija, et al. Time interval between final
protocol approval (FPA) and inclusion of the first patient into randomized
clinical trials (RCTs) performed by the Central European Cooperative Oncology Group (CECOG): A 10-year experience. J Clin Oncol. 2009;27:15s (suppl;
abstr 6546).
13. The International Adjuvant Lung Cancer Trial Collaborative Group.
Cisplatin-Based Adjuvant Chemotherapy in Patients with Completely Resected NonSmall-Cell Lung Cancer. N Engl J Med. 2004;350:351-360.
14. Grigorescu Al, Ciuleanu T, Firoiu E, et al. A randomized phase II trial
of sequential gemcitabine plus vinorelbine followed by gemcitabine plus
ifosfamide versus gemcitabine plus cisplatin in the treatment of chemo-naive
patients with stages III and IV non-small cell lung cancer (NSCLC). Lung
Cancer. 2007;57:168-174.
15. Langer CJ, OByrne KJ, Socinski MA, et al. Phase III Trial Comparing
Paclitaxel Poliglumex (CT-2103, PPX) in Combination with Carboplatin
Versus Standard Paclitaxel and Carboplatin in the Treatment of PS 2
Patients with Chemotherapy-Naive Advanced Non-small Cell Lung Cancer.
J Thorac Oncol. 2008;3:623-630.
16. Shepherd FA, Pereira JR, T Ciuleanu T, et al. Erlotinib in Previously
Treated Non-Small-Cell Lung Cancer. N Engl J Med. 2005;353:123-132.
17. Thatcher N, Chang A, Parikh P, et al. Gefitinib plus best supportive
care in previously treated patients with refractory advanced non-small-cell
lung cancer: Results from a randomised, placebo-controlled, multicentre
study (Iressa Survival Evaluation in Lung Cancer). Lancet. 2005;366:15271537.
18. Brodowicz T, Krzakowski M, Zwitter M, et al. Cisplatin and gemcitabine first-line chemotherapy followed by maintenance gemcitabine or best
supportive care in advanced non-small cell lung cancer: A phase III trial for
the Central European Cooperative Oncology Group (CECOG). Lung Cancer.
2006;52:155-163.
19. Socinski MA, Smit EF, Lorigan R, et al. Phase III Study of Pemetrexed
Plus Carboplatin Compared With Etoposide Plus Carboplatin in
Chemotherapy-Naive Patients With Extensive-Stage Small-Cell Lung Cancer. J Clin Oncol. 2009;27:4787-4792.
20. OBrien MER, Ciuleanu TE, Tsekov H, et al. Phase III Trial Comparing
Supportive Care Alone With Supportive Care With Oral Topotecan in Patients With Relapsed Small-Cell Lung Cancer. J Clin Oncol. 2006;24:54415447.
21. Le Pechoux C, Dunant A, Senan S, et al. Standard-dose versus
higher-dose prophylactic cranial irradiation (PCI) in patients with limitedstage small-cell lung cancer in complete remission after chemotherapy and
thoracic radiotherapy (PCI 99-01, EORTC 22003-08004, RTOG 0212, and
IFCT 99-01): a randomised clinical trial. Lancet Oncol. 2009;10:467-474.
22. Gatzemeier U, Ciuleanu T, Dediu M, et al. XM02, the First Biosimilar
G-CSF, is Safe and Effective in Reducing the Duration of Severe Neutropenia
and Incidence of Febrile Neutropenia in Patients with Small Cell or Nonsmall Cell Lung Cancer Receiving Platinum-Based Chemotherapy. J Thorac
Oncol. 2009;4:736-740.
23. Ciuleanu T, Stelmakh L, Cicenas S, et al. Efficacy and safety of
erlotinib versus chemotherapy in second-line treatment of patients with
advanced, non-small-cell lung cancer with poor prognosis (TITAN): a randomised multicentre, open-label, phase 3 study. Lancet Oncol. 2012; [epub
ahead of print].
441
LEVERAGING VIRTUAL PATIENT COMMUNITIES

FOR OPTIMAL CLINICAL CARE AND RESEARCH
CHAIR
Howard J. West, MD
Swedish Cancer Institute
Seattle, WA
SPEAKERS
Dana-Farber Cancer Institute, Harvard Medical School
Boston, MA
Dave deBronkart, SB
Boston, MA
A New Model: Physician-Patient Collaboration

in Online Communities and the Clinical
Practice of Oncology
By Howard J. West, MD, Dave deBronkart, SB, and George D. Demetri, MD
Overview: The practice of medicine is in the midst of a

fundamental transformation based on the new availability of
health information through the Internet and other sources
accessible by the broad lay public, as well as on the easy
sharing of experiences and content through social media. This
is occurring at a time when the volume of new information
required for optimal medical care is exceeding that which an
individual physician can feasibly follow and master. The
changes in cancer care are especially acute as we experience
an ongoing reclassification of many disease entities to reflect
divisions by molecular variables, often with new clinical options now optimized for very limited patient subsets. The
increasing complexity of the field, combined with the high
stakes of optimizing treatment decisions and the growing
HE INTERNET has transformed many industries,

ranging from news media to travel, real estate, and
politics, with the practice of medicine also in the midst of its
own belated disruption as well. For centuries, our dominant
medical model has been based on a unidirectional flow of
information, in which only physicians had access to the
knowledge sufficient to weigh differential diagnoses and to
recommend optimal treatments. Patients readily accepted
physician wisdom and recommendations as the only readily
available source of medical information.
Among the factors that have dramatically altered the
course of medical practice, and oncology in particular, are
two central issues. The first is that the sheer amount of new
medical information emerging over the past few decades has
ballooned to the point that no single physician could possibly
maintain expertise in the full breadth of topics required to
care optimally for a full range of patients and clinical
problems. In particular, oncology has exploded with new
content that is a boon in defining a new era of mechanistic
understanding of cancer pathophysiology, as well as necessary for the rational development and use of molecularly
targeted therapies; however, it is simply infeasible for an
oncologist to internalize the depth and breadth of new
content. The second core element is that this content is now
readily accessible to the lay public, including patients and
caregivers, essentially without limitation and in real time
perhaps even before the clinician has become aware of it.
Engaged Patients Usher in a New Model of

Medical Practice
There is clear evidence that the amount of medical information for any one doctor to consume and digest has grown
remarkably. PubMed (www.ncbi.nlm.nih.gov/pubmed/) currently categorizes more than 21 million citations, with new
publications being added at a rate of approximately one per
minutea rate that has more than doubled over the past
20 years.1 The number of medical journals has also increased, and this is especially true for oncology, where there
are now approximately 180 journals covering just this one
area of clinical subspecialty practice.2 As new research
availability of a wide range of information in the public domain,

make oncology an area in which patients and caregivers are
most motivated to become active seekers of medical information and participants in their care decisions.
The credibility of the available online information in such a
situation has emerged as a critical issue, but physicians have
historically been reluctant to create content or interact with
the lay public in online patient communities. Here we will
highlight several examples of collaborative engagement between health care professionals and motivated patients in an
online environment that illustrate how a new bidirectional or
even networked model that is a product of the Internet age can
accelerate clinical research and improve delivery of cancer
care.
results are released early online and new trial data are often
released via a press release long before being presented
and/or published in a peer-reviewed setting, physicians only
have more sources of practice-changing information to interpret and incorporate into their management recommendations. This broader distribution of medical information is
also occurring as physicians are typically pressed to see
more patients in less time.
Concurrent with these changes, patients and caregivers
are seeking both support and medical information online
at a rapidly escalating pacefrom 25% of U.S. adults in
2,000% to 61% by 20103with this trend rapidly accelerating as fast Internet connections become more readily available. People also have become increasingly comfortable with
seeking and expecting to find relevant content online free
of charge. Among health care topics, those that are lifethreatening and/or chronic lend themselves best to online
patient communities and to Internet-based searches for
information.
This is unquestionably a mixed blessing for patients and
physicians alike. Internet searches can instantaneously return reliable and timely information but also might deliver
content from unreliable sources that instead prey on the
desperation of people by peddling the false hope of a Miracle
Cancer Cure (www.theCancerCureMiracle.com) (Fig. 1).
Beyond charlatans misleading patients and caregivers for
profit, online communities provide a wide range of recommendations from people who might be knowledgeable and
well intended but who might definitively promote treatment
ideas that are not proven to be superior to others or are
possibly even detrimental. Such information may directly
compete with thoughtful recommendations from qualified
medical and scientific professionals. The fact that the Inter-
From the Department of Thoracic Oncology, Swedish Cancer Institute, Seattle, WA.
Address reprint requests to Howard West, MD, Swedish Cancer Institute, 1221 Madison
St., Suite 1020, Seattle, WA 98104; email: howard.west@swedish.org.
1092-9118/10/1-10
443
WEST, DEBRONKART, AND DEMETRI
net makes it easier than ever to share ideas and information

means that the difference between this being a net benefit or
harm depends entirely on the quality of the information
available.
Practicing physicians have historically been reluctant to
create content or to participate in online communities for the
lay public. The most frequently cited reasons for physician
reluctance to engage online are wariness about liability,
patient privacy concerns, and a lack of time or reimbursement for these efforts (Fig. 2).4 Although there are a growing
number of online resources with professionally vetted information for patients and caregivers, these are few and far
between compared with the volume of content from nonprofessional sources. In addition, although a growing proportion of physicians participate in online communities for
physicians, these are generally completely segregated from
patient-oriented communities, so that there are very few
online settings in which physicians and patients interact
together online.
Yet, the potential benefits for online physician engagement are very substantial, not only for patients but also for
physicians and for the global practice of medicine. Patients
KEY POINTS
444
Medicine in general, and oncology in particular, is

now experiencing a rapid growth in new content and
a reclassification of many cancers into smaller, molecularly defined subgroups that has made it infeasible for an individual physician to maintain sufficient
expertise to remain the sole source of treatment
information.
Patients and caregivers are increasingly turning to
online sources to supplement what they learn from
their own medical care teams, and this can be a
beneficial or detrimental change depending entirely
on the quality of the medical information available
online.
Physicians have historically remained wary to engage in online communities or to provide information
via the Internet, largely because of concerns about
legal liability, patient privacy, and time limitations.
Despite these challenges, there is a wide range of
examples in which physicians have produced credible
online content and/or partnered with communities of
motivated patients and caregivers facilitated by
Internet-based platforms to distribute highly relevant and high-quality information and to conduct
valued clinical research.
Reliance on Internet-based information will only escalate as more patients and caregivers become connected, making it more compelling for physicians to
embrace the constructive possibilities of collaborative
engagement and interaction between patients and
caregivers online rather than passively cede the ability to influence these online communities in a constructive way.
Fig. 1. Medical advice from unqualified sources is plentiful online

(theCancerCureMiracle.com).
rank a physicians input as the most influential factor in

shaping their care decisions (see Table 1),5 and timely and
credible information from experts can help overcome the gulf
between the prohibitive expanse of information for a patients care and for the amount that any single physician can
review and retain. With the expanse of new content, physicians are now apt to become a bottleneck and to limit care
options if medical practice follows an outdated unidirectional model. Increasingly, patients with access to a network
of committed advocates, potentially including many other
fellow patients, in the context of highly accessible medical
information can collaborate with their own local medical
team to shape a more bidirectional or even a networked
model in which patients conduct their own research and
discuss their own views regarding the most appropriate or
preferred options with their physicians. This respects the
physicians role as an expert and caring provider, while also
recognizing that no single physician can be expected to be
the sole source of all medical knowledge. In this new model,
the physician becomes the pivotal individual to provide
context and to shape recommendations for a patient who is
increasingly engaged in key decisions.
Provision of Readily Available, Vetted Content
through Physician Engagement
The substantial benefits from this new model are all

predicated on high-quality information being available online from identifiable, credible sources to counter the abundance of less reliable information, or even misinformation,
being promulgated by less qualified or less competent
sources. Practically speaking, physicians can engage by
offering pushed online content, in which knowledge is
offered and can be consumed by a limitless number of people
who can view this digital content without any additional
effort from its producer. Examples of this type of content
include blog posts, audio and video podcasts, or information
found on micro-blogging platforms such as Twitter. The
PHYSICIAN ENGAGEMENT IN ONLINE PATIENT COMMUNITIES
Fig. 2.
Top concerns of physicians regarding interaction with patients online (respondents choose up to three).4
alternative, pulled contentpatient solicitation of answers on an interactive online community or participation

in discussions via Facebook or Twitter offer the benefit of
great appeal to the lay audience because of potentially more
individualized and, therefore, valuable answers. These efforts create challenges, however, by being more time consuming and not scalable like pushed content. In addition,
pulled, interactive content that involves individualized
case discussions could potentially entail an implied medical
recommendation in the absence of broad disclaimers and
careful wording. Additionally, there is a risk of oversimplification because the optimal practice of medicine always
depends on the totality of details and accuracy of the
information. In the most egregious example, patients with a
mistaken diagnosis might not even have the disease that
they think they have.
Patients can clearly benefit from becoming far more informed about appropriate treatment options and may learn
about standard or clinical research-based options about
which their own physician might not be aware. They can
also be comforted by the ability to play an active role in
achieving a consensus about the optimal treatment for a
complex situation. However, benefits are also readily available for physicians who are willing to invest the time to
impart their knowledge into global online discussions. Aside
from the very substantial intrinsic value of enabling more
patients to participate directly in their own care while
armed with vetted information, physicians who engage with
patients online may need to spend less time covering these
same topics repeatedly in individual discussions, may be
Table 1. Physicians Remain Most Valued Source of

Health Information 5
Who is more helpful when you need . . .
An accurate medical diagnosis

Information about prescription drugs
Information about alternative treatments
A recommendation for a doctor or specialist
A recommendation for a hospital or other
medical facility
Fellow patients, Both

Professional
equally
friends, and
sources, e.g.,
(%)
family (%)
MD and RN (%)
91
85
63
62
62
5
9
24
27
27
2
3
5
6
6
better trusted by their patients, and may attract more new

patients based on their online outreach.
The medical community can potentially curate highquality content, such as by creating an online resource akin
to a Wikipedia or Khan Academy of medical information accessible to physicians and to the lay public alike. In
many ways, this was the impetus for ASCO to develop the
professionally vetted Web resource CancerNet (www.cancer.
net) to facilitate these new professional roles and responsibilities with the public. Effective professional activities in
this regard will engender two critical and increasingly
necessary fundamental changes. First, physicians will no
longer need to hold an unmanageable capacity of specialized
knowledge but can work with patients to access and interpret the best data to create an optimal management strategy. By necessity, this change will also be accompanied by
an increased physician acceptance of outside sources of new
information as relevant and potentially valid. Secondly,
rather than have the same content recreated and recapitulated for thousands of patients thousands of times individually, this information can be shared communally, freeing
time otherwise spent in reduplicated efforts.
One example of successful engagement of multiple oncologists providing timely content to a patient and caregiver
population is provided by the Global Resource for Advancing
Cancer Education (GRACE, CancerGRACE.org). GRACE is
a nonprofit organization comprised primarily of an expert
physician-mediated online forum that started with a focus
on lung cancer and recently expanding into other cancer
subtypes, in which one of the authors (Dr. West) along with
other oncology experts distill the latest trial results and
summarize current best practices, along with personal perspectives, in accessible language and multiple formats.
These formats include blog posts, video and audio podcasts,
and a very popular interactive discussion forum in which
patients and caregivers can ask experts, and each other,
questions about the best current treatments and new research concepts just emerging. Although requiring time and
effort to develop and maintain, this resource offering both
pushed and interactive content efficiently delivers expertquality information, enabling tens of thousands of highly
motivated people every month from all over the world to
445
become extremely sophisticated about the leading treatments and promising new trial-based options.
Other important new online vehicles are being developed
by oncology professionals along these lines as well. Cancer
Commons (cancercommons.org) is a nonprofit, open-science
initiative that focuses on several forms of cancers for which
the molecular understanding is rapidly evolving, with important implications for current clinical research and care
options. Using highly interactive online models as teaching
tools, this site aims to illuminate complex scientific pathways and bring them into the clinical context in a meaningful way for patients, caregivers, and health care
professionals.
Online Medical Information: Facilitating Research in
Small and Geographically Distributed Subgroups
Centralized online content provided by medical experts

can help overcome an emerging new challenge that is a
by-product of the recognition that patients with any given
cancer diagnosis viewed as a single monolithic group are
actually far more heterogeneous, comprising many, much
smaller subgroups. The rapid evolution of molecular oncology and molecular diagnostics over the past few years has
led the erosion of large populations of stage-specific lung
cancer or breast cancer into groups defined by clinically
relevant molecular markers that redefine diagnostic groups,
natural history, prognosis, optimal treatments and appropriate trial populations such as ALK-positive nonsmall cell
lung cancer or triple-negative breast cancer. In fact, it is
widely expected that cancer as a term will cede to the more
accurate representation of subtypes of cancers as a multidimensional group of many different types of rare diseases.
The implications of this shift for clinical research and care
are clear and significant. When an eligible trial population
transitions from an easily accessible pool of more than
tens of thousands of potentially eligible patients to a much
smaller subgroup of perhaps less than 1,000 potential candidates who are broadly geographically distributed with a
very low density, researchers are no longer able to pursue
the classical model in which hundreds of centers offer the
same clinical trial (e.g., standard chemotherapy with or
without novel agent X).
Instead, research will be facilitated if these much smaller
yet geographically diverse populations of rare subsets can be
identified (or ideally, self-identify) and, increasingly, if these
patients are enabled to actively seek out treatment at a
very limited number of highly specialized centers offering a
menu of unique clinical trials for this limited subgroup. To
borrow a sartorial phrase, this could be considered as
bespoke clinical research. Although inconvenient and expensive, this model has been proven to be effective by the
development of agents such as crizotinib for nonsmall cell
lung carcinoma driven by genomic aberrancies in the ALK
kinase, as well as vemurafenib for V600E BRAF mutationdriven melanomasettings in which the anticipated benefits far exceed those we typically see for broad, untargeted
populations.
Patients with rare cancers, acquired resistance to effective
therapies, and other narrowly defined subgroups have even
self-aggregated online to facilitate clinical research, an effort that has been extremely fruitful when these coordinated
efforts by motivated patients are encouraged and facilitated
446
by medical professionals. More than a decade ago, at least

two different self-organized collections of patients with gastrointestinal stromal tumors (GISTs) and their caregivers
were formed, including the Life Raft Group and GIST
Support International. Similar groups subsequently formed
internationally, such as Das Lebenhaus in Germany and
Ensemble Contre le GIST in France. These online groups
began sharing and aggregating their experiences with this
rare disease, which had been inconsistently identified or
diagnosed before the year 2000, brought together by the
breakthrough results demonstrated with the tyrosine kinase
inhibitor, imatinib. This patient-driven online community
derived, in part, from prior online collaborations with one
the authors (Dr. Demetri), who was participating actively in
patient-moderated online discussions in the late 1990s under the auspices of the nonprofit organization known as
the Association of Online Cancer Resources (acor.org). This
organization was founded by a private philanthropic individual, Gilles Frydman, who had been frustrated by the
difficulties he encountered in finding up-to-date information
about cancer. The GIST online patient communities were
very helpful in driving awareness of this new diagnosis in a
relatively rare group of patients, and this awareness helped
stimulate interest and participation in critically important
clinical trials and clinical research initiatives linking investigators directly with the patient population that would be
the subject and ultimate beneficiary of this research. Interestingly, certain groups were facilitated in their operations
by support from the pharmaceutical firms that were developing and marketing imatinib and other targeted therapies,
representing a complex intersection of corporate-derived
philanthropic support to patient support groups with
corporate-sponsored outreach to a target population of consumers and patients. These patient communities have been
highly visible, promulgating viewpoints and even making
public presentations of the collated, self-reported patient
experiences from their database including a qualitative
patient-developed scale for rating severity of toxicity.6 Although the reliability and rigor of this methodology might be
questionable, it is certainly a testament to the engagement
of this very dedicated community of patients and caregivers
with a devotion to this disease.
The website PatientsLikeMe (PatientsLikeMe.com) has
now enrolled more than 100,000 patients with a wide range
of medical conditions, as a means to share experiences,
provide mutual support, and also facilitate clinical research
on the aggregated patient populations. Its own representatives have even conducted an observational study of a
relatively uniform population of patients with amyotrophic
lateral sclerosis who were on lithium carbonate treatment.
Representatives also collected and recently published efficacy and toxicity data on participating patients who were
compared statistically with matched controls.7
Finally, the accelerating ability of Internet-based patient
recruitment for clinical research on rare and globally distributed populations is illustrated by other examples. A
relatively large study aiming to accrue 1,000 people who
have been diagnosed with any form of sarcoma is in progress
under the sponsorship of 23andMe (www.23andMe.com), a
company that aims to link genome-wide screening analyses
with elements of online community and social networking. 8
This study may subsequently expand to evaluate first- and
PHYSICIAN ENGAGEMENT IN ONLINE PATIENT COMMUNITIES
second-generation relatives to assess risk factors that may

be related to why people may be at risk of developing such
rare diseases. Another example is a small trial that focused
on the rare clinical problem of spontaneous coronary artery
dissection. This was conducted by investigators at the Mayo
Clinic and was able to enroll its target of 12 patients within
just one week of being granted institutional review board
approval.9 Although a small and limited trial outside of
the cancer setting, it is clear that the partnership of engaged
medical professionals with a coordinated and motivated
online patient population can facilitate research otherwise
infeasible without Internet-based participation.
It is critical to emphasize that each of these novel mechanisms of identifying appropriate trials for patients requires
the input and guidance of a patients primary oncologist to
provide context and guidance to review the range of options
and facilitate participation in an optimal choice, if one
exists. These network-furnished opportunities alter but do
not obviate the relationship between the patient and the
local oncologist.
Conclusions: Picking Up the Gauntlet
The Internet and related information technology has ushered in disruptive changes in the practice of medicine. The
volume of new information has grown to a point where
individual physicians increasingly find themselves unable to
feasibly master the range of material required by any but
the most specialized clinics, especially with increasingly
clinic and paperwork demands. Most significantly, this information is no longer exclusively available to physicians;
rather, it is now also available to motivated patients who
seek relevant content in hopes of better understanding their
condition and participating more actively in their own care
decisions. These trends are especially true in cancer care,
where the chronicity and often life-threatening nature of the
disease leads patients to be exceptionally motivated to learn
about a field that has become exponentially more complex
and yet also more mechanistic with this new era of molecular oncology.
As the proportion of the general public seeking health care
information online grows steadily, the quality of the content
they encounter is a concerning variable. Although patientcentered online communities often feature very knowledgeable members of the lay public who have become extremely
sophisticated, such communities are also a potential source
for misinformation that can be detrimental to good care, or

at least will compete for patient attention with more accurate, constructive educational material. The need for openminded and expert professional input is critical despite the
fact that physicians have historically been wary about engaging in such public dissemination of information by producing vetted online content.
Although the investment of time and effort for such
activities is significant, the potential for practical benefits
cannot be overstated. Patients and caregivers will seek
assistance from online sources in greater numbers regardless of whether health care professionals provide content
that is professionally reviewed for quality, accuracy, methodology, fair balance, and reliability. The physicians ability
to engage and largely direct the conversation, as well as
ensure the quality of online content, is likely to translate to
the difference between whether the newly defined relationship between patient and physician will become more oppositional or more collaborative. As we move toward an era of
bidirectional rather than unidirectional flow of health care
information, physicians have the potential to leverage the
efficiency of digital content to convey the most current
expert information broadly, opening up the possibilities of
new strategies for molecularly-guided clinical research that
will capitalize on the ability of the Internet to connect small
groups of geographically distributed people, along with the
motivation and communication within patient online communities.
The role of the oncologist, as for nearly all other physicians, is being altered in real time, and the opportunity (if
not responsibility) to provide high-quality, vetted medical
information remains crucial. There are enough examples
now of the realized potential of these efforts that more
physicians should feel compelled to engage in the professional discourse with patients online.
Importantly, no public online source will have the details
of a patients case that can inform care decisions, and even
knowledgeable sources may offer a range of perspectives on
questions with no absolute correct answer. It is therefore
critical to underscore that, even as the role of serving as a
patients primary oncologist evolves to increasingly integrate a plurality of sources of knowledge, it remains the
pivotal mechanism for vetting and prioritizing content and
its applicability for a particular patients context.
447
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Amgen; ARIAD;
Daiichi Sankyo;
Genentech;
GlaxoSmithKline;
Idera
Pharmaceuticals;
Infinity; Johnson
& Johnson;
Kolltan
Pharmaceuticals;
Merck Serono;
Momenta
Pharmaceuticals;
Novartis; Pfizer;
Plexxikon;
Ziopharm
Oncology
Champions
Biotechnology;
EmergingMed;
Kolltan
Pharmaceuticals;
Plexxikon
Honoraria
Research
Funding
Expert
Testimony
Amgen; ARIAD;
Bristol-Myers
Squibb; Daiichi
Sankyo;
Genentech;
Infinity; Johnson
& Johnson;
Novartis; Pfizer;
PharmaMar
ARIAD (U);
Infinity (U);
Johnson &
Johnson (U);
Novartis (U);
Pfizer (U);
PharmaMar (U)
Other
Remuneration
Howard J. West*
Dave deBronkart*
George D. Demetri
Novartis; Pfizer
REFERENCES
1. U.S. National Library of Medicine, Institutes of Health Web site.
http://www.nlm.nih.gov/bsd/medline_cit_counts_yr_pub.html. Accessed February 23, 2012.
2. Cancer Journals Web site. http://www.cancerindex.org/clinks9.htm. Accessed February 23, 2012.
3. Fox S, Jones S. The Social Life of Health Information. Pew Internet and
American Life Project, 2009 Web site. http://www.pewinternet.org/Reports/
2009/8-The-Social-Life-of-Health-Information/02-A-Shifting-Landscape/2-61of-adults-in-the-US-gather-health-information-online.aspx. Accessed February
23, 2012.
4. QuantiaMD web site. www.quantiamd.com. Accessed February 23, 2012.
4. Modahl M, Tompsett L, Moorhead T. Doctors, Patients & Social Media:
QuantiaMD, 2011. http://www.quantiamd.com/q-qcp/DoctorsPatientSocial
Media.pdf. Accessed February 23, 2012.
5. Fox S. Peer-to-Peer Healthcare: Pew Internet and American Life Project,
448
2011. Available at http://www.pewinternet.org//media//Files/Reports/2011/

Pew_P2PHealthcare_2011.pdf, accessed 2/10/12.
6. Call J, Scherzer NJ, Josephy PD, et al. Evaluation of self-reported
progression and correlation of imatinib dose to survival in metastatic gastrointestinal stromal tumors: An open cohort study. J Gastrointest Cancer.
2010;41(1):60-70.
7. Wicks P, Vaughan TE, Massigli MP, et al. Accelerated clinical discovery
using self-reported patient data collected online and a patient-matching
algorithm. Nat Biotechnol. 2011;29(5):411-414.
8. 23andMe Sarcoma Community: A Patient-Driven Revolution in Sarcoma Research. https://www.23andme.com/sarcoma/. Accessed February 23,
2012.
9. Tweet MS, Gulati R, Aase LA, et al. Spontaneous coronary artery
dissection: A disease-specific, social networking community-based study.
Mayo Clinic Proc. 2011;86(9):845-850.
LUNG CANCER SCREENING 101

CHAIR
Christine D. Berg, MD
Bethesda, MD
SPEAKERS
Denise R. Aberle, MD
University of California, Los Angeles David Geffen School of Medicine
Los Angeles, CA
Douglas E. Wood, MD
Seattle, WA
Lung Cancer Screening: Promise and Pitfalls

By Christine D. Berg, MD, Denise R. Aberle, MD, and Douglas E. Wood, MD
OVERVIEW: The results of the National Lung Screening Trial

(NLST) have provided the medical community and American
public with considerable optimism about the potential to
reduce lung cancer mortality with imaging-based screening.
Designed as a randomized trial, the NLST has provided the
first evidence of screening benefit by showing a 20% reduction
in lung cancer mortality and a 6.7% reduction in all-cause
mortality with low dose helical computed tomography (LDCT)
screening relative to chest X-ray. The major harms of LDCT
screening include the potential for radiation-induced carcinogenesis; high false-positivity rates in individuals without lung
cancer, and overdiagnosis. Following the results of the NLST,
the National Comprehensive Cancer Network (NCCN) published the first of multiple lung cancer screening guidelines
under development by major medical organizations. These
OR DECADES, the early detection of common cancers

has been advocated in an attempt to improve the chance
for long-term survival and cure. Breast, colon, and prostate
cancer all have established screening programs that are
covered by insurers, embraced by physicians and the public,
endorsed by professional societies and policy-makers, and
touted as critical public-health measures as the U.S. health
care system strives to prevent rather than treat disease.
Lung cancer, the leading cause of cancer death in the United
States and the world has lagged behind. There are many
reasons, including that patients with lung cancer may suffer
from the publics and policy-makers perception of lung
cancer as a self-inflicted disease. Also, the poor long term
survivorships of lung cancer patients has compromised advocacy efforts. However, we are now at the beginning of a
new and exciting era for patients with lung cancer. Revolutions in molecular genetics and modern technology have
begun to have an effect on the course of this here-to-for
highly lethal disease. For adenocarcinomas, in particular,
several targeted therapies, such as the use of erlotinib, have
emerged. Because of the National Lung Cancer Screening
Trial (NLST), medical imaging advances have recently assessed the utilization of computerized tomographic scanning
of the lung with a low radiation dose technique and provided
the medical community and patients with optimism.1
Evidence for Benefit
Previously, trials that tested lung cancer screening with

chest x-ray (CXR) showed disappointing results. Four randomized trials included one study in the Czech Republic and
three National Cancer Institute (NCI)-sponsored trials that
included sputum cytology in two trials that demonstrated no
effect on lung-cancer specific mortality.2-55 The Mayo Lung
Project (MLP), in particular, demonstrated a known problem
with screening, i.e., of overdiagnosis detecting lesions so
indolent that they are not medically significant, with 17%
more lung cancers detected in the screened arm, an excess
that persisted for at least 20 years.6 (Interestingly, a recent
reanalysis of the MLP and the Johns Hopkins Lung Project
shows some possible evidence of a very small beneficial
mortality effect with sputum analysis.7) However, as a
consequence of these studies, no major medical group recommended lung cancer screening until recently.
450
recommendations amalgamated screening cohorts, practices,

interpretations, and diagnostic follow-up based on the NLST
and other published studies to provide guidance for the
implementation of LDCT screening. There are major areas of
opportunity to optimize implementation. These include standardizing practices in the screening setting, optimizing risk
profiles for screening and for managing diagnostic evaluation
in individuals with indeterminate nodules, developing interdisciplinary screening programs in conjunction with smoking
cessation, and approaching all stakeholders systematically to
ensure the broadest education and dissemination of screening
benefits relative to risks. The incorporation of validated biomarkers of risk and preclinical lung cancer can substantially
enhance the effectiveness screening programs.
The Prostate, Lung, Colorectal, and Ovarian Cancer

screening trial (PLCO) was launched in 1993 as a multimodal screening trial with ambitious goals. One was to
assess with a large sample size the effect of CXR (posteroanterior only) screening (three rounds for nonsmokers and
four rounds for current or former smokers) on lung cancer
mortality. The trial enrolled 154, 901 individuals, 10% of
whom were current smokers and 41.5% former smokers. The
result unfortunately confirmed that this approach did not
affect lung cancer mortality.8 There was perhaps some
evidence of overdiagnosis but not of the magnitude seen
with the Mayo Lung Project.
Computed tomography (CT) has been clinically available
in the United States for decades; however, image-acquisition
time was slow until the advent of the helical CT. Also, until
it was demonstrated that a low-dose technique could reliably
image the lung parenchyma, valid concerns about radiation
dose and subsequent carcinogenesis discouraged its use for
screening a healthy population.9 With the advent of low-dose
helical computed tomography (LDCT), several groups undertook screening of at-risk individuals and reported promising results. Among these was the Early Lung Cancer
Action Program (ELCAP), in which 1000 individuals at risk
of lung cancer underwent combined chest-x-ray and LDCT
screening. A high proportion of early stage lung cancers
were observed using LDCT, and the ELCAP was among the
major studies to firmly introduce LDCT screening into the
American consciousness. A number of downsides emerged
from these various studies, including high false positivity
rates and the challenges of distinguishing true mortality
benefit from the well-known biases of lead-time, length, and
overdiagnosis that arise from single arm screening studies.
From the Early Detection Research Group, Division of Cancer Prevention, National
Cancer Institute, Bethesda, MD; Radiological Sciences, David Geffen School of Medicine at
UCLA, Los Angeles, CA; Division of Cardiothoracic Surgery, University of Washington,
Seattle, WA.
Address reprint requests to Christine Berg, MD, National Cancer Institute, Early
Detection Research Group, Division of Cancer Prevention, Executive Plaza North, Room
3112, 6130 Executive Boulevard, MSC 7346, Bethesda, MD 20892; email: bergc@
mail.nih.gov.
1092-9118/10/1-10
LUNG CANCER SCREENING
The concern in much of the scientific community was that

given these limitations, as well as the highly aggressive and
heterogeneous biology of lung cancer, the validity of LDCT
screening would require demonstrate of a mortality benefit
in order to have great public health importance. Therefore,
NCI decided that a randomized, controlled clinical trial
should be designed and conducted to determine reliably the
effect of LDCT on lung cancerspecific mortality.
The NLST was designed to answer one question in the
shortest, most definitive manner: does screening with LDCT
in an at-risk population lower lung cancerspecific mortality? A 20% mortality reduction was judged to be clinically
important and feasible to assess. The entry criteria for the
trial were set to bring a high-risk group into screening.
These criteria included current or former smokers 55 to74
years, 30 pack-years of smoking, if former smokers having
quit within 15 years. The participants had to be asymptomatic and healthy enough to withstand surgery, and could not
have had prior invasive cancers, and also not have had a CT
within the last18 months of enrollment. The trial was to
compare three rounds of screening at 12 month intervals
with LDCT compared to postero-anterior CXR. Consideration was given to having a third arm of no screening, but, as
the PLCO was ongoing, it was decided to compare the NLST
results to a matched cohort in the PLCO.11 Also, an analysis
was done to assess whether or not adding rounds of screening would have any major, additional benefit; results showed
that they did not. With these parameters, the needed sample
size was 50,000 with 90% power and an of 5% to determine
a 20% mortality reduction (p 0.05). The NCI launched the
trial, which was managed by both the Lung Screening
Study, under contract to the Division of Cancer Prevention,
and the American College of Radiology Imaging Network, a
cooperative group through the Division of Cancer Treatment
and Diagnosis.
Accrual was rapid. Over 20 months from August 2002 to
April 2004, 53,454 individuals were enrolled. Overall, the
trial participants were younger, of higher educational status, and more likely to be former smokers than those who
also matched the NLST entry criteria in the U. S. population.12 Screening was accomplished with a high degree of
KEY POINTS
Screening with low dose computed tomography

(LDCT) has been shown to reduce lung cancer by 20%
relative to chest X-ray.
Complication rates from screening and downstream
diagnostic procedures are low in individuals with
positive screens.
The major harms of screening relate to high false
positivity rates, the potential for radiation-induced
carcinogenesis, and overdiagnosis.
LDCT screening implementation should be interdisciplinary and integrated with smoking cessation programs to derive maximum benefit.
Successful dissemination of LDCT screening will require a systematically approach to address the
unique challenges of the screening centers, primary
care environment, and population at risk.
compliance in both arms95% in LDCT and 93% in CXR.

Follow-up continued after screening was completed. Regular
assessments were done through questionnaires to participants and searches through tumor registries, National
Death Index, and other sources to determine whether or not
a participant developed lung cancer, whether he or she died,
and the cause of death. A detailed endpoint verification
process was undertaken to ensure the highest degree of
accuracy and consistency in determining cause of death,
particularly for lung cancer and deaths possibly from complications related to procedures done to evaluate for lung
cancer. In October 2010, the Data and Safety Monitoring
Board observed that a stopping boundary had been crossed
and recommended that the trial cease. In November 2010,
the initial findings from the NLST were released. On June
29, 2011, the primary results were published online in the
New England Journal of Medicine and appeared in the print
issue on August 4, 2011.1
Screening with LDCT resulted in a 20% decrease in
lung-cancer specific mortality. A subset of the PLCO that
matched the NLST was analyzed. There was no evidence
that when compared with community care there was any
mortality reduction with CXR, and the lung cancerspecific
mortality in this cohort in the PLCO was the same as in the
CXR arm of the NLST. The NLST was the first ever report
from a randomized clinical trial documenting that lung
cancer mortality could be reduced with a screening modality.
Overall mortality was also lowered. However, when lung
cancer deaths were removed, this difference was no longer
statistically significant (p 0.05). Compared with CXR,
LDCT screening was associated with a stage shift towards
earlier stages for all histologies of non-small cell lung
cancer. There was no stage shift with small cell lung cancers. Unfortunately, the detection of limited small cell carcinoma was not enhanced with LDCT compared with CXR.
During screening in the LDCT arm, 649 cases of lung cancer
were diagnosed after a positive screen and 44 cases as
interval cancer, whereas in the CXR arm, 279 cases were
diagnosed after a positive screen, with 137 as interval
cancer. A total of 1060 cases of lung cancer occurred in the
LDCT arm compared with 941 in the CXR arm. Therefore,
129 additional cases of lung cancer were diagnosed in the
LDCT arm than in the CXR arm; this absolute number is not
an estimate of the amount of overdiagnosis. More follow-up
would be helpful to determine precisely the numbers of
excess cancers detected with LDCT compared with CXR that
would not come to clinical detection during a participants
life time. Alternatively, modeling can be done to better
estimate the amount of overdiagnosis. Several approaches to
determining overdiagnosis exist and further work is planned
on this topic.
A positive result of suspected lung cancer was defined as
a nodule 4 mm, or other findings potentially related to
lung cancer. The average percentage of positive screens was
high: 24.2% of LDCTs and 6.9% of CXRs. The chance for a
participant overall after three screens to have one positive
result was 39.1% in the LDCT arm and 16.0% in the CXR
arm. Other significant abnormalities were also found more
frequently in the LDCT arm than the CXR arm (7.5%
compared with 2.1%, respectively). During the trial, there
were guidelines for the evaluation of a positive screen both
in the LDCT arm and the CXR arm. These were not
mandatory, as it was judged important to leave follow-up
451
BERG, ABERLE, AND WOOD
in the hands of the radiologists and physicians caring for

the individual patients, taking into account regional differences and patient-specific preferences. Fortunately, much of
the evaluation of abnormalities could be conducted noninvasively. In general, a clinical evaluation and a diagnostic CT
were performed. For a diagnosis of malignancy, invasive
procedures were performed. The number of invasive procedures per malignancy diagnosed was relatively low. Complications were few (1.4% in the LDCT arm and 1.6% in the
CXR arm). Major complications were primarily seen in
individuals with underlying lung cancer, i.e., in the LDCT
arm the rate of major complications in those with lung
cancer was 11.2% compared to 0.06% in those without.1
Reader variability studies done in the NLST reported that
radiologists have a low level of agreement in detecting
nodules and in measuring the growth of nodules.13,14 A
panel of radiologists reviewed NLST baseline studies and
reported the total number of abnormalities detected and
classified as pulmonary nodules. There was up to a two-fold
difference among radiologists. For cases classified as positive, consistency among recommendations for follow-up
was poor. A similar study was done to assess changes in
nodule morphology between two annual scans. Out of 95
nodules originally interpretend as present on both scans, 19
were judged by at least one of nine independent reviewers
not to be present initially.14
Another potential harm from an imaging test with ionizing radiation is that of radiation carcinogenesis. The medical
physics group involved with the trial was diligent in setting
image acquisition parameters to keep the dose low and to
keep image quality high.15,16 Effective doses were estimated
using volume CT dose index (CTDI) for the 97 scanners, and
a whole-body mean effective dose was calculated. This was
1.4 millisievert (mSv). This compares with the average
whole-body effective dose of 7 mSv from diagnostic CT.17 An
estimate of radiation-induced cancers in an individual
screened at 55, 56 and 57 years with the NLST LDCT
technique is 1 to 3 lung cancer deaths per 10,0000, and
breast cancers induced is 0.3 per 10,000. This compares with
30 lung cancer deaths prevented per 10,000 screened three
times. Of note, the cancers caused by radiation would occur
many years after the screen whereas deaths prevented occur
within a few years.18
A detailed assessment of cost-effectiveness is planned
utilizing data from the NLST. This will take into account not
only screening and diagnostic evaluation costs for positive
screens but those evaluations undertaken in those screened
who had other abnormalities or entered the medical system
as a consequence of the screen. A preliminary report indicated that the incremental cost per year of life gained in the
NLST was $38,000.19 Additionally, work with the Cancer
Intervention and Surveillance Network (CISNET) is ongoing. The CISNET groups will validate and improve as
needed their models using the NLST results. Other questions then that are very important when considering implementation of screening in the population, such as age at
which to start screening, other smoking intensities, as well
as other frequencies and durations of screening, will be
addressed.
Concerns for Implementation
Clearly this is a major advance for patients at risk for lung

cancer and will mean a major policy change by payers,
452
policy-makers, guideline groups, and patient advocates.

However, this enthusiasm, although deserved, must be
tempered with caution. Another view of the NLST data
reveals that it is necessary to screen 320 individuals every
12 months for three rounds for each lung cancer death
avoided. Many patients will be exposed to the emotional
and physical risks of lung-cancer screening to achieve
the desired benefit. A careful, measured approach is
important for the institution of lung cancer screening nationwide.
For the radiology community, the following considerations
apply: the screening process itself should be standardsdriven. Image-acquisition protocols must be consistent to
ensure adequate image quality at the lowest reasonable
radiation exposure17,20; this is particularly important if
computer-aided diagnosis (CAD) is incorporated into routine
nodule detection and characterization.21,22 Imagers experienced in the management of lung nodules should provide
interpretations and use consistent follow-up guidelines.23,24
Viable commercial solutions to track patients and nodules
do not currently exist but would have a major beneficial
effect on screening effectiveness and efficiencies; this critical
need should be the basis for developing partnerships between screening centers and industry to understand how
software technologies can facilitate workflow. There are
several questions that remain to be addressed by the imaging community. Among them are the following:
Interpretation guidelines. Screening interpretations in
the NLST were largely dichotomous, based on considerations of nodule size and morphology. The NELSON trial
being conducted in the Netherlands and Denmark uses a
two-tiered interpretation paradigm in which nodules falling
between certain size thresholds are considered indeterminate, which mandate a 3-month follow-up LDCT to determine whether the screen is negative or positive.23 Using this
algorithm, the positive predictive value of LDCT was substantially improved. Although it may be argued that medical
resource utilization is not significantly different between the
two interpretation paradigms, the implications of the
screening result using the NELSON model more closely
approximate lung cancer risk in individuals with indeterminate nodules, which has significant implications for both
the individual patient and her/his provider.
Results communication. Screening centers should not
only communicate results to the one being screened and his
or her provider but have the necessary resources to follow up
individuals with indeterminate nodules while keeping the
primary provider fully informed.
Role of image analysis. Screening interpretation in the
NLST was based on visual assessment. The European
screening trials have predicated results interpretation on
quantitative nodule volumetry. The incorporation of quantitative software into the screening process will impose
modifications to workflow in imaging practice and will
probably result in the expansion of trained allied personnel
who can oversee software analysis before formal radiologist
review.
The implications of lung cancer screening in the primarycare setting are substantial. Primary-care providers will
need to be convinced that LDCT can be effective in reducing
lung-cancer mortality and that the benefits of LDCT screening outweigh the potential harms of radiation exposure, high
false positivity rates, and potential overdiagnosis. The in-
Fig. 1. Guidelines published by the National Comprehensive Cancer Network for the diagnostic evaluation of positive screens in which solid
or part-solid nodules are detected. Abbreviations: LDCT, low-dose helical computed tomography; PET/CT, positron emission tomography/
computed tomography. Reproduced with permission from the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines for Lung Cancer
Screening V.1.2012). 2012 National Comprehensive Cancer Network, Inc. All rights reserved. The NCCN Guidelines and illustrations herein
may not be reproduced in any form for any purpose without the express written permission of the NCCN. To view the most recent and complete
version of the NCCN Guidelines, go online to NCCN.org. NATIONAL COMPREHENSIVE CANCER NETWORK, NCCN, NCCN GUIDELINES, and all
other NCCN Content are trademarks owned by the National Comprehensive Cancer Network, Inc.
corporation of lung cancer screening adds additional complexity to a busy out-patient clinic, erodes already limited
clinical time and resources, and disrupts workflow. The
intent of screening and its importance must be communicated unambiguously to patients, screening exams must be
scheduled, patients with significant findings on screens
must be referred for additional testing, and patients counseled about likely outcomes and their implications. Finally,
screening implementation will compete for attention with
other validated and less costly health care measures like
breast cancer or colon cancer screening, smoking cessation,
weight loss, and exercise.25,26
There are formidable challenges to the implementation of
lung cancer screening from the community perspective.
Successful implementation will require the diffusion of
screening across all socioeconomic strata. Community engagement will be an important element of implementation
in many cultural settings and requires trust and a successful
dialogue in which community members can be informed
about the health consequences of smoking, lung-cancer risk,
and the balance of benefits versus risks of early detection,

while also educating the medical profession regarding community priorities. The diffusion of screening and preventive
services is particularly important in underserved and minority populations because these communities are disproportionately adversely affected by lung cancer; they are
commonly diagnosed at advanced stages, less commonly
undergo surgical resections, and, in particular, black men
have a lower overall survival from lung cancer. If lungcancer screening is to be equitably administered to all
individuals at risk, the following barriers must be addressed: lack of awareness and low prioritization of lungcancer prevention and early detection; cultural concerns of
trust, fatalism, and stigmatization; financial constraints;
and geographical barriers to access.27,28
The NLST enrolled only patients at high risk of lung
cancer. Although it would be nave and narrow-minded to
not recognize that additional patients, outside of the NLST
criteria, may have substantial risk of lung cancer that
warrants screening, one must be very cautious in extrapo-
453
lating the NLST results to other patient populations and

recognize the unintended consequences of screening these
patients. Second, the NLST centers included only programs
with substantial experience and resources in radiology,
pulmonary medicine, thoracic surgery, and pathology, along
with a disciplined and multidisciplinary approach to nodule
management. Ninety-six percent of lung nodules found were
ultimately determined to be false positives, and 39% of
patients had at least one positive result during the study. A
highly organized and disciplined approach to management
is the only way to mitigate the potential harms caused to
patients by excessive and unnecessary testing and the
morbidity of invasive procedures. The margin between net
benefit and net harm in lung-cancer screening is likely
small, and the benefit to patients could easily be lost if a
higher percentage of the patients with false positive findings
undergo unnecessary work-up and invasive testing. Successful implementation of lung-cancer screening will require the
following: (1) pragmatic and thoughtful guidelines that
define patients eligible for screening (not limited to NLST
criteria yet reasonably narrow in scope); (2) experienced
radiologists to interpret screening studies and minimize
false positives; (3) a protocolized approach for the management of screen detected nodules; (4) diagnostic and therapeutic surgical procedures performed by board-certified
thoracic surgeons in order to optimize staging and minimize
morbidity; and (5) experienced multidisciplinary oncology
management with thoracic surgery, medical, and radiation
oncology to optimize oncology treatment and outcomes.
Guideline Development
The development of lung-cancer screening guidelines is

underway, with the first being published by the National
Comprehensive Cancer Network (NCCN) in October
2011.29 The NCCN has a strong history and experience
in the development of cancer guidelines. The group assembled a panel of 26 professionals, representing thoracic surgery, radiology, pulmonary medicine, medical oncology,
epidemiology, pathology, internal medicine, and patient advocacy, and worked together to produce the first lung cancer
screening guidelines developed after the NLST. Most notable in the NCCN guidelines is the extrapolation of high-risk
patients beyond the inclusion criteria of the NLST, to
include patients 50 to 54 years (NLST included only 55 to 74
years), and patients with 20 pack per year smoking
history (NLST required 30 pack-years) if the patient had
another lung-cancer risk factor as well (chronic obstructive
pulmonary disease, pulmonary fibrosis, radon or occupational exposure, cancer history, or family history). Another
extrapolation was to recommend that screening continue
annually until the individual reached 74 years. The NCCN
also recommended a highly protocolized approach to the
follow-up, work-up, and invasive testing of positive findings,
similar to those recommended by the Fleischner Society24
and others.
The biggest challenge in lung-cancer screening is the
thoughtful management of screen- detected nodules, the
majority of which are benign. There are several variations
on the management of screen-detected lung nodules, proposed by the Fleischner Society,24 the International Early
Lung Cancer Action Program (I-ELCAP),30 the NLST,1 the
NELSON Trial,23 and specific recommendations regarding
nonsolid nodules by Godoy and colleagues.31 The NCCN
454
Lung Screening Panel has amalgamated these recommendations into a pragmatic algorithm for nodule management
(Fig. 1 and Fig. 2).29 The NCCN recommendations are less
aggressive than the I-ELCAP for the work-up of baseline,
new solid, and part solid nodules 6 mm. The NCCN
recommendations are also slightly different in recommending a contrast enhanced CT or positron emission tomography (PET) in the evaluation of solid or part solid nodules
8 mm. Finally, the NCCN defined nodule growth as either
an increase in the mean diameter of 2 mm or more for
nodules 15 mm or in the solid portion of a part-solid
nodule, or an increase of 15% or more in the mean diameter
for nodules 15 mm. This definition of nodule growth is
simplified compared with I-ELCAP and should result in
fewer false positive results than seen in the NLST. Of note,
surveys of compliance with the Fleischner Society guidelines
have shown only 35% to 60% compliance by members of the
Radiological Society of North America,32 and 27% compliance by members of the Society of Thoracic Radiology,33 with
an overall trend toward over-management. It will be important for the successful application of screening programs to
assure an algorithmic and disciplined approach to nodule
work-up and follow-up in order to minimize the serious
potential harms from excessive and invasive testing in these
patients undergoing screening.
Once a nodule has been identified, the involvement of an
experienced thoracic surgeon will help the multidisciplinary
team refine a strategy for further work-up, including biopsy
and/or resection. The Lung Cancer Early Detection and
Prevention Clinic at the University of Washington incorporates a Nodule Board, consisting of specialists from thoracic radiology, pulmonary medicine, and thoracic surgery.
This group reviews clinical details and imaging and develops a management plan based on a treatment algorithm
and informed by the combined expertise of the involved
specialists. The NonSmall Cell Lung Cancer Panel of the
NCCN now recommends assessment and management of
presumed or proven lung cancer by board certified thoracic
surgeons who perform lung cancer surgery as a prominent
part of their practice.34 This recommendation is based on
the data that as much as 50% of lung cancer surgery in the
United States continues to be performed by general surgeons and that surgical outcomes (morbidity and mortality),
as well as oncology outcomes (correct staging, extent of
resection, and cancer survival) are better when performed
by specialists in thoracic surgery.35,36 There are multiple
potential adverse consequences of nonspecialist surgery,
which are even more profound for recipients of lung-cancer
screening: unnecessary surgery in cases where follow-up or
other diagnostic testing may have been preferred, inadequate staging before and/or during lung cancer surgery,
underutilization of minimally invasive surgery for both
diagnostic and resection procedures, and a lack of advanced
techniques (segmentectomy or sleeve resection) to minimize the extent of pulmonary resection. Specialist thoracic
surgeons, working with a multidisciplinary lung cancer
team, are best equipped to help maximize the benefit of early
detection. They are an important part of avoiding the adverse consequences of unnecessary procedures or substandard cancer outcomes that potentially could result in more
harm than good from lung-cancer screening programs applied without adherence to guidelines and necessary professional expertise.
Fig. 2. Guidelines published by the National Comprehensive Cancer Network for the diagnostic evaluation of positive screens in which
non-solid (ground glass) nodules are detected. Abbreviations: LDCT, low-dose helical computed tomography. Reproduced with permission from
the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines for Lung Cancer Screening V.1.2012). 2012 National Comprehensive
Cancer Network, Inc. All rights reserved. The NCCN Guidelines and illustrations herein may not be reproduced in any form for any purpose
without the express written permission of the NCCN. To view the most recent and complete version of the NCCN Guidelines, go online to
NCCN.org. NATIONAL COMPREHENSIVE CANCER NETWORK, NCCN, NCCN GUIDELINES, and all other NCCN Content are trademarks owned
by the National Comprehensive Cancer Network, Inc.
Opportunities
There are major opportunities to be gained in the process

of screening implementation. The NLST successfully addressed the critical endpoint of differential lung cancer
mortality, and secondary analyses will inform the costeffectiveness of LDCT in older, heavy smokers. However,
broad-scale implementation of LDCT screening is predicated
on several variables that the NLST does not directly address
and for which further research is crucial. Among these are
considerations of the optimal risk profile of those who are
screened, and how risk profiles might be used to guide
diagnostic strategies.
Morphologic features of indeterminate lung nodules on
CT have been studied as potential predictors of lung cancer. Most analyses have relied on subjective visual assessment of nodule features such as: size (diameter), consistency
(ground glass, part-solid, or solid), border definition, and
internal features, such as reticulation, air bronchograms
and bubble-like lucencies.37 Quantitative analysis of lung
nodules using CAD seeks to characterize nodules by math-
ematical feature descriptors. We are at the cusp of validating analytic software that can reproducibly characterize
lung nodules across a range of nodule types.38,39 Such
nodule characterization could become standard in the diagnostic stratification of individuals with indeterminate nodules.
Between 80% and 90% of lung cancers occur in tobacco
smokers, yet only 10% to 15% of chronic smokers develop
lung cancer. Prospective studies have also shown that approximately 25% of smokers develop COPD as defined by
spirometry, whereas 50% to 80% of patients with lung
cancer have COPD.40,41 Relative to smokers with normal
lung function, those with COPD have up to a six-fold
increased risk of lung cancer, making COPD by far the
greatest risk factor for lung cancer in ever smokers.41 These
observations suggest an inherently greater risk of lung
cancer among smokers with COPD than smokers with
normal lung function. Although COPD and lung cancer have
in common smoking exposure, several lines of evidence now
support underlying shared genetic susceptibility that acts in
455
concert with the shared risk of smoking-related genetic and

epigenetic effects. Genome-association studies have identified several heritable susceptibility or protective loci thought
to affect both COPD and lung cancer development: single
nucleotide polymorphisms on loci 15q25 that regulate cholinergic nicotine receptors (CHRNA3/5); several haplotypes
involved in the xenobiotic metabolism of tobacco lung carcinogens, and; genes involved in cell-cycle control, apoptosis,
airway inflammation, and repair.42,43
Emphysema has recently been found to be associated
with lung cancer, independent of airflow obstruction on
spirometry. Emphysema can be directly quantified on LDCT
with high reproducibility, and commercial software is also
available that can objectively quantify the severity of
smoking-related airway remodeling.44 In patients with indeterminate nodules, such characterization could factor into
diagnostic algorithms and may ultimately inform the determination of screening frequency at the individual patient
level.
Finally, the peripheral blood serves as a repository of
lung cancer-associated cytokines, soluble proteins, and
microRNAs that derive from the tumor microenvironment
and that exhibit molecular signatures similar to those
in tumor tissues.45,46 Similarly, samples of airway epithelium obtained through bronchoscopy, sputum expectoration, or nasal cellular brushings express aberrant methylation and microRNA patterns observed in lung cancers.47,48
If validated, these lung cancer-specific molecular signatures fromeasily accessible tissues will enable their
translation into clinical practice and will substantially alter how we define lung-cancer risk and screening in the
future. At 15 of the NLST centers, sponsored by the American College of Radiology Imaging Network, participants
volunteered to provide serial blood, sputum, and urine
specimens. Lung cancer and other tissue specimens were
collected across the trial and used to construct tissuemicroarrays. These specimens, when combined with the
voluminous data from the study, may be useful in enhancing
this molecular-signature research. The biospecimens are
available to the research community through a peerreviewed process.49
As we begin to more systematically define lung-cancer risk
through combinations of clinical, phenotypic, and molecular
profiling, we will be better positioned to distinguish between
individuals who have lung cancer versus no cancer. Such
discrimination can significantly lower the harms of screening by reducing unnecessary interventions, minimizing
anxiety, and lowering costs while promoting early diagnosis
and intervention. Finally, the integration of biologic and
imaging-based biomarkers to define risk provides significant
opportunity to stimulate the motivational tension to stop
smoking, which is most important in the prevention of lung
cancer and all smoking-related diseases. The goal is to bring
this epidemic of smoking-related disorders to an end.
Author
Christine D. Berg*
Denise R. Aberle*
Douglas E. Wood*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. The National Lung Screening Trial Research Team. Reduced lungcancer mortality with low-dose computed tomographic screening. N Engl
J Med. 2011;365:395-409.
2. Melamed MR, Flehinger BJ, Zaman MB, et al. Screening for early lung
cancer: Results of the Memorial Sloan-Kettering study in New York. Chest.
1984;86:44-53.
3. Tockman MS, Levin ML, Frost JK, et al. Screening and detection of lung
cancer. In Aisner J (ed). Lung Cancer. New York: Churchill Livingstone,1985;
25-26.
4. Fontana RS, Sanderson DR, Woolner LB, et al. Lung cancer screening:
The Mayo program. J Occup Med. 1986;28:746-750.
5. Kubik A, Parkin DM, Khlat M, et al. Lack of benefit from semi-annual
screening for cancer of the lung: Follow-up report of a randomized controlled
trial on a population of high-risk males in Czechoslovakia. Int J Cancer.
1990;45:26-33.
6. Marcus PM, Bergstralh EJ, Fagerstrom RM, et al. Lung cancer mortality
in the Mayo Lung Project: Impact of extended follow-up. J Natl Cancer Inst.
2000;92:1308-1316.
7. Doria-Rose VP, Marcus PM, Szabo E, et al. Randomized controlled trials
of the efficacy of lung cancer screening by sputum cytology revisited:a combined mortality analysis from the Johns Hopkins Lung Project and the
Memorial Sloan-Kettering Lung Study. Cancer. 2009;115:5007-5017.
8. Oken MM, Hocking WG, Kvale PA, et al. Screening by chest radiograph
and lung cancer mortality: The Prostate, Lung, Colorectal, and Ovarian
(PLCO) randomized trial. JAMA. 2011;306:1865-1873.
9. Naidich DP, Marshall CH, Gribbin C, et al. Low-dose CT of the lungs:
preliminary observations. Radiology. 1990;175:729-731.
10. The National Lung Screening Trial Research Team. The National
456
Lung Screening Trial: Overview and study design. Radiology. 2011;258:243253.

11. Church TR, National Lung Screening Trial Executive Committee.
Chest radiography as the comparison for spiral CT in the National Lung
Screening Trial. Acad Radiol. 2003;10:713-715.
12. The National Lung Screening Trial Research Team. Baseline characteristics of participants in the randomized National Lung Screening Trial.
13. Gierada DS, Pilgrim TK, Ford M, et al. Lung cancer: Interobserver
agreement on interprestion of pulmonary findings at low-dose CR screening.
Radiology. 2008;246:265-272.
14. Singh S, Pinsky P, Fineberg NS, et al. Evaluation of reader variability
in the interpretations of follow-up CT Scans at lung cancer screening.
Radiology. 2011;259:263-270.
15. Cagnon CH, Cody DD, McNitt-Gray MF, et al. Description and implementation of a quality control program in an imaging-based clinical trial.
Acad Radiol. 2006;13:1431-1441.
16. Gierada DS, Garg K, Nath H, et al. CT quality assurance in the
lung screening study component of the National Lung Screening Trial:
Implications for multicenter imaging trials. AJR Am J Roentgenol. 2009;193:
419-424.
17. Larke FJ, Kruger RL, Cagnon CH, et al. Estimated radiation dose
associated with low-dose chest CT of average-size participants in the National
Lung Screening Trial. AJR Am J Roentgenol. 2011;197:1165-1169.
18. Berrington de Gonzalez A, Kim KP, Berg CD. Low-dose lung computed
tomography screening before age 55: estimates of the mortality reduction
required to outweigh the radiation-induced cancer risk. J Med Screen.
2008;15:153-8.

19. Montes RP. The cost of CT screening: year gained via lung cancer
screening could cost $38,000. The Cancer Letter. 2011;37:15-16.
20. CT Technique Chart. National Lung Screening Trial-American College
of Radiology Imaging Network Protocol 6654. http://www.acrin.org/Default.
aspx?tabid282. Accessed February 19, 2011.
21. Daqing M, Linning E. Volumetric measurement pulmonary groundglass opacity nodules with multi-detector CT: Effect of various tube current
on measurement accuracya chest CT phantom study. AJR. 2009;16:934939.
22. Park CM, Goo JM, Lee HJ, et al. Persistent pure ground-glass nodules
in the lung: Interscan variability of semiautomated volume and attenuation
measurements. AJR. 2010;195:W408-414.
23. Van Klaveren RJ, Oudkerk M, Prokop M, et al. Management of
lung nodules detected by volume CT scanning. N Engl J Med. 2009;361:22212229.
24. MacMahon H, Austin JHM, Gamsu G, et al. Guidelines for management of small pulmonary nodules detected on CT scans: a statement from the
Fleischner Society. Radiol. 2005;237:395-400.
25. Raupach T, Merker J, Hasenfuss G, et al. Knowledge gaps about
smoking cessation in hospitalized patients and their doctors. Eur J Cardiovasc Prev Rehabil. 2011;18:334-341.
26. Tong EK, Strouse R, Hall J, et al.. National survey of US health
professionals smoking prevalence, cessation practices, and beliefs. Nicotine
Tobac Res. 2010;12:724-733.
27. Stuber J, Galea S, Link BG. Smoking and the emergence of a stigmatized social status. Soc Science Med. 2008;67:420-430.
28. Niederdeppe J, Levy AG. Fatalistic beliefs about cancer prevention and
three prevention behaviors. Cancer Epidemiol Biomarkers Prev. 2007;16:9981003.
29. Wood DE, Eapen GA, Ettinger DS, et al. Lung cancer screening. J Natl
Compr Canc Netw. 2012;10:24-65.
30. International Early Lung Cancer Action Program: Enrollment and
Screening Protocol. http://www.ielcap.org/professionals/docs/ielcap.pdf. Accessed March 27, 2012.
31. Godoy MC, Naidich DP. Subsolid pulmonary nodules and the spectrum
of peripheral adenocarcinomas of the lung: Recommended interim guidelines
for assessment and management. Radiology. 2009;253:606-622.
32. Eisenberg RL, Bankier AA, Boiselle PM. Compliance with Fleischner
Society guidelines for management of small lung nodules: A survey of 834
radiologists. Radiology. 2010;255:218-224.
33. Esmaili A, Munden RF, Mohammed TLH. Small pulmonary nodule
management: A survey of the members of the Society of Thoracic Radiology
with comparison to the Fleischner Society guidelines. J Thorac Imaging.
2011;26:27-31.
34. Ettinger DS, Akerley W, Bepler G, et al. Non-small cell lung cancer.
35. Wood DE, Farjah F. Surgeon Specialty is Associated with Better
Outcomes: The Facts Speak for Themselves. Ann Thorac Surg. 2009;88:13931395.
36. Farjah F, Flum DR, Varghese TK, et al. Surgeon specialty and longterm survival following resection for lung cancer. Ann Thorac Surg. 2009;87:
995-1006.
37. Travis WD, Brambilla E, Noguchi M, et al. International Association
for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol. 2011;6:244-285.
38. Iwano S, Nakamura T, Kamioka Y, et al. Computer-aided differentiation of malignant from benign solitary pulmonary nodules imaged by highresolution CT. Comput Med Imag Graph. 2008;32:416-422.
39. Way TW, Sahiner B, Chan H-P, et al. Computer-aided diagnosis of
pulmonary nodules on CT scans: Improvement of classification performance
with nodule surface features. Med Phys. 2009;36:3086-3098.
40. Lokke A, Lang P, Scharling H, et al. Developing COPD: A 25-year
follow up study of the general population. Thorax. 2006;61:935-939.
41. Young RP, Hopkins RJ, Christmas T, et al. COPD prevalence is
increased in lung cancer, independent of age, sex, and smoking history. Eur
Respir J. 2009;34:380-386.
42. Adcock IM, Caramori G, Barnes PJ. Choric obstructive pulmonary disease
and lung cancer: New molecular insights. Respiration. 2011;81:265-284.
43. Young RP, Hopkins RJ, Whittington CF, et al. Individual and cumulative effects of GWAS susceptibility loci in lung cancer: Associations after
sub-phenotyping for COPD. PLos ONE. 2011;6:e16476.
44. Hasegawa M, Nasuhara Y, Onodera Y, et al. Airflow limitations and
airway dimensions in chronic obstructive pulmonary disease. Am J Respir
Crit Care Med. 2006;173:1309-1315.
45. Punturieri A, Szabo E, Croxton T, et al. Lung cancer and chronic
obstructive pulmonary disease: Needs and opportunities for integrated research. J Natl Cancer Inst. 2009;101:554-559.
46. Taylor DD, Gercel-Taylor C. MicroRNA signatures of tumor-derived
exosomes as diagnostic biomarkers of ovarian cancer. Gynecol Oncol. 2008;
110:13-21.
47. Mitchell PS, Parkin RK, Kroh EM, et al. Circulating microRNAs as
stable blood-based markers for cancer detection. Proc Natl Acad Sci U S A.
2008;105:10513-10518.
48. Palmisano WA, Divine KK, Saccomanno G, et al. Predicting lung cancer
by detecting aberrant promoter methylation in sputum. Cancer Res. 2000;60:
5954-5958.
49. NLST-ACRIN Biorepository. http://www.acrin.org/RESEARCHERS/
POLICIES/NLSTACRINBIOREPOSITORY.aspx. Accessed March 28, 2012.
457
PERSONALIZED MEDICINE IN LUNG

CANCER IN 2012
CHAIR
Boston, MA
SPEAKERS
Ignacio I. Wistuba, MD
Houston, TX
Tetsuya Mitsudomi, MD, PhD
Aichi Cancer Center Hospital
Nagoya, Japan
Molecular Testing of NonSmall Cell Lung

Carcinoma Biopsy and Cytology Specimens
By Ignacio I. Wistuba, MD
Overview: During the past decade, substantial progress has

been made in the characterization of molecular abnormalities
in nonsmall cell carcinoma (NSCLC) tumors that are being
used as molecular targets and predictive biomarkers for
selection of targeted therapy. These recent advances in
NSCLC targeted therapy require the analysis of a panel of
molecular abnormalities in tumor specimens, including gene
mutations (e.g., EGFR, KRAS, BRAF, DDR2), gene amplifications (e.g., MET, FGFR1), and fusions (e.g., EML4-ALK) by
applying different methods to tumor tissue (biopsy) and cell
(cytology) samples. However, the biopsy and cytology samples
UNG CANCER continues to be the most common and

deadly malignant tumor worldwide.1 The main challenge to improve the poor survival rate (5-year survival of
approximately 15%) of this disease is to develop novel
strategies to better stratify high-risk populations for early
diagnosis and to select the adequate treatment for different
lung cancer subsets.2
NSCLC represents more than 80% of lung cancers.3 Adenocarcinoma (40%) and squamous cell carcinoma (30%) are
the most frequent histologic subtypes, but there are also
less frequent types, including large cell, adenosquamous,
and sarcomatoid carcinomas. Although most NSCLCs are
associated with smoking, some of them (approximately
15%), mostly adenocarcinoma, also occur in never-smoker
patients.
Lung tumors are the result of a multistep process in
which normal lung cells accumulate multiple genetic and
epigenetic abnormalities and evolve into cells with malignant biological capabilities.4 Recent advances in understanding the complex biology of NSCLC, particularly the
activation of oncogenes by mutation, translocation, and
amplification, have provided new treatment targets
and allowed the identification of subsets of tumors with
unique molecular profiles that can predict response to
therapy in this disease.5 The identification of specific genetic
and molecular abnormalities in tumor tissue specimens and
the administration of specific inhibitors to those targets are
the basis of personalized cancer treatment.5
The successful development of personalized therapy depends on the identification of a specific molecular target that
drives cancer growth, subsequent validation of a clinically
applicable biomarker, and development of a clinically sound
and rational endpoint, coupled with understanding of the
molecular mechanisms associated with the tumors resistance. In this process, the role of the pathologist in the
analysis of molecular changes in lung cancer tumor tissue
specimens is becoming increasingly important. These
changes in the paradigms of lung cancer diagnosis and
treatment have posed multiple new challenges for pathologists to adequately integrate both routine histopathologic
assessment and molecular testing into the clinical pathology
for tumor diagnosis and subsequent selection of the most
appropriate therapy.
available for molecular testing in advanced metastatic NSCLC

tumors are likely to be small specimens, including core needle
biopsies and/or fine needle aspiration, which may limit the
molecular and genomic analysis with currently available methods and technologies. In this process, the role of the pathologist is becoming increasingly important to adequately
integrate both routine histopathologic assessment and molecular testing into the clinical pathology for proper tumor
diagnosis and subsequent selection of the most appropriate
therapy.
Molecular Abnormalities of NSCLC
During the past decade, substantial progress has been

made in the characterization of molecular abnormalities in
NSCLC tumors that are being used as molecular targets and
predictive biomarkers for selection of targeted therapy (Table 1). In lung adenocarcinoma, at least two different major
pathways have been identified in its pathogenesis: a
smoking-associated activation of KRAS signaling and a
nonsmoking-associated activation of EGFR signaling.6
Lung adenocarcinomas arising in never or light smokers are
characterized by markedly higher frequencies of a series of
targetable oncogenes abnormalities,5 including EGFR and
HER2 tyrosine kinase (TK) domainactivating mutations6
and EML4-ALK (2;5)(p23q35) translocation.7 Recently, an
additional potentially targetable gene translocation, KIF5BRET (10p;11q)(p11.22; q11-21), has been identified in lung
adenocarcinoma from never and ever smokers.8-11
Squamous cell carcinoma of the lung has been less histologically and molecularly studied than adenocarcinoma.
Squamous cell carcinoma also harbors genetic abnormalities, resulting in activation of oncogenes, including EGFRvIII (deletion of exons 27) and DDR2 mutations and FGFR1
(8p12) gene amplification (Table 1).12,13 Other potentially
targetable genetic abnormalities have been detected in both
major NSCLC histologic subtypes, including, among others,
PIK3CA mutation and amplification, MET amplification
(7q21-q31), and AKT1 and MAP2K1 mutations.5,14
The most frequent clinically relevant driver gene abnormalities that define new molecular subsets of NSCLC are
reviewed below.
EGFR Mutation
Epidermal growth factor receptor (EGFR) molecular abnormalities are common events in NSCLC and include
gene-activating mutations, gene amplification, and overex-
From the Departments of Pathology and Thoracic/Head and Neck Medical Oncology,
University of Texas M. D. Anderson Cancer Center, Houston, TX.
Address reprint requests to Ignacio I. Wistuba, MD, Department of Pathology and
Thoracic/Head and Neck Medical Oncology, Unit 85, University of Texas M. D. Anderson
Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030; email: iiwistuba@mdanderson.org.
1092-9118/10/1-10
459
IGNACIO I. WISTUBA
Table 1. Summary of Molecular Abnormalities Associated
with the Lung Adenocarcinoma and Squamous Cell
Carcinoma Histologies
Gene
EGFR
HER2
EML4-ALK
KIF5B-RET
KRAS
BRAF
FGFR1
DDR2
PIK3CA
Molecular Change
Adenocarcinoma
Squamous cell
Carcinoma
Mutation
Amplification/CNG
IHC overexpression
Mutation
Amplification
Translocation
Translocation
Mutation
Mutation
Amplification
Mutation
Amplification/CNG
Mutation
1040%
15%
1540%
2%
4%
7%
2%
1030%
13%
Not reported
Not reported
26%
2%
Very rare
30%
60%
Very rare
2%
Very rare
Not reported
Very rare
Very rare
20%
4%
30%
2%
Abbreviations: CNG, copy number gain; IHC, immunohistochemistry.
pression of the protein and its ligands.4 Mutations of EGFR

occur in approximately 24% of adenocarcinomas and up to
60% in tumors from never smokers. The mutations are
limited to the first four exons of the TK domain (exons 18 to
21).15,16 The most frequent mutations are in-frame deletions
in exon 19 (44% of all mutations) and missense mutations in
exon 21 (41% of all mutations).17 In addition, in-frame
duplications or insertions occurring in exon 20 have been
described in approximately 5% of the mutant cases, and rare
missense mutations occur in multiple sites.17 EGFR mutations occur predominantly in adenocarcinoma (approximately 20% to 48% vs. approximately 2% for other NSCLC
histologic subtypes) and are more frequent in never smokers
(54% vs. 16% in ever smokers) and female patients (49% vs.
19% in male patients).17 EGFR mutation is the most important criterion used to select patients for EGFR TK inhibitor
therapy in lung cancer.18
HER2 Mutation
In lung cancer, HER2/Neu abnormalities include gene

mutation, gene amplification, and overexpression of the
protein.19 HER2 gene mutations have been detected in 2%
KEY POINTS
460
Non-small cell lung carcinoma (NSCLC) can be molecularly classified in multiple subtypes for selection
of targeted therapy.
Currently, a panel of gene abnormalities (mutations,
amplifications, and translocations) can be tested in
NSCLC tumor biopsy and cytology specimens.
The role of the pathologist is crucial to integrate
histology diagnosis and molecular testing of small
tumor samples.
Quality control of tumor tissue and cell specimens for
adequacy is extremely important for successful molecular testing.
Novel methodologies, including multiplex mutation
detection platforms and next-generation sequencing,
are useful to test multiple genetic aberrations in
small tumor specimens.
of adenocarcinomas and occur mostly in exon 20.19 HER2

mutations have been described predominantly in patients
with lung cancer with East-Asian ethnic background and a
history of never smoking. HER2 amplification has been
reported in 2 to 4% of NSCLCs and is more frequent in
adenocarcinoma (4%).20
ALK Fusion Genes
In lung cancer, aberrant ALK expression has been identified in a subset of adenocarcinomas, and this abnormality
consists of the formation of a fusion transcript with celltransforming activity, which is the product of a inverted
translocation of EML4 gene located at chromosome 2p21 and
the ALK gene located at 2p23.7 EML4-ALK translocations
have multiple distinct isoforms (up to 9) with demonstrated
transforming activity. EML4-ALK translocation has been
detected in 7% of lung adenocarcinomas, particularly in
patients with a history of never or light smoking, and is
associated with early onset of tumor.21 Histologically,
EML4-ALKrearranged adenocarcinomas have been described to have a predominantly solid pattern with signet
ring cells, but combined acinar and cribriform patterns
also have been described in these tumors.21 The standard
method to assess EML4-ALK fusion in lung cancer tumors is
fluorescence in situ hybridization (FISH) using a breakapart probe, and samples are considered to have positive
FISH results for EML4-ALK fusion if more than 15% of
scored tumor cells have split ALK 5 and 3 probe signals or
have isolated 3 signals.22 There are some reports that
suggest that ALK protein expression assessment by immunohistochemistry (IHC) correlates with the presence of
EML4-ALK fusion, and there are ongoing studies testing
ALK protein expression as a screening method for ALK
fusions.23
BRAF and KRAS Mutations
BRAF oncogene can be activated in NSCLC, particularly

adenocarcinoma (1% to 3%), by gene point mutations.5
Contrary to melanoma, most BRAF mutations detected in
lung cancer are non-Val600Glu mutations that affect exons
11 and 15, and they are mutually exclusive to EGFR and
KRAS mutations.
KRAS mutations are more common in lung adenocarcinoma than other NSCLC histologic types and are more
frequently found in tumors from patients with a smoking
history (approximately 30%).6 In lung cancer, KRAS mutations are found in codons 12, 13, and 61 (42% of all mutations), which are mainly GGT to TGT transversions that
produce glycine to cysteine amino acid changes.6 KRAS
mutations are rarely detected in EGFR-mutant tumors.
RAS is considered of untargetable molecule; therefore, recent studies have evaluated the RAS downstream pathway,
RAS/RAF/MEK, as a potential target for therapy in lung
cancer.
FGFR1 Amplification
FGFR1 is a transmembrane TK and member of the

fibroblast growth factor receptor (FGFR) TK family that
comprises four kinases (FGFR-1 to -4).13 In lung cancer,
amplification of FGFR1 (chromosome 8p11-12) is a driver
event in NSCLC and is predominantly detected in squamous
cell carcinomas (approximately 20%) compared with adenocarcinomas (1% to 3%).13 Currently, the preferred method to
MOLECULAR TESTING AND NSCLC

Fig 1. Left, Photomicrographs of representative examples of core needle biopsy (CNB) and fine needle
aspiration (FNA) specimens frequently available for histologic diagnosis of advanced metastatic non-small cell
lung cancer (NSCLC). The insert in the CNB photomicrograph corresponds to a tissue paraffin block. Right, In
NSCLC, diagnosis of the histologic subtype (adenocarcinoma or squamous cell carcinoma) is the first step. In
tumors with poorly differentiated histologic subtypes,
the diagnosis of NSCLC not otherwise specified is frequently performed; however, a more specific histologic
diagnosis should be reached by using a limited immunohistochemistry panel: thyroid transcription factor1 is
marker of adenocarcinoma, and p63 is a marker of
squamous cell carcinoma. After assessment of tissue
quality for molecular testing, the sample should be submitted for a panel of tests, including gene fusions and
amplification analyses by fluorescent in situ hybridization and DNA extraction for gene mutation analysis. It
has been recommended that all lung adenocarcinomas
be tested for EML4-ALK fusion and EGFR mutation, while
squamous cell carcinomas should be tested for other
gene abnormalities (DDR2 mutation and FGFR1 amplification). However, the utilization of multiplex platforms
to test mutations in tumor samples allows testing of all
NSCLC histologies for a panel of mutations and other
gene abnormalities regardless of their histology.
Biopsy or
Cytology
NSCLC Small Biopsy

and Cytology
asses FGFR1 copy number is FISH, but the definitions of

copy number gain and gene amplification still must be
determined.
DDR2 Mutations
Mutations of this TK have been described in 4% of lung

squamous cell carcinomas.12 Mutations were found in both
the kinase domain and other regions of the protein sequence
without hotspots, which makes the analysis of mutations
of this gene challenging. Tumors established from a DDR2mutant cell line were sensitive to dasatinib, and a patient
with squamous cell carcinoma that responded to dasatinib
and erlotinib treatment harbored a DDR2 kinase domain
mutation.12
PIK3CA Mutations and Amplification
In NSCLC, copy number gain ( 3 copies per cell) of

PIK3CA is a common abnormality, predominantly in squamous cell carcinomas (33% to 35%) compared with adenocarcinomas (2% to 6%).14 Mutations in the helical or kinase
domain of PIK3CA have been reported in very low frequencies (2%) in NSCLC.14 PI3K and its downstream effectors
PTEN, mTOR, and AKT are potential therapeutic targets
for NSCLC therapy and are being evaluated in clinical trials
for lung cancer.
Molecular Testing of NSCLC Tissue and
Cell Specimens
The recent advances in NSCLC targeted therapy require

the analysis of a panel of molecular abnormalities in tumor
specimens (Fig. 1), including gene mutations, amplifications,
and fusions, by applying different methods to tumor tissue
specimens.24 However, the tissue (biopsy) and cell (cytology)
samples available for molecular testing in advanced metastatic tumors are likely to be small specimens, including core
needle biopsies (CNB) and/or fine needle aspiration (FNA)
(Fig. 1), which may limit molecular and genomic analysis
with currently available methods and technologies. There is
a need to adapt and incorporate the current and new
NSCLC NOS
CNB
Adenocarcinoma
FNA
IHC:
TTF-1
p63
Squamous Cell Ca
FISH:
EML4-ALK Fusion
KIF5B-RET Fusion
FGFR1 Amp
MET Amp
Mutation:
EGFR
DDR2
PI3KCA
BRAF
HER2
K-, N-, H-RAS
MEK
AKT1, etc
emerging technologies to the molecular analysis of small

tissue specimens, such as CNB and FNA, obtained from
NSCLC patients.
There are several scientific and methodologic challenges,
as well as practical barriers, to the use of widespread
molecular testing using lung biopsy and cytology specimens.
The ideal specimens for molecular testing would be freshly
obtained tumor tissues followed by immediate snap freezing.
However, these samples are usually available only for research purposes in academic centers and are typically used
for discovery purposes.2
In pathology laboratories, the diagnostic clinical tumor
tissue specimens (e.g., CNB, bronchoscopy samples, surgical
resections) are fixed in formalin and embedded in paraffin
for the histologic process. Both formalin fixation and paraffin embedding compromise the integrity of protein and
nuclei acids (RNA, DNA) for molecular testing, particularly
when nonbuffered formalin is used and the specimens are
fixed in formalin for greater than 24 hours. The cytology
specimens (e.g., bronchial brushes, bronchoalveolar lavages,
pleural fluids, and FNAs) are usually fixed in alcohol, which
is optimal for preservation of nucleic acids. When the cytology specimen has abundant material, the sample can be
fixed in formalin and processed as a tissue specimen (cell
block) to obtain histologic sections. Although tissue specimens are preferable for molecular testing, cytology samples
with abundant malignant cells can be successfully used for
molecular testing.
Among others, there are two important aspects that must
be addressed when small lung cancer biopsy or cytology
specimens are received for diagnosis in a pathology laboratory. First, the handling of the biopsy and cytology specimen
for histologic analysis and subsequent molecular testing
requires thoughtful prioritization of the utilization of the
sample to prevent the use of tissue in less important analysis than the definitive molecular testing required for selection of therapy. Second, the pathologist should determine
whether the amount of malignant cells available in the
specimen is adequate for nuclei acid extractions and also for
461
IGNACIO I. WISTUBA
histologic section based molecular tests (e.g., FISH and

IHC).
On the other hand, our growing understanding of the
cancer biology of NSCLC, particularly the molecular evolution of tumors during local progression and metastasis and
the identification of molecular abnormalities contributing to
resistance to TK inhibitor therapies, emphasizes the importance of characterizing the molecular abnormalities of the
disease at every stage of its evolution. For molecular testing
of advanced metastatic NSCLC, it is important to sample
and analyze the tumors sample at each time point of clinical
decision making.
Relevant Molecular Pathologic Analysis Methods
Currently, most of the predictive molecular markers available for therapy selection in NSCLC are oncogene mutations
and amplifications. However, other molecular and genetic
changes modulate the sensitivity of tumor to targeted therapies, including protein overexpression, gene methylation,
and gene expression abnormalities. The need for analysis of
multiple molecular and genetic changes in small, clinically
relevant biopsy and cytology specimens has driven the
scientific community and the molecular pathology laboratories to develop multiplex approaches for molecular testing
of small tumors samples, particularly for gene mutation
analysis.
Mutation Analysis
Currently, direct nucleic acid sequencing with previous

polymerase chain reaction (PCR) amplification of extracted
DNA is the most commonly used technique for gene mutation analysis of tumor biopsy and cytology samples to detect
mutations of clinically relevant genes. Several sequencing
methods are available for mutation analysis of DNA extracted from tumor tissue and cell specimens, especially
for formalin-fixed paraffin-embedded (FFPE) samples. The
current PCR-based sequencing mutation analysis methods
can be divided into uniplex (e.g., Sanger sequencing and
pyrosequencing) and multiplex (e.g., matrix-assisted laser
desorption ionization time-of-light mass spectrometry and
primer extension assay) methods. In the uniplex method,
one hotspot sequence is examined at a time, while the
multiplex technique multiple hotspot mutations are examined simultaneously. The multiplex approaches are clearly
the preferred methods since they allow more efficient mutation analysis of small amounts of DNA for multiple hotspots
(approximately 100 to 200) from a panel of genes (approximately 10 to 30).
Sanger Sequencing
In solid tumors, including lung, Sanger sequencing is the

most commonly used sequencing method to detect hotspot
mutations of oncogenes (e.g., EGFR, KRAS, BRAF). This
type of sequencing can detect essentially all base substitutions, small insertions, and deletions. Its main disadvantage
is the relatively low sensitivity (approximately 20%) for the
detection of mutant alleles in the DNA sample extracted
from tumor specimens. These DNA samples are usually a
mixture of mutant and wild-type alleles as a result of the
presence of malignant and nonmalignant (from adjacent
normal or tumor stroma) cells in the tumor tissue specimens.25
462
Pyrosequencing
Pyrosequencing uses sequencing by a synthesis method

to sequence nucleic acids and relies on the detection of
pyrophosphate release on nucleotide incorporation. Pyrosequencing is considered to be more sensitive than Sanger
sequencing and detects approximately 5% of mutant compared with wild-type alleles.26
Matrix-Assisted Laser Desorption Ionization Time-of-Light
Mass Spectrometry
Matrix-assisted laser desorption ionization time-of-light

mass spectrometry (Sequenom) utilizes a high-throughput
PCR-based sequencing assay to detect multiple hotspot
mutations simultaneously using small amounts of DNA
obtained from biopsy and cytology specimens.27 It has a high
level of sensitivity (approximately 5% of the mutant alleles)
and allows quantification of the percentage of mutant
DNA.27 The Sequenom method is also useful to assess gene
amplification.
Primer Extension Assay
Primer extension assay is a primer extension based

method that allows simultaneous analysis of up to 10 different mutations.28 It is a sensitive, low-cost, and rapid method
to screen for mutations and to analyze methylation. This
assay uses the SNaPshot Multiplex Kit (Applied Biosystems
Inc), which contains a reaction mix of four differentially
fluorescently labeled ddNTPs, allowing the interrogation of
each base at a mutation site.28
Translocation and Gene Copy Number Analyses
FISH is a cytogenetic technique that uses fluorescentlabeled probes to hybridize specific sequences of DNA on
chromosomes.29 It is applied to visualize chromosome deletions, amplification, and structural rearrangements. The
main advantage of this technique is that it allows in situ localization of the specific sequences and the simultaneous
detection of multiple sites by using hybridization probes
labeled with different fluorophores. The main disadvantage
of FISH is the need of additional equipment for analysis,
such as dark-field microscopy and multiband fluorescent
filters. Chromogenic in situ hybridization is a variant of the
in situ hybridization technique that visualizes the specific
DNA sequence by a peroxidase reaction and allows the
visualization in a light microscope. The main disadvantage
of this method is that it limits the use of different label
probes to target multiple sites; however, recent advances in
methodologies permit the use of dual colors to target two
sequencing sites and enable the visualization of tissue
architecture and cytomorphologic analysis. Although there
are other methods available for gene copy number and
fusion genes analyses, such as DNA quantitative PCR assay
for copy number assessment and messenger RNA quantitative PCR assay for gene fusion analysis, FISH continues to
be the preferred method for gene copy number (e.g., MET,
FGFR1) and gene fusion (e.g., EML4-ALK) in lung cancer.
Protein Expression Analysis
IHC is a widely used technique in pathology laboratories

to detect the presence and levels of expression of a specific
protein in FFPE tumor cytology and cytology specimens. In
lung cancer, the use of IHC is currently used for histopatho-
MOLECULAR TESTING AND NSCLC
logic diagnosis and classification of tumors, particularly

when small tissue specimens are examined. Currently, IHC
markers are frequently used by pathologists to clinically
subtype NSCLC; for example, cytokeratin 7 and thyroid
transcription factor 1 are positive in most adenocarcinomas,
whereas p63, p40, and cytokeratin 5/6 are positive in most
squamous cell carcinomas.24 Despite being readily available,
there are no validated molecular markers based on protein
expression by IHC being used to predict response to therapy
in lung cancer. One of the most important advantaged of
IHC is that it allows the identification of the protein expression in specific types of cells as well as distinct subcellular
localization.
Conclusion
Most of the current biomarkers discovered and now used

in clinical applications to date consist of a single genetic
mutation, gene amplification, or translocation, but these
aberrations are rare and not sufficient to select the majority
of patients for targeted therapies. It is also known that, in
many cases, multiple changes in tumor cells, rather than a
single modification, lead to activation of selective and often
interactive molecular pathways promoting tumor growth
and survival. In addition, various targeted treatment regimens have been shown to result in the activation of alternative, compensatory molecular pathways that continue to
promote cancer cell survival.
The development of new technologies, such as highthroughput arrays, has allowed researchers to screen the
whole genome, proteome, and transcriptome for new biomarkers in tumor tissue, serum, plasma, or other human
body fluids and develop genomic and proteomic profiles, or
signatures, to better reflect the complex molecular aberrations within a single tumor.
The rapid development of technologies for large-scale
sequencing (next-generation sequencing [NGS]) of DNA

and RNA has facilitated high-throughput molecular analysis, holding various advantages over traditional sequencing. These new technologies provide capabilities to fully
sequence large numbers of genes in a single test and
simultaneously detect deletions, insertions, copy number
alterations, translocations, and exome-wide base substitutions (including known hotspot mutations) in all known
cancer-related genes. The amount of starting material (DNA
or RNA) needed for the newest NGS applications is getting
smaller, and currently, the analysis of a panel (approximately 200 to 300) of gene mutations and fusions can be
performed even in DNA extracted from FFPE tumor tissue
specimens. However, one of the potential difficulties in this
process is the large computing capacities needed to manage
the billions of small sequence readouts generated and to
assemble those with large databases to interpret the raw
data. Another challenge for the NGS is the identification of
meaningful driver mutations and the separation of true
mutations among a background of intrinsic sequence variations.30 In addition, verifying and validating the discovered
driver mutations will require experimental and detailed
classic molecular pathology studies to bring NGS into clinical context.
In summary, the recent advances in NSCLC targeted
therapy require the analysis of a panel of molecular abnormalities of tumor specimens, including gene mutations,
amplifications, and fusions, by applying different methods
to the samples. In this new era of personalized therapy
in lung cancer using targeted agents, the role of the
pathologist in the analysis of molecular changes in lung
cancer tumor tissue specimens is becoming increasingly
important to properly integrate routine histopathologic diagnosis and molecular testing into the clinical practice.
Author
Ignacio I. Wistuba*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
2010;60:277-300.
2. Wistuba II, Gelovani JG, Jacoby JJ, et al. Methodological and practical
challenges for personalized cancer therapies. Nat Rev Clin Oncol. 2011;8:135141.
3. Travis WD, Brambilla E, Muller-Hermelink HK, et al. Tumours of
the lung. In: Travis WD, Brambilla E, Muller-Hermelink HK, et al
(eds). Pathology and Genetics: Tumours of the Lung, Pleura, Thymus
and Heart. World Health Organization Classification of Tumours. Pathology
& Genetics. Lyon, France: International Agency for Research on Cancer;
2004:9-124.
4. Herbst RS, Heymach JV, Lippman SM. Lung cancer. N Engl J Med.
2008;359:1367-1380.
5. Pao W, Girard N. New driver mutations in non-small cell lung cancer.
Lancet Oncol. 2011;12:175-180.
6. Sun S, Schiller JH, Gazdar AF. Lung cancer in never smokersa
different disease. Nat Rev Cancer. 2007;7:778-790.
7. Soda M, Choi YL, Enomoto M, et al. Identification of the transforming
EML4-ALK fusion gene in non-small-cell lung cancer. Nature. 2007;448:561566.
8. Ju YS, Lee WC, Shin JY, et al. A transforming KIF5B and RET gene
fusion in lung adenocarcinoma revealed from whole-genome and transcriptome sequencing. Genome Res. 2012;22:436-445.
9. Lipson D, Capelletti M, Yelensky R, et al. Identification of new ALK and
RET gene fusions from colorectal and lung cancer biopsies. Nat Med. Epub
2012 Feb 12.
10. Kohno T, Ichikawa H, Totoki Y, et al. KIF5B-RET fusions in lung
adenocarcinoma. Nat Med. Epub 2012 Feb 12.
11. Takeuchi K, Soda M, Togashi Y, et al. RET, ROS1 and ALK fusions in
lung cancer. Nat Med. Epub 2012 Feb 12.
12. Hammerman PS, Sos ML, Ramos AH, et al. Mutations in the DDR2
kinase gene identify a novel therapeutic target in squamous cell lung cancer.
Cancer Discov. 2011;1:78-89.
13. Weiss J, Sos ML, Seidel D, et al. Frequent and focal FGFR1 amplification associates with therapeutically tractable FGFR1 dependency in squamous cell lung cancer. Sci Transl Med. 2010;2:62ra93.
14. Yamamoto H, Shigematsu H, Nomura M, et al. PIK3CA mutations
and copy number gains in human lung cancers. Cancer Res. 2008;68:69136921.
15. Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer:
463
IGNACIO I. WISTUBA
correlation with clinical response to gefitinib therapy. Science. 2004;304:14971500.
16. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the
epidermal growth factor receptor underlying responsiveness of non-small-cell
lung cancer to gefitinib. N Engl J Med. 2004;350:2129-2139.
17. Shigematsu H, Gazdar AF. Somatic mutations of epidermal growth
factor receptor signaling pathway in lung cancers. Int J Cancer. 2006;118:
257-262.
18. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatinpaclitaxel in pulmonary adenocarcinoma. N Engl J Med. 2009;361:947-957.
19. Shigematsu H, Takahashi T, Nomura M, et al. Somatic mutations of
the HER2 kinase domain in lung adenocarcinomas. Cancer Res. 2005;65:
1642-1646.
20. Hirsch FR, Varella-Garcia M, Franklin WA, et al. Evaluation of
HER-2/neu gene amplification and protein expression in non-small cell lung
carcinomas. Br J Cancer. 2002;86:1449-1456.
21. Inamura K, Takeuchi K, Togashi Y, et al. EML4-ALK lung cancers are
characterized by rare other mutations, a TTF-1 cell lineage, an acinar
histology, and young onset. Mod Pathol. 2009;22:508-515.
inhibition in non-small-cell lung cancer. N Engl J Med. 2010;363:16931703.
23. Mino-Kenudson M, Chirieac LR, Law K, et al. A novel, highly sensitive
464
antibody allows for the routine detection of ALK-rearranged lung adenocarcinomas by standard immunohistochemistry. Clin Cancer Res. 2010;16:15611571.
24. Kerr KM. Personalized medicine for lung cancer: new challenges for
pathology. Histopathology. 2012;60:531-546.
25. Anderson SM. Laboratory methods for KRAS mutation analysis. Expert
Rev Mol Diagn. 2011;11:635-642.
26. Tsiatis AC, Norris-Kirby A, Rich RG, et al. Comparison of Sanger
sequencing, pyrosequencing, and melting curve analysis for the detection of
KRAS mutations: diagnostic and clinical implications. J Mol Diagn. 2010;12:
425-432.
27. Fumagalli D, Gavin PG, Taniyama Y, et al. A rapid, sensitive, reproducible and cost-effective method for mutation profiling of colon cancer and
metastatic lymph nodes. BMC Cancer. 2010;10:101.
28. Su Z, Dias-Santagata D, Duke M, et al. A platform for rapid detection
of multiple oncogenic mutations with relevance to targeted therapy in
non-small-cell lung cancer. J Mol Diagn. 2011;13:74-84.
29. Varella-Garcia M. Chromosomal and genomic changes in lung cancer.
Cell Adh Migr. 2010;4:100-106.
30. Cronin M, Ross JS. Comprehensive next-generation cancer genome
sequencing in the era of targeted therapy and personalized oncology. Biomark
Med. 2011;5:293-305.
THYMOMA AND THYMIC CARCINOMA:

UPDATE ON MANAGEMENT
CHAIR
Gregory J. Riely, MD, PhD
New York, NY
SPEAKERS
Nicolas Girard, MD, MSc
Hopital Louis Pradel
Bron, France
Frank C. Detterbeck, MD
New Haven, CT
Multidisciplinary Management of
Thymic Carcinoma
By Gregory J. Riely, MD, PhD, James Huang, MD, and Andreas Rimner, MD, PhD
Overview: Thymic carcinomas represent approximately 10%

of thymic tumors. In our approach to patients with thymic
carcinoma, we emphasize multimodality treatment with close
communication between the pathologist, thoracic surgeon,
medical oncologist, and radiation oncologist. Given the paucity of high-quality clinical research data, treatment decisions
are guided by a small amount of prospective trial data,
retrospective reports, and clinical experience. Surgical management of thymic carcinoma must account for the more
aggressive biology, higher degree of local invasion of neighboring structures, greater propensity for nodal metastasis,
and higher risk of distant metastatic disease. Although surgi-
HYMIC CARCINOMAS represent approximately 10%

of thymic tumors, with thymoma and thymic carcinoids
encompassing the remaining 90%. In our approach to patients with thymic carcinoma, we emphasize multimodality
treatment with close communication between the pathologist, thoracic surgeon, medical oncologist, and radiation
oncologist. Given the paucity of data with which to evaluate
approaches to therapy, treatment decisions are guided by
a small amount of prospective trial data, retrospective
reports, and institutional experience. For patients who present with localized or locally advanced disease for which
surgical resection is feasible, we routinely recommend preoperative chemotherapy with regimens evaluated in the
advanced disease setting followed by complete surgical resection with RT dependent on the findings at the time of
resection. For patients with unresectable localized disease,
we use an approach combining chemotherapy and RT. Patients who have evidence of distant metastatic disease or
recurrent disease after prior surgery or RT, palliative chemotherapy is the mainstay of treatment.
Pathology
Thymic carcinomas are malignant epithelial tumors with

overt cytologic atypia. Thymic carcinoma cells are thought
to be derived from thymic epithelial cells. Although they
may have different histologic features (similar to neuroendocrine, mucoepidermoid, and lymphoepithelioma), the most
common is squamous differentiation with large polyhedral
cells, sometimes with keratinization. Thymic carcinomas
are routinely found to be immunoreactive to antibodies to
CD5 and CD117 (c-Kit). Given the squamous histology,
histology review alone cannot distinguish squamous cell
carcinoma of the lung from thymic carcinoma, so clinical
correlation is required for this diagnosis. Although not
clinically used, copy number data show different patterns
and frequencies of chromosomal gains and losses for thymic
carcinoma tumors as compared with squamous cell carcinomas of the lung, emphasizing that they are different disease
entities.1 Of note, it is rare for patients with thymic carcinoma to have associated paraneoplastic syndromes such as
myasthenia gravis and pure red cell aplasia, which are
relatively common in patients with thymoma.
To better understand thymic carcinoma and its association with thymomas, a number of groups have evaluated
466
cal resection remains the most important component in the

management of localized thymic tumors, radiation therapy
(RT) may be used as adjuvant therapy after surgical resection
or as the definitive treatment modality in patients who are
deemed unresectable because of medical comorbidities or
technical reasons. Systemic therapy for thymic carcinoma is
used in two clinical scenarios: preoperative treatment and
palliative therapy. First-line, platinum-based chemotherapy
regimens are associated with response rates between 22%
and 75%. Recent data from targeted therapy trials do not
reveal a clear role for targeted therapies for patients with
thymic carcinoma.
specific molecular features of thymic carcinoma. Further

characterization of thymic carcinomas has found that the
epidermal growth factor receptor is expressed at high levels
in virtually all thymic carcinomas (when detected by immunohistochemistry).1 The insulin-like growth factor receptor
(IGF1-R) is also expressed at high levels in most, but not
all, thymic carcinomas.2 With targeted mutational analyses,
KIT mutations have been identified as well as a limited
number of KRAS mutations).1 Unfortunately, there are no
chromosomal gains or losses that have been uniformly
identified in all thymic carcinomas.3
Surgical Management of Thymic Carcinoma
Surgery is considered the mainstay of treatment for most

thymic tumors. However, thymic carcinoma presents challenges for management given its more aggressive behavior.
The literature regarding the surgical treatment of thymic
carcinoma is sparse because of its rarity and is limited to
small retrospective case series with heterogeneous treatments. Furthermore, many series have grouped thymic
carcinoma with thymoma in their published experiences, not
surprisingly since thymic carcinoma was labeled as type C
thymoma by the World Health Organization up until 2004,
when they recognized thymic carcinoma as a separate distinct entity. Similarly, some series of thymic carcinomas
have also included thymic neuroendocrine tumors, such as
carcinoids.
Surgical considerations must account for the more aggressive biology exhibited by thymic carcinoma, which is manifested by a higher degree of local invasion of neighboring
structures, greater propensity for nodal metastasis, and
higher risk of distant metastatic disease. Taken together,
these factors result in a lower resectability rate, higher
From the Thoracic Oncology Service, Division of Solid Tumor Oncology, Department of
Medicine, Memorial Sloan-Kettering Cancer Center, Weill Cornell Medical College, New
York, NY; Thoracic Surgery Service, Department of Surgery, Memorial Sloan-Kettering
Cancer Center New York, NY; Department of Radiation Oncology, Memorial SloanKettering Cancer Center, New York, NY.
Address reprint requests to Gregory J. Riely, MD, PhD, Department of Medicine,
Memorial Sloan-Kettering Cancer Center, Weill Cornell Medical College, Room 1261, 300
East 66th St., New York, NY; email: rielyg@mskcc.org.
1092-9118/10/1-10
MANAGEMENT OF THYMIC CARCINOMA
recurrence rate, and shorter survival relative to patients

with thymoma. Thymic carcinoma rarely presents at an
early stage. A small, well-encapsulated thymic carcinoma is
an uncommon occurrence and would likely only be diagnosed
after being resected. More typically, thymic carcinoma presents in a locally advanced setting, with evidence of local
invasion, adenopathy, or pleural or pericardial involvement.
In a recent series at our institution, 75% of thymic carcinomas presented as Masaoka stage III or IV.4
In patients who present with locally advanced disease,
achieving a complete resection is often difficult. Within the
anatomic confines of the mediastinum, the ability to obtain
a clear surgical margin is frequently limited by the great
vessels of the heart. Published complete resection rates vary
from 20% to 60%, and our institutional complete resection
rate of 52% is consistent with these results.4-6 Extended
resections involving resection of the innominate vein or
superior vena cava can be done in appropriate settings. In
one series, one-third of the patients required partial or
complete resection of the vena cava.7 Circumferential resection of the vena cava also necessitates resection of the
phrenic nerve, and careful assessment of pulmonary function in the preoperative planning is mandatory, with consideration given to diaphragm plication at the time of surgery
if indicated. It is debatable whether a tumor that invades
the aorta, pulmonary trunk, or a cardiac chamber can still
be resected. Although these situations may be technically
feasible with the use of cardiopulmonary bypass, there is no
data to suggest that the outcomes would be better than the
alternative of debulking and R1 resection followed by radiotherapy for the residual disease.7 There is insufficient evidence of benefit to support the added morbidity and risk of
such measures.
There is no consensus regarding the role of mediastinal
lymph node dissection for patients with thymic carcinoma.
Patients with thymic carcinoma much more frequently
have nodal involvement than those with thymoma.8 Any
clinically involved nodes must be resected en bloc with the
specimen, and a careful search of the mediastinum should
entail the prevascular, aortopulmonary, internal mammary,
and cervical stations. Whether a systematic mediastinal
lymphadenectomyincluding subcarinal and paratracheal
stationsis warranted is unclear and must be individualized to the patient.9
Local invasion of the lung is common and can be fre-
KEY POINTS
Thymic carcinomas are an uncommon subset of thymic tumors.

There are limited prospective data regarding treatment.
The outcomes of patients with thymic carcinoma are
worse than that seen with thymoma.
In localized thymic carcinoma, complete surgical resection, if possible, should be performed.
Radiation therapy may be used as adjuvant therapy
after surgical resection or as the definitive treatment
modality in patients with localized thymic carcinoma
that is deemed unresectable.
quently managed with wedge resections or lobectomy if

necessary. Individual pleural metastases can be easily resected in some circumstances. However, pleural involvement that presents as innumerable miliary metastases, or
conversely, as bulky confluent disease will preclude complete resection. Extrapleural pneumonectomy has been advocated for stage IVa thymoma by several institutions
including our own, but the poorer prognosis for patients with
thymic carcinoma and the morbidity associated with the
procedure engenders less enthusiasm for this aggressive
approach.7,10
RT
Although surgical resection remains the most important

component in the management of all thymic tumors, RT
may be used as adjuvant therapy after surgical resection or
as the definitive treatment modality in patients who are
deemed unresectable because of medical comorbidities or
technical reasons. No prospective trials have established a
clear role for RT, but a series of retrospective studies have
demonstrated that excellent local control rates can be
achieved when surgical resection and RT are combined.
The largest study was reported by Kondo et al.8 Onehundred and eighty-six patients with thymic carcinoma at
multiple institutions had surgically excision and treatment
with adjuvant chemotherapy, RT, both, or no adjuvant
therapy. Fifty-one percent of patients had undergone a
complete resection. The most important prognostic factor
for overall survival was a complete resection. Nevertheless,
51% of patients developed a recurrence. Although RT improved the results in incompletely resected tumors, there
was no clear benefit in patients who had undergone a
complete resection.
Similarly, a retrospective study of 40 cases with long-term
follow-up (median follow-up time of 87 months for surviving
patients) found that a complete resection was the most
important prognostic factor on multivariate analysis.11 All
patients in this study were treated with either adjuvant RT
in resectable patients or definitive RT in unresectable cases.
They achieved 100% in-field local control in 16 patients who
underwent a complete resection followed by adjuvant RT
with a median dose of 50 Gy. Other significant prognostic
factors included a KPS of 70% or greater and low-grade
histology. The 5- and 10-year overall survival rates for the
whole patient group was 38% and 28%, respectively.
Hsu et al reported their experience of 26 patients with a
minimum follow-up of 40 months, all of which were treated
with adjuvant RT at a median dose of 60 Gy after a total
or subtotal resection.12 Excellent 5-year local control rates
of 92% after a complete surgical resection and adjuvant RT
were obtained. Even after an incomplete resection and
adjuvant RT, 5-year local control rates of 88% were
achieved. The 5-year overall survival in this study was 77%.
RT techniques
Most published studies to date have reported outcomes on

patients who were treated over a long period of time with
often times old two-dimensional RT techniques using opposed fields. However, the development of more conformal
RT (CRT) techniquessuch as three-dimensional CRT in
the early 1990s and more recently intensity-modulated RT
(IMRT) or proton therapy has allowed significantly im-
467
RIELY, HUANG, AND RIMNER

Table 1. Selected Prospective and Larger Retrospective Series of Patients with Thymic Carcinoma Treated with
Reference
18
19
20
21
22
23
24
25
16
Therapy
ADOC/ADOCb
Cisplatin/irinotecan
CODE
Carboplatin, paclitaxel
Carboplatin, paclitaxel
Etoposide, ifosfamide, cisplatin
Etoposide, Ifosfamide, cisplatin
Doxorubicin, cyclophosphamide, cisplatin, vincristine
Pemetrexed
Prospective or
Retrospective
Patients
R
R
R
P
P
R
P
P
P
34
9
12
23
11
4
8
8
11
RR (%)
50%
56%
42%
22%
36%
25%
25%
75%
0
Median PFS/TTP
(months)
Median OS
(months)
8
6
5
8
21
34
46
20
23
Comment
Included type B3 thymoma
19
5
Abbreviations: RR, recurrence rate; PFS, progression-free survival; TTP, time to progression; OS, overall survival; ADOC, doxorubicin, cyclophosphamide, vincristine,
cisplatin; ADOCb, doxorubicin, cyclophosphamide, vincristine, carboplatin; CODE, cisplatin, vincristine, doxorubicin, etoposide.
proved sparing of normal organs surrounding the tumor/

tumor bed. In addition, four-dimensional treatment
planning now allows a quantitative determination of tumor
motion caused by respiration at the time of simulation and
provides further accuracy and confidence in our treatment
delivery. Expert guidelines recommend that these techniques be used for thymic malignancies where available.13 It
needs to be noted that the increased conformality of these
techniques harbors the risk of underdosing the tumor, which
may lead to marginal recurrences. Therefore, a clear set of
definitions and guidelines for the simulation, treatment
planning, and delivery using CRT techniques is necessary.
Recently, the International Thymic Malignancy Interest
Group (ITMIG) has published its first set of guidelines to
increase consistency in how RT is delivered to patients with
thymic malignancies.14 These may allow a more systematic
approach to determine the best use of RT in thymic tumors.
RT dose
The optimal RT dose has not yet been determined. Most

published studies used between 40 and 70 Gy in 1.8 to 2.0 Gy
daily fractions. One study on thymic malignancies, which
included six patients with thymic carcinoma, found that an
RT dose of 50 Gy or greater was associated with improved
local control.15 However, this has not been confirmed in
other studies.12 In general, doses of 45 to 50 Gy for negative/
close margins, 50 to 54 Gy for microscopically positive
margins, and 60 Gy for gross residual disease in conventional daily fractions of 1.8 to 2.0 Gy are recommended
(National Comprehensive Cancer Network guidelines
v2.2012).
Organs at Risk
The organs at risk for short- and long-term toxicities from

RT for thymic carcinomas include the heart, lungs, esophagus, and spinal cord. More modern RT techniques, such as
three-dimensional CRT and IMRT, allow better sparing of
these organs at risk without compromising target coverage.
General dosimetric guidelines include a mean heart dose of
less than 26 Gy, mean lung dose 20 Gy or less and V20 at
30% to 35% or less, mean esophageal dose less than 34 Gy,
and Dmax to the spinal cord less than 45 Gy.14
Systemic Therapy
Systemic therapy for thymic carcinoma is used in two

clinical scenarios, preoperative treatment and palliative
468
therapy. Given the significance of a complete surgical resectionif significant tumor response occursthe probability
of complete surgical resection can be improved. In comparison with more common diseases, identifying the appropriate
chemotherapy regimen to choose for initial or subsequent
therapy for patients with thymic carcinoma is challenging.
As with surgical and RT data, prior reports of thymic tumor
management have often combined patients with thymoma
and thymic carcinoma into small studies. However, there
are relatively large retrospective series and subsets from
prospective series that provide some data to guide therapy.
All chemotherapy regimens evaluated in thymic carcinoma have been studied either as part of series evaluating
thymoma or used chemotherapy regimens developed for
thymoma. In general, the combination chemotherapy regimens most widely evaluated have combined platinum analogs, anthracyclines, along with other agents (Table 1). For
first-line, platinum-based chemotherapy regimens, the response rates have ranged between 22% and 75%; however,
the varying chemotherapy regimens, different ways data
were collected, and the small sample sizes limit the conclusions that can be drawn about individual chemotherapy
regimens. Despite our inability to differentiate individual
regimens, it seems clear that the most commonly used
treatments are cisplatin-based chemotherapy, sometimes in
combination with an anthracycline.
Despite the relatively short median overall survival times
reported, there is a frustrating lack of data to guide use of
second-line cytotoxic therapies for treatment of patient with
thymic carcinoma. The exception is pemetrexed in patients
with previously treated advanced thymic carcinoma; unfortunately, there were no radiographic responses observed,
but there was a median time to progression of 5 months.16
This summary of data underscores the need for more prospective evaluation of the cytotoxic chemotherapies often
used to treat patients with thymic carcinoma.
Targeted Systemic Therapies
The absence of high rates of radiographic response and

short progression-free survival associated with cytotoxic
chemotherapy has blunted enthusiasm for further evaluation of such drugs and more recent studies have focused on
newer targeted therapies, often with preclinical rationales
supporting evaluation of a given drug (Table 2). With two
MANAGEMENT OF THYMIC CARCINOMA

Table 2. Selected Prospective and Larger Retrospective Series of Patients with Thymic Carcinoma Treated with
Targeted Therapies
Reference
Therapy
26
27
28
29
30
31
32
33
Prospective or
Retrospective
Patients
RR (%)
P
P
P
P
P
P
P
P
11
5
3
7
7
16
5
3
0
0
0
0
0
0
0
67%
Imatinib
Imatinib
Imatinib
Erlotinib, bevacizumab
Gefitinib
Belinostat
Octreotide
PHA 848125
Median PFS/TTP
(months)
Median OS
(months)
Comment
All patients were KIT or PDGF by IHC

No patients had KIT mutations
No patients had KIT mutations
3
5
12
23
Part of a phase I trial
Abbreviations: RR, recurrence rate; PFS, progression-free survival; TTP, time to progression; OS, overall survival; PDGF, platelet-derived growth factor receptor; IHC,
immunohistochemical.
exceptions, testing of targeted therapies in thymic carcinoma has met with no significant success. Several prospective trials of imatinib were launched after initial case
reports identified mutated KIT in patients with thymic
carcinoma and subsequent significant response to multitargeted tyrosine kinase inhibitors. These studies included
patients with thymic carcinoma, sometimes requiring KIT
(or platelet-derived growth factor expression), but none
required the presence of KIT mutations for enrollment. In
the absence of patients with documented KIT mutations, no
patients had radiographic responses. Because of high epidermal growth factor receptor (EGFR) expression in thymic
carcinoma, both gefitinib and erlotinib (with bevacizumab)
have been prospectively evaluated in a total of 14 patients
without radiographic response. No evidence of EGFR dependence in thymic tumors has been reported (no EGFR mutations and no significant focal EGFR gene amplification). One
targeted therapy has shown initial success in patients with
thymic carcinoma who were enrolled in a phase I trial. Two
of three patients with thymic carcinoma treated with PHA848125ACa dual cyclin-dependent kinase 2/thropomyosin
receptor kinase A inhibitor had radiographic responses
and long responses to therapy. This initial success has led to
a multicenter phase II study of this drug in patients with

thymic carcinoma (clinicaltrials.gov NCT01011439).
Future Directions
The rarity of all thymic tumors, especially thymic carcinomas, complicates prospective clinical research in these
diseases. As noted above, at sites with expertise in this
disease, the largest series of patients prospectively evaluated was 23 and none of these studies has led to a breakthrough in therapy for this disease. Given the uncommon
nature of thymoma and thymic carcinoma, empiric approaches to drug evaluation conducted at a limited number
of centers are likely to be of little benefit. Under the
guidance of the ITMIG, researchers are developing a large
retrospective and prospective database to identify potential
improvements in care that can be derived from current
practices. In addition, ITMIG has put forth standards for
clinical research to allow comparison of clinical trial results
across groups.17 Finally, the thymic cancer research community needs to harness recent advances in sequencing technology to vastly improve our understanding of the biology
of this disease to generate testable clinical research hypotheses.
Author
Gregory J. Riely
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
ARIAD;
Boehringer
Ingelheim;
Chugai Pharma;
Daiichi Sankyo;
Novartis;
Tragara
Research
Funding
Expert
Testimony
Other
Remuneration
Bristol-Myers
Squibb
James Huang
Bristol-Myers
Squibb
Andreas Rimner*
REFERENCES
1. Girard N, Shen R, Guo T, et al. Comprehensive genomic analysis reveals
clinically relevant molecular distinctions between thymic carcinomas and
thymomas. Clin Cancer Res. 2009;15:6790-6799.
2. Girard N, Teruya-Feldstein J, Payabyab EC, et al. Insulin-like growth
factor-1 receptor expression in thymic malignancies. J Thorac Oncol. 2010;5:
1439-1446.
3. Zettl A, Strobel P, Wagner K, et al. Recurrent genetic aberrations in
thymoma and thymic carcinoma. Am J Pathol. 2000;157:257-266.
4. Huang J, Rizk NP, Travis WD, et al. Comparison of patterns of relapse
in thymic carcinoma and thymoma. J Thorac Cardiovasc Surg. 2009;138:2631.

5. Chalabreysse L, Etienne-Mastroianni B, Adeleine P, et al. Thymic
carcinoma: A clinicopathological and immunohistological study of 19 cases.
Histopathology. 2004;44:367-374.
6. Takeda S, Sawabata N, Inoue M, et al. Thymic carcinoma. Clinical
institutional experience with 15 patients. Eur J Cardiothorac Surg. 2004;26:
401-406.
7. Huang J, Rizk NP, Travis WD, et al. Feasibility of multimodality
469
RIELY, HUANG, AND RIMNER

therapy including extended resections in stage IVA thymoma. J Thorac
Cardiovasc Surg. 2007;134:1477-1483, discussion 1483-1484.
8. Kondo K, Monden Y. Therapy for thymic epithelial tumors: A clinical
study of 1,320 patients from Japan. Ann Thorac Surg. 2003;76:878-884.
9. Detterbeck FC, Moran C, Huang J, et al. Which way is up? Policies and
procedures for surgeons and pathologists regarding resection specimens of
thymic malignancy. J Thorac Oncol. 2011;6:S1730-S1738.
10. Wright CD. Stage IVA thymoma: Patterns of spread and surgical
management. Thorac Surg Clin. 2011;21:93-97, vii.
11. Ogawa K, Toita T, Uno T, et al. Treatment and prognosis of thymic
carcinoma a retrospective: Analysis of 40 cases. Cancer. 2002;94:3115-3119.
12. Hsu HC, Huang EY, Wang CJ, et al. Postoperative radiotherapy in
thymic carcinoma: Treatment results and prognostic factors. Int J Radiat
Oncol Biol Phys. 2002;52:801-805.
13. Gomez D, Komaki R. Technical advances of radiation therapy for
thymic malignancies. J Thorac Oncol. 2010;5:S336-S343.
14. Gomez D, Komaki R, Yu J, et al. Radiation therapy definitions and
reporting guidelines for thymic malignancies. J Thorac Oncol. 2011;6:S1743S1748.
15. Mayer R, Beham-Schmid C, Groell R, et al. Radiotherapy for invasive
thymoma and thymic carcinoma: Clinicopathological review. Strahlenther
Onkol. 1999;175:271-278.
16. Loehrer PJ, Yiannoutsos CT, Dropcho S, et al. A phase II trial of
pemetrexed in patients with recurrent thymoma or thymic carcinoma. J Clin
Oncol. 2006;24:7079.
17. Girard N, Lal R, Wakelee H, et al. Chemotherapy definitions and
policies for thymic malignancies. J Thorac Oncol. 2011;6:S1749-S1755.
18. Agatsuma T, Koizumi T, Kanda S, et al. Combination chemotherapy
with doxorubicin, vincristine, cyclophosphamide, and platinum compounds
for advanced thymic carcinoma. J Thorac Oncol. 2011;6:2130-2134.
19. Okuma Y, Hosomi Y, Takagi Y, et al. Cisplatin and irinotecan combination chemotherapy for advanced thymic carcinoma: Evaluation of efficacy
and toxicity. Lung Cancer. 2011;74:492-496.
20. Yoh K, Goto K, Ishii G, et al. Weekly chemotherapy with cisplatin,
vincristine, doxorubicin, and etoposide is an effective treatment for advanced
thymic carcinoma. Cancer. 2003;98:926-931.
21. Lemma GL, Lee JW, Aisner SC, et al. Phase II study of carboplatin
470
and paclitaxel in advanced thymoma and thymic carcinoma. J Clin Oncol.

2011;29:2060-2065.
22. Igawa S, Murakami H, Takahashi T, et al. Efficacy of chemotherapy
with carboplatin and paclitaxel for unresectable thymic carcinoma. Lung
Cancer. 2010;67:194-197.
23. Grassin F, Paleiron N, Andre M, et al. Combined etoposide, ifosfamide,
and cisplatin in the treatment of patients with advanced thymoma and
thymic carcinoma. A French experience. J Thorac Oncol. 2010;5:893-897.
24. Loehrer PJ Sr., Jiroutek M, Aisner S, et al. Combined etoposide,
ifosfamide, and cisplatin in the treatment of patients with advanced thymoma
and thymic carcinoma: An intergroup trial. Cancer. 2001;91:2010-2015.
25. Koizumi T, Takabayashi Y, Yamagishi S, et al. Chemotherapy for
advanced thymic carcinoma: Clinical response to cisplatin, doxorubicin,
vincristine, and cyclophosphamide (ADOC chemotherapy). Am J Clin Oncol.
2002;25:266-268.
26. Salter JT, Lewis D, Yiannoutsos CT, et al. Imatinib for the treatment of
thymic carcinoma. J Clin Oncol. 2008;26 (suppl; abstr 8116).
27. Giaccone G, Rajan A, Ruijter R, et al. Imatinib mesylate in patients
with WHO B3 thymomas and thymic carcinomas. J Thorac Oncol. 2009;4:
1270-1273.
28. Palmieri G, Marino M, Buonerba C, et al. Imatinib mesylate in thymic
epithelial malignancies. Cancer Chemother Pharmacol. 2012;69:309-315.
29. Bedano PM, Perkins S, Burns M, et al. A phase II trial of erlotinib plus
bevacizumab in patients with recurrent thymoma or thymic carcinoma. J Clin
Oncol. 2008;26 (suppl; abstr 19087).
30. Kurup A, Burns M, Dropcho S, et al. Phase II study of gefitinib
treatment in advanced thymic malignancies. J Clin Oncol. 2005;23:7068.
31. Giaccone G, Rajan A, Berman A, et al. Phase II study of belinostat in
patients with recurrent or refractory advanced thymic epithelial tumors.
J Clin Oncol. 2011;29:2052-2059.
32. Loehrer PJ Sr, Wang W, Johnson DH, et al. Octreotide alone or with
prednisone in patients with advanced thymoma and thymic carcinoma: An
Eastern Cooperative Oncology Group Phase II Trial. J Clin Oncol. 2004;22:
293-299.
33. Weiss GJ, Hidalgo M, Borad MJ, et al. Phase I study of the safety,
tolerability and pharmacokinetics of PHA-848125AC, a dual tropomyosin
receptor kinase A and cyclin-dependent kinase inhibitor, in patients with
advanced solid malignancies. Invest New Drugs. Epub 2011 Dec 9.
The Creation of the International Thymic

Malignancies Interest Group as a Model for
Rare Diseases
By Frank C. Detterbeck, MD
Overview: Similar to other orphan diseases, little progress

has been made in the past decades in thymic malignancies. A
determination to make a difference, despite the challenges
facing a rare disease, led to the formation of the International
Thymic Malignancies Interest Group (ITMIG) in 2010. This
organization has brought together the majority of those focused on the management of thymic malignancies and has
built a foundation for scientific collaboration, including con-
HYMOMA IS a relatively rare malignancy. The agestandardized incidence has been reported to be 2.5 and
2.8 per million in Denmark and Iceland, respectively.1,2
Studies in the United Kingdom and United States have
reported incidence rates of 0.72 to 1.5 per million, but these
series may have missed many smaller thymomas (i.e., those
previously thought to be benign).3,4 A study of the Surveillance Epidemiology and End Results (SEER) database in the
United States from 1973 to 2006 found a modest and
consistent increase in incidence (Fig. 1).5 This was true for
all subtypes and stages, suggesting that the increased incidence was unlikely to be an artifact (e.g., because of a higher
prevalence of computed tomography imaging).
A recent comparative analysis of SEER data from 1988 to
2003 found that there was no improvement in survival
during this period (Fig. 2).6 For each stage of disease, there
was no consistent evidence of even a trend toward better
survival in more recent years, despite potential advances in
medicine and surgery.
Why has no progress in outcomes been seen over the past
2 decades? A major factor is certainly that the disease is
relatively rare and physicians have largely worked independently. The treatment approach has been primarily empiric,
based on individual judgment with little supporting data.
In addition, most published studies are retrospective series
spanning many decades during which many changes have
occurred and provide only a vague idea of what can be
learned from this experience (Fig. 3). Clinical trials have
been rare, involving only limited numbers of patients in
phase II studies. Also, as it is with any rare disease, research
funding mechanisms and health care structures make it
difficult to establish a scientific basis for approaching the
disease. Thus funding is not available because there is no
scientific basis to build on, and there is no scientific basis
because there is no funding.7
In the case of thymoma, several other issues have hampered progress. First, a common misconception is that many
thymomas are benign. The data do not support this, and this
misnomer should be abandoned.8 However, this misconception together with the view that the thymus (in adults) is an
involuted functionless organ contribute to a lack of interest
in and focus on thymic malignancy. This is worsened by the
fact that cardiac surgeons see the thymus every day as
inconsequential tissue and frequently are willing to remove
or debulk a thymic malignancy with little understanding of
the disease process itself.
Another factor has been the inconsistency with which
sistent use of terms, an international database, and multidisciplinary engagement of clinicians and researchers from
around the world. ITMIG has embarked on the development of
novel approaches to research particularly suited to a rare
condition. ITMIG has gained substantial recognition for the
rapid progress that has been made and serves as a model for
the advancement of knowledge in a rare disease.
terms have been interpreted. This makes published data

almost impossible to compare. Yet for a rare disease, collaboration is absolutely essential. For example, published literature contains a variety of outcome measures; because of
the nature of thymic malignancies, associated diseases and
causes of death, these yield dramatically different outcome
results (Fig. 4).9 These distinctions have rarely been appreciated, and no consistency in reporting results had emerged
until the advent of the ITMIG.
The Development of ITMIG
ITMIG was inaugurated as a formal not-for-profit organization in May 2010. ITMIG is an academic organization,
whose mission is to promote the advancement of science
related to thymic malignancies and other mediastinal conditions to achieve better outcomes for patients. The goal of
ITMIG is to develop an infrastructure that facilitates collaboration and to create innovative approaches that maximize
the progress that can be made.
The catalyst for the development of ITMIG came from the
Foundation for Thymic Cancer Research, an organization
formed by patients and family members who were frustrated
about having to search for prolonged periods before finding a
physician who was truly knowledgeable about thymic malignancy. This group held two conferences in 2007 and 2008
to which physicians active in this disease were invited. In
addition to stimulating discussion and some collaborative
projects, it became clear that real progress in a rare disease
such as thymic malignancy would require creating a scientific infrastructure to foster collaborative research. At a
third meeting held in 2009 at the National Institutes of
Health in Bethesda, Maryland, a provisional structure
was created and tasked with the formal development of
ITMIG.10
Many professional organizations have come forward to
support the creation of ITMIG, including the American
Association for Thoracic Surgery, the European Association
of Cardiothoracic Surgeons, the European Society for Myasthenia Gravis, the European Society of Thoracic Surgeons,
From the Yale University School of Medicine, Thoracic Surgery, New Haven, CT.
Address reprint requests to Frank C. Detterbeck, MD, Yale University School of Medicine,
330 Cedar St., BB205, New Haven, CT 06520-8062; email: frank.detterbeck@yale.edu.
1092-9118/10/110
471
FRANK C. DETTERBECK
Fig. 1. Changing incidence of thymoma and thymic cancer. Data

taken from the Surveillance Epidemiology and End Results (SEER)
database.5
the Fleischner Society, the General Thoracic Surgical Club,

the International Association for the Study of Lung Cancer
(IASLC), the Japanese Association of Chest Surgeons, the
Japanese Association for Research on the Thymus, the
Japanese Association for Thoracic Surgery, the mediastinal
workgroup of the European Society of Pathologists, the
Myasthenia Gravis Foundation of America, the Society of
Thoracic Radiologists, the Society of Thoracic Surgeons, and
the Thoracic Oncology Network of the American College of
Chest Physicians.
As of January 2012, ITMIG has nearly 400 members from
all continents (except Antarctica). ITMIG is a multispecialty
organization, involving thoracic surgeons, medical and radiation oncologists, pulmonologists, radiologists, pathologists,
neurologists, and basic science researchers. Perhaps the
most important early achievement of ITMIG has been the
enthusiastic engagement of nearly all of those around the
world who have been active in studying thymic malignancies. The ITMIG annual meetings in 2010 (in New York) and
in 2011 (in Amsterdam) have been well attended with many
submitted abstracts and posters.
ITMIG Projects and Accomplishments
A prerequisite for collaboration is the ability to speak the

same language. Differences in the interpretation of terms
were surprisingly wide in this field and largely unrecog-
Fig. 2. Trends in 5-year survival for thymoma and thymic cancer

by tumor extent. Data taken from the Surveillance Epidemiology and
End Results (SEER) database.6
nized. The ITMIG community organized a broad process to

clarify critical terms. Multiple workgroups were assembled,
and core members drafted initial proposals, which were
vetted with workgroup members. At a 2-day workshop at
Yale University with broad international representation,
these definitions were discussed and revised so that they
would be aligned with one another. These were then further
refined by the workgroups with input from the entire ITMIG
membership. The final documents were approved by the
ITMIG membership for use in all ongoing research and
publications. The level of engagement and broad consensus
in this process was in itself a major accomplishment of
ITMIG. These definitions were published in a supplement to
the Journal of Thoracic Oncology (JTO) in July 2011 and are
openly available for download from the ITMIG or JTO
websites (itmig.org and jto.org).
A fundamental aspect of cancer research is stage classification, but there is no formal classification system for
thymic malignancies. ITMIG partnered with IASLC, which
has extensive experience in developing the revision of the
staging system for lung cancer, to develop an official thymic
stage classification system for the next (eighth) edition of the
stage classification of tumors in 2017. This project is conducted under the auspices of the Union for International
Cancer Control and American Joint Committee on Cancer,
the entities that determine the official classification systems
for all tumors.
The histologic classification of thymic malignancies has
also been a source of confusion and controversy. There is
KEY POINTS
472
Progress in a rare disease cannot occur without global

collaboration.
The consistent use of terms across a field of study is
necessary for collaboration.
Details of how outcomes are reported for thymic
malignancies are critically important when comparing results.
The recognition of the limitations inherent in data
from small patient cohorts is important to appropriately interpret available results.
Compared with traditional approaches, innovative
approaches to statistics and clinical science provide a
research strategy better suited to a rare disease.
Fig. 3. Overview of published literature results from a PubMed

search from 1989 to 2009 for thymoma, grouped by type of paper
and size of patient cohort.
ITMIG AS A MODEL FOR RARE DISEASES
Fig. 4. Ten-year outcomes for different outcome measures in the

same cohort of resected patients with stage III thymoma.9
variability in how the current World Health Organization

classification system is interpreted and applied, and its
prognostic value is inconsistent. ITMIG has conducted two
international workshops to identify the sources of the difficulties and to develop a strategy for defining a better system.
Clinical science depends heavily on statistics to separate
what we know from perceptions or beliefs. However, a rare
disease presents many statistical challenges. The limited
number of patients magnifies misperceptions caused by
common practices about how clearly something has been
demonstrated. ITMIG invested effort in developing an understanding of the limitations and techniques to minimize or
at least evaluate the level of uncertainty (Fig. 5). As examples, confidence intervals around survival curves provide a
clearer picture of the findings, and identification of prognostic factors may carry a risk of false-positive or false-negative
findings that should be acknowledged. A description of such
issues has also been published in the JTO supplement.11
A sophisticated, detailed international ITMIG database
has been built on the HUBzero platform, thus benefiting
from the engineering expertise of Purdue University, over a
million users of this platform for multiple major initiatives,
and more than a decade of research and experience in
developing platforms for international collaboration involving multiple disparate types of data and analyses. Furthermore, the developers of this platform are focused on
providing tools to facilitate research rather than on promoting a commercial product. The database is linked to a virtual
tissue bank that is actively accruing samples.
Cancer Center at Indiana University and the Purdue Department of Engineering. This project, known as Cancer
Care Engineering (CCE), applies techniques of complex
modeling to cancer research. Developing an adaptive model
allows new insights to be quickly assessed in a virtual
manner and allows for more strategic planning of how to
prioritize and how best to attempt to validate early findings.
Such an approach has been successful in other types of
cancer, and the people involved in the CCE project had
independently come to the idea that the adaptation of this
approach would be particularly useful in a rare disease
coincident with ITMIGs initiative to find novel approaches
that would maximize progress. The initial accomplishments
of ITMIG provide a good foundation on which to build this
effort; it is now time to begin actual development of this
approach.
The traditional clinical research approach relies on providing clear proof of one approach over another through the
use of randomized clinical trials. Although this is part of the
scope of ITMIGs research plans, this strategy is also associated with major challenges, especially in a rare disease.
ITMIG therefore is also including other approaches, particularly the use of Bayesian statistics. These do not seek to
prove superiority or exclude expectations based on prior
data (i.e., biases) as with the traditional frequentist approach. Instead, the Bayesian approach makes use of prior
knowledge and quantifies the possibility that one treatment
is or is not better. Bayesian analyses have the advantage in
a rare disease of being applicable no matter how many
patients are available for inclusion and of refining predictions based on each observation as it happens, instead of
blinding for years until the data are mature.
Traditional research approaches also play a role, where
applicable, in the ITMIG approach. However, in a rare
disease this requires global collaboration. It is hard enough
negotiating the regulatory hurdles for a multi-institutional
study in one country much less meeting expectations across
many countries. ITMIG has partnered with the International Rare Diseases Initiative, which is a collaboration
between relevant organizations in the United States, the
United Kingdom, and Europe, to manage such issues.
A problem that is magnified for a rare disease is funding
for the infrastructure necessary to perform collaborative
Strategy for the Future
Making progress in a rare disease is more challenging

than in a common condition. Merely trying to duplicate the
measures taken in common will leave thymic malignancies
still far behind other areas, despite facilitating some degree
of progress. Therefore a specific focus of ITMIG is to seek out
novel and innovative approaches that allow ITMIG to leapfrog ahead. Each annual meeting has several lectures chosen specifically to promote insight and thinking in areas
that are not widely known but may have implications for
thymic malignancies. Innovative strategies and research
approaches are being explored as well.
A search for a novel strategy that would maximize the rate
of progress led ITMIG to a collaborative project at Simon
Fig. 5. Precision in estimating. The vertical bars are 95% confidence intervals for a 5-year study, based on a standard model of
exponentially decreasing survival, a constant rate of accrual, and no
loss to follow-up.9
Abbreviation: MST, median survival time.
473
FRANK C. DETTERBECK
work and to accomplish the work itself. Traditional mechanisms such as research grants are difficult to qualify for
projects without a large clinical effect or with limited data
existing have a low chance of competing for funding. Industry generally sees the market niche as small with little
return on investment. ITMIG has struggled with these
issues but has managed to stay afloat. The work performed
by the ITMIG members, of course, is purely donated time by
physicians, researchers, and other health care professionals
who feel that the opportunity to move forward is simply
something they have to support. Several industry sponsors
have donated unrestricted gifts in what represents primarily an altruistic gesture. Many related professional organizations have not only officially endorsed ITMIG but also
have been willing to provide start-up money. The bulk of
funding, however, comes from patients and their families
and friends, most notably from the Foundation for Thymic
Cancer Research.
However, one advantage for a rare disease is that a
relatively small amount of funding can have a substantial
effect precisely because it is a rare disease. This limits the
size and cost of the infrastructure needed and promotes

willingness on a broad front for people to volunteer a
manageable amount of time. Maintaining such willingness
broadly and consistently remains a challenge. ITMIG members are engaged and committed, in part because it is easier
to have a feeling of ownership in a smaller group and
because the effort seems to produce real progress.
Conclusion
Thymic malignancies and other mediastinal tumors represent rare diseases, in which there has been little progress
over many decades. For a rare disease, it is clear that
progress is only possible if international collaboration can be
achieved. ITMIG represents an organization that is devoted
to making progress in the management of rare diseases
through international collaboration. ITMIG has built an
infrastructure, has engaged a broad multidisciplinary group
of people in a global initiative, and sought novel approaches
to maximize the progress that can be made in improving
outcomes for patients with these orphan diseases.
Author
Frank C. Detterbeck
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
I-Flow
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Mariusdottir E, Nikulasson S, Bjornsson J, et al. Thymic epithelial
tumours in Iceland: incidence and histopathology, a population-based study.
APMIS. 2010;118:927-933.
2. Engel P, Marx A, Muller-Hermelink HK. Thymic tumours in Denmark.
A retrospective study of 213 cases from 1970-1993. Pathol Res Pract. 1999;
195:565-570.
3. Engels EA, Pfeiffer RM. Malignant thymoma in the United States:
demographic patterns in incidence and associations with subsequent malignancies. Int J Cancer. 2003;105:546-551.
4. dos Santos Silva I, Swerdlow AJ. Thymus cancer epidemiology in
England and Wales. Br J Cancer. 1990;61:899-902.
5. Schwartz A, Kostun L, Henson D. Thymoma and thymic carcinomas: an
analysis of 2,189 cases from the SEER database. Out 4. International Thymic
Malignancy Interest Group (ITMIG) 2nd Annual Conference. Amsterdam, The
Netherlands. July 7-8, 2011.
474
6. Detterbeck F, Morgensztern D. Progress in outcomes for patients with

Thymoma International Thymic Malignancy Interest Group (ITMIG) 2nd
Annual Conference. Amsterdam, The Netherlands: July 7-8, 2011 (abstr
02.01).
7. Lilford RJ, Thornton JG, Braunholtz D. Clinical trials and rare diseases:
A way out of a conundrum. BMJ. 1995;311:1621-1625.
8. Detterbeck FC, Parsons AM. Thymic tumors. Ann Thorac Surg. 2004;
77:1860-1869.
9. Huang J, Wang Z, Loehrer P, et al. Standard Outcome Measures for
Thymic Malignancies. J Thorac Oncol. 2010;5:2017-2023.
10. Detterbeck F. ITMIG: A Way Forward. J Thorac Oncol. 2010;5:s365s370 (suppl 4).
11. Gonen M. Bias, Biostatistics & Prognostic Factors. J Thorac Oncol.
2011;6:s1705-s1709 (suppl 3).
Thymoma: From Chemotherapy to

Targeted Therapy
By Nicolas Girard, MD, PhD
Overview: Thymic malignancies are rare epithelial tumors

that may be aggressive and difficult to treat. Thymomas are
frequently eligible for upfront surgical resection. However,
nearly 30% of patients present with locally advanced tumor at
time of diagnosis, and chemotherapy is then used to reduce
the tumor burdenpossibly allowing subsequent surgery
HYMIC MALIGNANCIES represent a wide range of

clinical, histologic, and radiologic entities, which may
be aggressive and difficult to treat.1-3 The current histopathologic classification distinguishes thymomas from thymic carcinomas; thymomas are further subdivided into
different types (types A, AB, B1, B2, and B3), according to
the morphology of epithelial tumor cells (with an increasing
degree of atypia from type A to type B3), the relative
proportion of the nontumoral lymphocytic component (decreasing from types B1 to B3), and the resemblance to
normal thymic architecture.2 More than 25% of thymomas
actually exhibit morphologic heterogeneity, whereas 10% to
15% combine different histologic types. Tumor invasiveness,
as evaluated by the Masaoka staging system, is a major
predictor of outcome.4,5
Surgery is the mainstay of the curative-intent treatment
of thymic tumors,1 and complete resection represents the
most noteworthy favorable prognostic factor.1,5 Contrary to
thymic carcinomas, recurrences after complete surgical
resection of thymomas are rare and mostly occur locoregionally,5 which limits the rationale for postoperative chemotherapy. Nearly 30% of patients present with locally
advanced tumor at time of diagnosis, with invasion of
intrathoracic neighboring structures, and/or dissemination
to the pleura and the pericardium. In such cases, primary
chemotherapy has been used both to reduce the tumor
burdenpossibly allowing subsequent surgery and/or radiotherapyand to achieve prolonged disease control.1,6
Finally, chemotherapy is also the palliative-intent treatment for unresectable, metastatic, and recurrent thymic
tumors.1,6
As a consequence of the rarity of thymomas, our knowledge regarding chemotherapy in this setting has mainly
been based on retrospective series, most of which were
published years ago. Only a few prospective trials have been
conducted. Most studies included both thymomas and thymic carcinomas, and did not report detailed results by
histologic type. Taken together, these studies, despite recruiting limited numbers of patients over extended period of
times and being heterogeneous with regard to patient selection criteria, therapeutic sequence, and intent of the treatment, demonstrated the chemosensitivity of thymoma to
various cytotoxic agent combinations.1,6 Interpretation of
outcomes data, especially overall survival and response
rates, should integrate 1) the fact that only half of patients
with thymoma actually die as a result of tumor progression,
whereas 25% of deaths are related to thymoma-associated
immunologic manifestations, including myasthenia7; 2) the
correlation between histology and stage, considering that
type A to B1 thymomas more frequently present as stage I to
and/or radiotherapy. Metastatic and recurrent thymic malignancies may be similarly treated with chemotherapy. More
recently, the molecular characterization of thymoma led to the
identification of potentially druggable targets, laying the foundation to implement personalized medicine for patients.
II disease, whereas type B2 to B3 thymomas are usually

diagnosed as stage III to IV disease8; this specific feature
may potentially confound the prognostic or predictive value
of these variables; and 3) the potential effect on lymphocytic
thymomas (types AB, B1, and B2) of corticosteroids, which
are usually delivered concurrently with chemotherapy, and
may produce a substantial reduction of lesion size at imaging studies through lymphocytic depletion, with no antitumor effect.9
Novel strategies are still needed, especially for type B3
thymomas, which, similarly to thymic carcinomas, carry
a poor prognosis despite multimodal treatment.5 In the
past, insights in the biology of thymic tumors were originally made after anecdotal clinical responses to targeted
therapies,3 and substantial efforts were subsequently conducted to dissect the molecular pathways involved in carcinogenesis.3,10-13 Research is hampered by the rarity of
these tumors, evolution of histopathologic concepts, and a
lack of established cell lines and animal models. However,
these studies led to the identification of potentially druggable targets, laying the foundations to implement personalized medicine in the field.
Chemotherapy for Thymoma
Curative-Intent, Preoperative Chemotherapy
In locally advanced thymic malignancies (i.e., unresectable Masaoka stage III and IVA disease at time of diagnosis), chemotherapy aims at making feasible subsequent R0
resection to achieve long-term survival.1,6 Several chemotherapy regimens have been used in this setting, mostly
consisting of doxorubicin- and/or platinum-based multiagent
combinations (Table 1). Usually two to four cycles of chemotherapy are administered before imaging reassessment. In
this setting, response rates to chemotherapy ranged from
70% to 80% in the largest studies (Table 1). Patients for
whom R0 resection was thought to be feasible underwent
surgery, and complete resection was achieved in approximately 50% of cases (Table 1).
In those studies, when the patient was not deemed to be a
From the Department of Respiratory Medicine, Pilot Unit for the Management of Rare
Intrathoracic Tumors, National Expert Center for Thymic Malignancies, Louis Pradel
Hospital, Hospices Civils de Lyon; and UMR 754 Retrovirus and Compared Pathology,
Claude Bernard University, Lyon, France.
Address reprint requests to Nicolas Girard, Service de Pneumologie, Hopital Louis
Pradel, 28, Avenue Doyen Lepine, 69677 Lyon (Bron), France; email: nicolas.girard@chulyon.fr.
1092-9118/10/1-10
475
NICOLAS GIRARD
Table 1. Selected Studies Reporting on Preoperative Chemotherapy or Chemoradiation for Locally Advanced Thymic Tumors
Subsequent Treatment (%)
Study
Chemotherapy
Macchiarini et al 199114
Berruti et al 199315
Rea et al 199316
Berruti et al 199917
Venuta et al 200318
Bretti et al 200419
Kim et al 200420
Lucchi et al 200521
Jacot et al 200522
Yokoi et al 200723
Kunitoh et al 200924
Chemoradiation
Loehrer et al 199725
Wright et al 200826
Primary
Chemotherapy
Regimen
Tumor
No. of
Patients
Type
CEE
ADOC
ADOC
ADOC
CEE
ADOC/PE
CAPP
CEE
CAP
CAMP
CODE
7
6
16
16
15
25
22
36
5
14
21
T/TC
T
T
T
T/TC
T/TC
T
T/TC
T/TC
T/TC
T
III
III-IVA
III-IVA
III-IVA
III
III-IVA
CAP/54 Gy
PE, ADOC, CAP,
CEE/4560 Gy
23
10
T/TC
T/TC
Surgery
Response
Rate (%)
Any
Surgery
Complete
Resection
Radiotherapy
None
III-IVA
III-IVA
III, IV
III
Phase II
Phase II
Retrosp
Phase II
Retrosp
Retrosp
Phase II
Retrosp
Retrosp
Retrosp
Phase II
100
83
100
81
66
72
77
67
75
93
62
100
NA
100
56
100
68
100
69
38
64
62
57
17
69
56
NA
44
72
78
25
14
43
0
NA
0
31
NA
NA
0
19
50
14
24
0
NA
0
13
NS
NA
0
3
12
21
14
III-IVA
III-IVA
Phase II
Retrosp
70
40
0
100
0
80
0
0
100
0
Stage
Design
Abbreviations: T, thymoma; TC, thymic carcinoma; Retrosp, retrospective; CAP, cisplatin (50 mg/m 2 /3 weeks), doxorubicin (50 mg/m 2 /3 weeks), cyclophosphamide
(500 mg/m 2 /3 weeks); ADOC, doxorubicin (40 mg/m 2 /3 weeks), cisplatin (50 mg/m 2 /3 weeks), vincristine (0.6 mg/m 2 /3 weeks), cyclophosphamide (700 mg/m 2 /3 weeks);
PE, cisplatin (60 mg/m 2 /3 weeks), etoposide (120 mg/m 2 3/3 weeks); CODE, cisplatin (25 mg/m 2 /week), vincristine (1 mg/m 2 /week), doxorubicin (40 mg/m 2 /week),
etoposide (80 mg/m 2 3 days/week); CEE, cisplatin (75 mg/m 2 /3 weeks), epirubicin (100 mg/m 2 /3 weeks), etoposide (120 mg/m 2 3 days/3weeks); CAMP, CAP plus
prednisolone (1000 mg/m 2 4 days, 500 mg/m 2 2 days/3 weeks); NA, not available.
surgical candidate either because R0 resection was not

thought to be achievable or because of poor performance
status or coexistent medical condition curative-intent, definitive radiotherapy was delivered. Loehrer and colleagues
reported a phase II trial including 23 patients with thymoma and treated with cyclophosphamide, doxorubicin, and
cisplatin (CAP) chemotherapy followed by radiotherapy to a
total dose of 54 Gy.25 Five-year survival was 53%.
Of note, two studies including patients with locally advanced thymic tumors reported on the use of primary
chemotherapy associated with sequential or concurrent ra-
KEY POINTS
476
Surgery is the mainstay of the treatment of thymic

malignancies. Chemotherapy is recommended in unresectable, locally advanced, and metastatic tumors.
Thymomas are chemosensitive. Major regimens are
based on doxorubicin and cisplatin. No randomized
trial is available that compares different cytotoxic
agent combinations.
Preoperative chemotherapy is recommended for unresectable, locally advanced thymomas, aiming at
allowing subsequent R0 resection, which is the major
predictor of long-term survival.
Adjuvant chemotherapy is not recommended for thymomas, but may be used for thymic carcinomas.
Targeted therapies have been developed empirically
in refractory thymomas, with limited rationale and
poor patient selection. The use of targeted agents in
thymic tumors is currently investigational, not a
routine practice, because other options may exist for
refractory tumors.
diotherapy in a preoperative intent (Table 1). Available

retrospective data do not provide interpretable results comparing chemotherapy with chemoradiotherapy in the preoperative setting. A randomized phase II trial is currently
ongoing, evaluating cisplatin and etoposide chemotherapy
combined with conformal or intensity-modulated radiotherapy as primary treatment of locally advanced thymoma and
thymic carcinoma (clinicaltrials.gov ID: NCT00387868).
After primary chemotherapy, if radiotherapy was not
feasible because of a large tumor burden that precluded safe
delivery of appropriate doses or because of comorbidities
increasing the risks of radiation-induced toxicity, treatment
consisted of chemotherapy alone in a strategy that may
ultimately be considered palliative. In the reported literature, approximately 15% to 20% of patients with locally
advanced thymic tumors receiving upfront chemotherapy
did not receive either surgery or radiation therapy or other
local treatment (Table 1). Survival of these patients was
limited.
Consolidation Chemotherapy
Consolidation chemotherapy refers to chemotherapy delivered after multimodal, curative-intent treatment, aiming
at treating possible residual microscopic disease after surgery. This strategy has been reported by investigators
from the M. D. Anderson Cancer Center (Houston, TX).20
Patients with stage III to IV thymoma received upfront
chemotherapy with three cycles of CAPP (CAP plus prednisone), followed by surgery and adjuvant radiotherapy, followed by consolidation chemotherapy with three cycles of
CAPP. Contrary to adjuvant chemoradiation, which usually
consists of chemotherapy followed by radiotherapy, consolidation chemotherapy was delivered after adjuvant radiotherapy. The role of consolidation chemotherapy within such
multimodal strategy has not been specifically evaluated.
SYSTEMIC TREATMENT OF THYMOMA

Table 2. Landmark Studies Reporting on Palliative Chemotherapy Regimens in Advanced Thymic Malignancies
Study
No. of
Patients
Period of
Accrual
(years)
Single-agent chemotherapy
Bonomi et al 199327
Highley et al 199928
21
15
27
Combination chemotherapy
Fornasiero et al 199030
Tumor
Type
Design
4
12
1
T/TC
T/TC
T/TC
Phase II
Retrosp
Phase II
Cisplatin
Ifosfamide
Pemetrexed
32
11
Retrosp
ADOC
30
T/TC
Phase II
CAP
Giaccone et al 199632
16
Phase II
PE
34
T/TC
Phase II
VIP
Lemma et al 201134
46
T/TC
Phase II
Carbo-Px
Palmieri et al 201135
15
T/TC
Phase II
CAP-GEM
Okuma et al 201136
TC
Retrosp
Cisplatin-Irinotecan
Regimen
Agents
Doxorubicin
Cisplatin
Vincristin
Cyclophosphamide
Cisplatin
Doxorubicin
Cyclophosphamide
Cisplatin
Etoposide
Etoposide
Ifosfamide
Cisplatin
Carboplatin
Paclitaxel
Capecitabine
Gemcitabine
Cisplatin
Irinotecan
Doses
Response
Rate (%)
50 mg/m2/3 weeks
1.5g/m2 5 days/3 weeks
500 mg/m2/3 weeks
10
46
17
40 mg/m2/3 weeks
50 mg/m2/3 weeks
0.6 mg/m2/3 weeks
700 mg/m2/3 weeks
50 mg/m2/3 weeks
50 mg/m2/3 weeks
500 mg/m2/3 weeks
60 mg/m2/3 weeks
120 mg/m2 3/3 weeks
75 mg/m2 4 days/3 weeks
1.2 g/m2 4 days/3 weeks
AUC 5/3 weeks
225 mg/m2/3 weeks
650 mg/m2 bid 14 days/3 weeks
80 mg/m2/4 weeks
91
51
56
32
43
40
56
Abbreviations: T, thymoma; TC, thymic carcinoma; Retrosp, retrospective.
Palliative-Intent, Definitive Chemotherapy
Palliative chemotherapy is administered as the sole treatment modality for thymic tumors, with no plan for surgery
or radiotherapy (e.g., in patients with metastatic or recurrent disease, or with locally advanced tumors that did not
sufficiently shrink after primary chemotherapy to be eligible
for subsequent focal treatment).6 The main objectives of
palliative-intent chemotherapy are to improve potential
tumor-related symptoms and to achieve tumor response.
Prolonged disease control is possible, but tumor eradication
is not expected. Several studies both prospective and retrospective described several regimens for definitive chemotherapy (Table 2), but because there are no randomized
studies, it is unclear which cytotoxic agents are best; multiagent combination regimens and anthracycline-based regimens appear to have improved response rates compared
with others. In general, a combination regimen is recommended, for at least three and no more than six cycles.
Overall, response rates are 20% to 40%, lower than those
observed in the preoperative setting.
In the palliative-intent setting, several consecutive lines
of chemotherapy may be administered when patients present with tumor progression. Delay of progression ranges
from 4 to 80 months in the literature. It is estimated that
50% to 70% of patients with thymoma recurrence receive
chemotherapy, while some recurrences may be eligible for
surgery and/or radiotherapy.7 Strategy may consist of the
readministration of a previously effective regimen,37 as well
as the use of less toxic agents, including paclitaxel or
pemetrexed (Table 2). The repeated use of anthracyclines is
limited by potential cardiac toxicity, which may be even
more relevant in the recurrence setting, in case of combination with radiotherapy and surgery, and given the possible
development of paraneoplastic myocarditis. Specific to recur-
rent thymic tumors after first course of treatment with definitive chemotherapy is the use of octreotide, which as single
agent produced objective tumor responses rates ranging from
10% to 37% in tumors showing increased uptake at OctreoScan
(Indium-111 pentetreotide; Covidien; Dublin, Ireland) scintigraphy.38 Of note, objective responses to octreotide have
been reported only in thymomas, not in thymic carcinomas.
Overall, given the modest results of chemotherapy in the
palliative-intent setting, novel strategies are needed. In this
way, amrubicin, a new generation anthracycline, is currently investigated in refractory thymic tumors (clinicaltrials.gov ID: NCT01364727).
Targeted Therapy for Thymic Malignancies
Despite the rarity of thymic tumors, substantial efforts

have been made to dissect the molecular pathways involved
in thymus carcinogenesis.3,10-13 The main signaling pathways with potential druggable targets that have been explored in thymic tumors are the epidermal growth factor
receptor (EGFR), the KIT/mast/stem-cell growth factor receptor (KIT/SCFR), and the insulin-like growth factor-1
receptor (IGF-1R) pathways. Antiangiogenic agents may
also be of interest.
EGFR Pathway Inhibitors
Overall, EGFR is overexpressed in approximately 70% of

thymomas and 50% of thymic carcinomas.3,10,11 EGFR expression, as well as EGFR gene amplification, was shown to
be higher in stage III to IV tumors. However, EGFR mutations are rare in thymic malignancies.3,10 Thus far, only two
cases of thymomas harboring EGFR mutations have been
found of a total of 158 tumors collectively analyzed. The
mutations were L858R in one case and G863D in the other
case, both of which are associated with response to EGFR
477
NICOLAS GIRARD
3,39
tyrosine kinase inhibitors in lung cancer.

Besides EGFR,
no mutation has been identified in other genes of the
pathway: PIK3CA, AKT1, ERBB2, MEK1, and PTEN.10 In
the Memorial Sloan-Kettering Cancer Center (New York,
NY) series, RAS mutations were observed in two thymomas:
one G12A KRAS mutation and one G13V HRAS mutation.10
These mutations are associated with primary resistance to
EGFR inhibitors.
From a clinical standpoint, the low frequency of EGFRactivating mutations in thymic tumors may explain why
responses to EGFR tyrosine kinase inhibitors have rarely
been observed.3,40,41 On the contrary, several observations
of heavily pretreated recurrent thymoma exhibiting partial
response to cetuximab have been reported.42,43 All tumors
harbored strong EGFR expression by immunohistochemistry. A phase II trial that evaluates the feasibility of delivering cetuximab in combination with CAP in unresectable
locally advanced thymomas (clinicaltrails.gov ID: NCT01025089) is ongoing.
spite the large tumor burden of thymic tumors and the

frequent abutment to mediastinal vascular structures, no
hemorrhagic adverse effect has been reported with the use of
these drugs in these observations.
IGF-1R Pathway Inhibitors
IGF-1R is a transmembrane receptor with tyrosine kinase

activity that was reported by immunohistochemistry to be
overexpressed in thymic tumors.3,11,13 Moderate to high
IGF-1R expression was more frequent in thymic carcinomas
than in thymomas (70% vs. 21%, respectively, p 0.001).
Clinically, figitumumab, an anti-IGF1-R antibody, showed
clinical activity in a patient with refractory thymoma.49 A
phase II trial is ongoing evaluating IMC-A12, another antiIGF-1R antibody, in advanced and refractory thymomas and
thymic carcinomas (clinicaltrials.gov ID: NCT00965250). In
vitro data in a thymoma cell line suggest that IGF-1R
overexpression may also be targeted by heat-shock protein
90 chaperone inhibitors.13
Other Targeted Therapies
KIT and Angiogenesis Inhibitors
KIT is a transmembrane growth factor with tyrosine

kinase activity. A recent pooled analysis of data reported in
the literature indicates that collectively, KIT is overexpressed in 2% of thymomas and 79% of thymic carcinomas.3
KIT gene mutations have been found exclusively in thymic
carcinomas, predicting the efficacy of specific inhibitors,
including imatinib, sunitinib, sorafenib, and dasatinib.3 Interestingly, most of these agents also potently inhibit other
kinases, including vascular endothelial growth factor
(VEGF) receptors (VEGFRs), platelet-derived growth factor
receptor (PDGFR), and Src. The effect of these drugs, especially in KIT-wild-type thymic carcinomas and in thymomas,
may then be partially related to off-target effects, especially
angiogenesis inhibition,44-46 because VEGF-A and VEGFR-1
and -2 are overexpressed by epithelial tumor cells.47
However, only sparse data are available regarding the use
of angiogenesis inhibitors in thymic malignancies. In a
phase II trial, bevacizumab was tested in combination with
erlotinib in 11 thymomas and seven thymic carcinomas.41
No tumor response was observed, but stable disease rate
was 60%. Motesanib diphosphate was reported to induce
partial response in a refractory thymoma.46 In a phase I
study combining docetaxel with aflibercept, one patient with
thymoma experienced partial response.48 Interestingly, de-
Belinostat, a histone deacetylase inhibitor, was evaluated

in thymic malignancies in a recently completed phase II trial
enrolling 41 patients (25 thymomas and 16 thymic carcinomas).50 Response and 2-year survival rates were 8% and
77%, respectively in thymomas, but the drug had no effect
in thymic carcinomas. A subsequent phase II trial is evaluating combination of belinostat with CAP as first-line
treatment for unresectable tumors (clinicaltrials.gov ID:
NCT01100944).
Cyclin-dependent kinase (CDK) proteins, which control
the cell cycle G1-S transition, are altered through p16INK4
loss in thymic tumors, related to a gene methylation mechanism.51 A phase II trial with a CDK inhibitor, PHA848125AC, in advanced thymoma (clinicaltrials.gov IDs:
NCT01301391) is currently ongoing.
For routine practice, the use of targeted agents in thymoma is currently investigational because other options
may exist for refractory tumors. Research efforts are currently being conducted to dissect the molecular biology of
thymic malignancies, and to better understand and predict
the efficacy of chemotherapy and targeted agents. Promising
results are emerging. Given the rarity of these tumors,
translation of preclinical findings to the clinic may be quick
and represents one of the most promising therapeutic approaches for advanced-stage thymoma.
Author
Nicolas Girard*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Girard N, Mornex F, Van Houtte P, et al. Thymoma: a focus on current
therapeutic management. J Thorac Oncol. 2009;4:119-126.
2. WHO histological classification of tumours of the thymus. In: Travis WB,
Brambilla A, Muller-Hermelinck HK, et al. World Health Organization
Classification of Tumour:. Pathology and Genetics of Tumours of the Lung,
Pleura, Thymus and Heart. Lyon: IARC Press, 2004;146.
3. Girard N. Thymic tumors: relevant molecular data in the clinic. J Thorac
Oncol. 2010;5:S291-S295.
4. Detterbeck FC, Nicholson AG, Kondo K, et al. The Masaoka-Koga stage
478
classification for thymic malignancies: clarification and definition of terms.

J Thorac Oncol. 2011;6:S1710-1716.
5. Kondo K, Monden Y. Therapy for thymic epithelial tumors: a clinical
study of 1,320 patients from Japan. Ann Thorac Surg. 2003;76:878-884.
6. Girard N, Lal R, Wakelee H, et al. Chemotherapy definitions and policies
for thymic malignancies. J Thorac Oncol. 2011;6:S1749-S1755.
7. Huang J, Detterbeck FC, Wang Z, et al. Standard outcome measures for
thymic malignancies. J Thorac Oncol. 2011;6:S1691-S1697.
8. Okumura M, Miyoshi S, Fujii Y, et al. Clinical and functional signifi-
SYSTEMIC TREATMENT OF THYMOMA

cance of WHO classification on human thymic epithelial neoplasms: a study of
146 consecutive tumors. Am J Surg Pathol. 2001;25:103-110.
9. Kirkove C, Berghmans J, Noel H, et al. Dramatic response of recurrent
invasive thymoma to high doses of corticosteroids. Clin Oncol. 1992;4:64-66.
10. Girard N, Shen R, Guo T, et al. Comprehensive genomic analysis
reveals clinically relevant molecular distinctions between thymic carcinomas
and thymomas. Clin Cancer Res. 2009;15:6790-6799.
11. Girard N, Teruya-Feldstein J, Payabyab EC, et al. Insulin-like growth
factor-1 receptor expression in thymic malignancies. J Thorac Oncol. 2010;5:
1439-1446.
12. Petrini I, Wang Y, Pineda M, et al. Complete genome sequencing of a
human B3 thymoma. J Thorac Oncol. 2011;6:S452 (suppl).
13. Breinig M, Mayer P, Harjung A, et al. Heat shock protein 90-sheltered
overexpression of insulin-like growth factor 1 receptor contributes to malignancy of thymic epithelial tumors. Clin Cancer Res. 2011;17:2237-2249.
14. Macchiarini P, Chella A, Ducci F, et al. Neoadjuvant chemotherapy,
surgery, and postoperative radiation therapy for invasive thymoma. Cancer.
1991;68:706-713.
15. Berruti A, Borasio P, Roncari A, et al. Neoadjuvant chemotherapy with
adriamycin, cisplatin, vincristine and cyclophosphamide (ADOC) in invasive
thymomas: results in six patients. Ann Oncol. 1993;4:429-431.
16. Rea F, Sartori F, Loy M, et al. Chemotherapy and operation for invasive
thymoma. J Thorac Cardiovasc Surg. 1993;106:543-549.
17. Berruti A, Borasio P, Gerbino A, et al. Primary chemotherapy with
adriamycin, cisplatin, vincristine and cyclophosphamide in locally advanced
thymomas: a single institution experience. Br J Cancer. 1999;81:841-845.
18. Venuta F, Rendina EA, Longo F, et al. Long-term outcome after
multimodality treatment for stage III thymic tumors. Ann Thorac Surg.
2003;76:1866-1872.
19. Bretti S, Berruti A, Loddo C, et al. Multimodal management of stages
III-IVa malignant thymoma. Lung Cancer. 2004;44:69-77.
20. Kim ES, Putnam JB, Komaki R, et al. Phase II study of a multidisciplinary approach with induction chemotherapy, followed by surgical resection, radiation therapy, and consolidation chemotherapy for unresectable
malignant thymomas: final report. Lung Cancer. 2004;44:369-379.
21. Lucchi M, Ambrogi MC, Duranti L, et al. Advanced stage thymomas
and thymic carcinomas: results of multimodality treatments. Ann Thorac
Surg. 2005;79:1840-1844.
22. Jacot W, Quantin X, Valette S, et al. Multimodality treatment program
in invasive thymic epithelial tumor. Am J Clin Oncol. 2005;28:5-7.
23. Yokoi K, Matsuguma H, Nakahara R, et al. Multidisciplinary treatment for advanced invasive thymoma with cisplatin, doxorubicin, and methylprednisolone. J Thorac Oncol. 2007;2:73-78.
24. Kunitoh H, Tamura T, Shibata T, et al. A phase II trial of dose-dense
chemotherapy, followed by surgical resection and/or thoracic radiotherapy, in
locally advanced thymoma: report of a Japan Clinical Oncology Group trial
(JCOG 9606). Br J Cancer. 2010;103:6-11.
25. Loehrer PJ Sr, Chen M, Kim K, et al. Cisplatin, doxorubicin, and
cyclophosphamide plus thoracic radiation therapy for limited-stage unresectable thymoma: an intergroup trial. J Clin Oncol. 1997;15:3093-3099.
26. Wright CD, Choi NC, Wain JC, et al. Induction chemoradiotherapy
followed by resection for locally advanced Masaoka stage III and IVA thymic
tumors. Ann Thorac Surg. 2008;85:385-389.
27. Bonomi PD, Finkelstein D, Aisner S, et al. EST 2582 phase II trial of
cisplatin in metastatic or recurrent thymoma. Am J Clin Oncol. 1993;16:342345.
28. Highley MS, Underhill CR, Parnis FX, et al. Treatment of invasive
thymoma with single-agent ifosfamide. J Clin Oncol. 1999;17:2737-2744.
29. Loehrer PJ, Yiannoutsos CT, Dropcho S, et al. A phase II trial of
pemetrexed in patients with recurrent thymoma or thymic carcinoma. J Clin
Oncol. 2006;24:7079.
30. Fornasiero A, Daniele O, Ghiotto C, et al. Chemotherapy of invasive
thymoma. J Clin Oncol. 1990;8:1419-1423.
31. Loehrer PJ Sr, Kim K, Aisner SC, et al. Cisplatin plus doxorubicin plus
cyclophosphamide in metastatic or recurrent thymoma: final results of an
intergroup trial. The Eastern Cooperative Oncology Group, Southwest Oncol-
ogy Group, and Southeastern Cancer Study Group. J Clin Oncol. 1994;12:
1164-1168.
32. Giaccone G, Ardizzoni A, Kirkpatrick A, et al. Cisplatin and etoposide
combination chemotherapy for locally advanced or metastatic thymoma. A
phase II study of the European Organization for Research and Treatment of
Cancer Lung Cancer Cooperative Group. J Clin Oncol. 1996;14:814-820.
33. Loehrer PJ Sr, Jiroutek M, Aisner S, et al. Combined etoposide,
ifosfamide, and cisplatin in the treatment of patients with advanced thymoma
and thymic carcinoma: an intergroup trial. Cancer. 2001;91:2010-2015.
34. Lemma GL, Lee JW, Aisner SC, et al. Phase II study of carboplatin and
paclitaxel in advanced thymoma and thymic carcinoma. J Clin Oncol.
2011;29:2060-2065.
35. Palmieri G, Merola G, Federico P, et al. Preliminary results of phase II
study of capecitabine and gemcitabine (CAP-GEM) in patients with metastatic pretreated thymic epithelial tumors (TETs). Ann Oncol. 2010;21:11681172.
36. Okuma Y, Hosomi Y, Takagi Y, et al. Cisplatin and irinotecan combination chemotherapy for advanced thymic carcinoma: evaluation of efficacy
and toxicity. Lung Cancer. 2011;74:492-496.
37. Lara PN Jr, Bonomi PD, Faber LP. Retreatment of recurrent invasive
thymoma with platinum, doxorubicin, and cyclophosphamide. Chest. 1996;
110:1115-1117.
38. Loehrer PJ Sr, Wang W, Johnson DH, et al. Octreotide alone or with
prednisone in patients with advanced thymoma and thymic carcinoma: an
Eastern Cooperative Oncology Group phase II trial. J Clin Oncol. 2004;22:
293-299.
39. Yoh K, Nishiwaki Y, Ishii G, et al. Mutational status of EGFR and KIT
in thymoma and thymic carcinoma. Lung Cancer. 2008;62:316-320.
40. Kurup A, Burns M, Dropcho S, et al. Phase II study of gefitinib
treatment in advanced thymic malignancies. J Clin Oncol. 2005;23 (suppl;
abstr 7068).
41. Bedano PM, Perkins S, Burns M, et al. A phase II trial of erlotinib plus
bevacizumab in patients with recurrent thymoma or thymic carcinoma. J Clin
Oncol. 2008:26 (suppl; abstr 19087).
42. Farina G, Garassino MC, Gambacorta M, et al. Response of thymoma to
cetuximab. Lancet Oncol. 2007;8:449-450.
43. Palmieri G, Marino M, Salvatore M, et al. Cetuximab is an active
treatment of metastatic and chemorefractory thymoma. Front Biosci. 2007;
12:757-761.
44. Fiedler W, Giaccone G, Lasch P, et al. Phase I trial of SU14813 in
patients with advanced solid malignancies. Ann Oncol. 2011;22:195-201.
45. Strobel P, Hohenberger P, Marx A. Thymoma and thymic carcinoma:
molecular pathology and targeted therapy. J Thorac Oncol. 2010;5:S286S290.
46. Azad A, Herbertson RA, Pook D, et al. Motesanib diphosphate (AMG
706), an oral angiogenesis inhibitor, demonstrates clinical efficacy in advanced thymoma. Acta Oncol. 2009;48:619-621.
47. Cimpean AM, Raica M, Encica S, et al. Immunohistochemical expression of vascular endothelial growth factor A (VEGF), and its receptors
(VEGFR1, 2) in normal and pathologic conditions of the human thymus. Ann
Anat. 2008;190:238-245.
48. Isambert N, Freyer G, Zanetta S, et al. A phase I dose escalation and
pharmacokinetic (PK) study of intravenous aflibercept (VEGF trap) plus
docetaxel (D) in patients (pts) with advanced solid tumors: preliminary
results. J Clin Oncol. 2008;26:177s (suppl; abstr 3599).
49. Haluska P, Shaw HM, Batzel GN, et al. Phase I dose escalation study
of the anti insulin-like growth factor-I receptor monoclonal antibody CP751,871 in patients with refractory solid tumors. Clin Cancer Res. 2007;13:
5834-5840.
50. Giaccone G, Rajan A, Berman A, et al. Phase II study of belinostat in
patients with recurrent or refractory advanced thymic epithelial tumors.
J Clin Oncol. 2011;29:2052-2059.
51. Hirabayashi H, Fujii Y, Sakaguchi M, et al. p16INK4, pRB, p53 and
cyclin D1 expression and hypermethylation of CDKN2 gene in thymoma and
thymic carcinoma. Int J Cancer. 1997;73:639-644.
479
CONTROVERSIES IN INDOLENT LYMPHOMA

CHAIR
Sonali M. Smith, MD
Chicago, IL
SPEAKERS
Gilles Salles, MD, PhD
Hospices Civils de Lyon and Universite de Lyon
Lyon, France
Ajay K. Gopal, MD
Seattle, WA
What Is the Best Strategy for Incorporating

New Agents into the Current Treatment of
Follicular Lymphoma?
By Sonali M. Smith, MD
Overview: Although there is increasing knowledge about the

pathobiology of follicular lymphoma (FL), the incorporation of
new agents is challenged by the long clinical course and
inherent heterogeneity of the disease. Furthermore, a longstanding concept in FL is that although most patients have an
indolent initial phase of disease, this is typically followed by
sequentially shorter remission durations and justifies the
continued intense search for new rationally designed agents.
Ideally, there would be personalized prognostic tools, preemptive target identification, and means to predict response
in individual patients. Short of having these tools, one conceptual approach is to consider FL as a series of clinical
disease states divided between treatment-nave (low tumor
S THE prototype of indolent lymphomas, follicular

lymphoma (FL) mandates a longitudinal perspective
when considering clinical management options. The median
life expectancy has improved, with approximately 40% of
patients surviving in excess of 10 years.1,2 During this time,
patients can have variable clinical courses and, although
clinical prognostic indices are helpful, they are often difficult
to apply to individual patients outside of a research setting.
Furthermore, treatment decisions made early in the disease
may influence future options, and the optimal sequencing of
treatments is poorly defined. The traditional paradigm has
been that with each successive therapy, the duration of
response shortens until a blatantly chemoresistant picture
emerges. However, today a clearer understanding of FL
biology has led to the emergence of a new repertoire of
agents with more rational and targeted mechanisms of
action. These agents challenge the notion of inevitably
shorter response durations, and offer hope of improved
clinical outcomes compared with traditional sequential cytotoxic therapy.
A major challenge to integrating new agents or new
approaches is the inherent clinical and biologic heterogeneity of follicular lymphoma. Ideally, there would be personalized prognostic tools, preemptive target identification, and
means to predict response in individual patients, followed by
randomized clinical trials validating each agent. Despite
major gains in dissecting the heterogeneity of FL (Table 1),
there are few existing ways of predicting an individual
patients overall disease course. One approach is to consider
FL not simply as one disease but as sequential clinical
disease states: treatment-nave (low tumor burden and high
tumor burden), relapsed (chemoimmunotherapy-sensitive),
and multiply relapsed/chemoimmunotherapy resistant disease (Fig. 1). At each point, the best approach to incorporating new agents into the current treatment of FL requires an
understanding of pathogenetic mechanisms, a consideration
of the current and most promising agents, and conscientious
trial design in the context of desired patient outcomes. This
review will present a conceptual framework of clinical disease states in FL, providing examples along the way of how
biologic insights and new agents are currently being integrated into existing treatment paradigms.
burden and high tumor burden), relapsed (typically still

chemoimmunotherapy-sensitive), and multiply relapsed (usually chemoimmunotherapy-resistant) disease. By applying
what is known about the biology of FL along with the available
agents, new treatment options can be better defined and
tested within these clinical contexts. During the last few
years, novel chemotherapeutics, biologic agents, monoclonal
antibodies, antibody drug conjugates, and maintenance strategies are all either replacing or adding onto existing strategies. These new agents and approaches challenge the notion
of inevitably shorter response durations, and offer hope of
improved clinical outcomes compared with traditional sequential cytotoxic therapy.
LowTumor Burden FL (Treatment Nave)
The vast majority of patients with FL have an asymptomatic presentation, with either palpable or radiographically
visible adenopathy discovered incidentally. At initial diagnosis, the pace of disease remains unknown for individual
patients, and there are no tools to accurately predict disease
behavior. Tumor grade is a rough measure of disease aggressiveness, with FL grade 3b behaving akin to diffuse large
B-cell lymphoma. But for patients with FL grade 1 to 2,
comprising the bulk of patients, grade itself is a poor
predictor of outcome. The FL International Prognostic Indices (FLIPI 1 and 2) are helpful, but it is important to
remember that all patients included in these retrospective
analyses were already being considered for therapy and
therefore do not adequately reflect a newly diagnosed,
asymptomatic patient with low tumor burden.3,4 Newer
biologic markers may help identify patients at high risk for
aggressive disease behavior, but are far from routine application.
Despite these hurdles, lowtumor burden FL is an ideal
setting in which to evaluate novel agents or novel application of existing agents. The National Lymphocare Study
found that nearly one-fifth of patients undergo a watch and
wait strategy, reflecting that no therapeutic intervention to
date has shown a survival benefit.5 Strong support for
watch and wait can be derived from randomized trials that
had shown no advantage to early therapy.6,7 Notably, these
studies were performed before the advent of monoclonal
antibodies, and the recent randomized trial of watch and
wait compared with a short course of the anti-CD20 monoclonal antibody rituximab followed by a maintenance strategy has reinvigorated the discussion regarding the timing of
treatment initiation in newly diagnosed, lowtumor burden
From the Section of Hematology/Oncology, Lymphoma Program, The University of

Chicago, Chicago, IL.
Address reprint requests to Sonali M. Smith, MD, Section of Hematology/Oncology,
Lymphoma Program, The University of Chicago, 5841 S. Maryland Ave., MC2115, Chicago,
IL 60637; email: smsmith@medicine.bsd.uchicago.edu.
1092-9118/10/1-10
481
SONALI M. SMITH
Table 1. Clinical, Biologic, and Genetic Factors with Prognostic Value in Follicular Lymphoma*
Factor
Histopathologic features
Tumor grade
Diffuse architecture
Proliferation rate
Clinical indices
IPI
Clinical Impact
Grade 12 is generally more indolent than grade 3
Controversial
Worsening survival with high scores

Worsening survival with high scores
FLIPI 2
Worsening PFS with high scores
VEGF, FGF, TNF, endostatin

Individual proteins or RNA
BCL6, CD10, PU.1
MCL1, MUM1, SOCS3, YY.1, BCLXL
Gene expression profiling
Immune response-1 and immune response-2
Unable to distinguish prognosis amongst vast majority of

patients who have low IPI scores
Analysis was done in a pre-rituximab population. Primary end
point was overall survival.
Analysis was done in patients receiving immunochemotherapy.
Primary end point was PFS.
Decreased survival with elevated values

Shorter FFP and decreased survival with elevated
values
Shorter FL
MYC
TP53
CCNB1
Part of IPI and FLIPI-1

Part of FLIPI-2
Small numbers of FL patients in most of these studies
Favorable
Unfavorable
Immune response-1 has 9-fold more favorable
survival
81-gene predictor model

Microenvironment
Macrophages
CD4 or CD8 T cells
T-regulatory cells
Molecular genetic biomarkers
BCL2
Reproducibility amongst pathologists can be low. Grade 3A is

treated similarly to grade 12. Grade 3B is treated similarly
to diffuse large B-cell lymphoma
Increased areas of diffuse architecture may increase risk of
transformation
Controversial
FLIPI 1
Blood markers
Lactate dehydrogenase
Beta-2 microglobulin
Comment
Model was able to predict disease aggressiveness and disease

progression but was done in pre-rituximab era.
Decreased survival
Controversial
Controversial
Conflicting reports in literature

Conflicting reports in literature
Typically diagnostic and not prognostic.

Worse prognosis
Worse prognosis
Improved survival
There may be differing prognostic implications of translocation

versus somatic mutations as source of BCL aberration.
Usually associated with transformation to high grade lymphoma.
Associated with histologic transformation.
Abbreviations: FGF, fibroblast growth factor; FL, follicular lymphoma; FLIPI, Follicular Lymphoma International Prognostic Index; IPI, International Prognostic Index;
PFS, progression-free survival; TNF, tumor necrosis factor; VEGF, vascular endothelial growth factor.
* There is little consensus on the prognostic applicability of these biomarkers on an individual basis. Data abstracted from Relander T, Johnson NA, Farinha P, et al.
Prognostic factors in follicular lymphoma. J Clin Oncol. 2010;28:29022913.
KEY POINTS
482
Follicular lymphoma is clinically and biologically

heterogeneous.
There are currently few validated predictive or prognostic tools available to guide treatment at an individual level.
Clinically, follicular lymphoma can be conceptualized
as a series of disease states with unique treatment
approaches and goals of treatment.
Advances in the pathobiology of follicular lymphoma
have led to a new generation of agents that are now in
clinical trials.
The optimal incorporation of these new agents will
likely build on existing treatment paradigms within
clinically distinct groups: treatment-nave (low tumor
burden and high tumor burden), chemoimmunotherapy-sensitive, and chemoimmunotherapy-resistant
disease.
FL.8 Patients in the maintenance rituximab arm were

significantly less likely to require cytotoxic therapy at 3
years (91% vs. 48%; hazard ratio [HR] 0.37; p 0.001);
there was no improvement in overall survival, which was
95% in all arms. The Eastern Cooperative Oncology
Group (ECOG) RESORT (Rituximab Extended Schedule
or Re-Treatment Trial) trial sought to determine whether
a maintenance rituximab strategy following induction
rituximab could improve time to treatment failure compared
with a rituximab re-treatment strategy.9 Although there
was no difference in the primary end point of time to
treatment failure (3.6 vs. 3.9 years for rituximab retreatment vs. maintenance arms, respectively), significantly
fewer patients in the maintenance arm had not yet required
cytotoxic therapy at 3 years (HR 2.5; p 0.027). Should
rituximab induction plus maintenance replace watch and
wait as the initial management approach for lowtumor
burden and asymptomatic patients? This is clearly a matter
of debate, but highlights the interest and need to integrate
new agents in this clinical setting in which delayed time to
cytotoxic therapy could be a reasonable goal for some patients.
TREATMENT OF FOLLICULAR LYMPHOMA
Fig 1. Conceptual clinical approach to

follicular lymphoma. Although the overall
goals of disease control and improved
survival apply to all phases of disease, there
are unique clinical considerations and treatment options according to tumor burden
and known sensitivity versus resistance to
chemotherapy and immunotherapy.
Abbreviations: Allo HCT, allogeneic hematopoietic cell transplantation; MR, maintenance rituximab; R-chemo, rituximab plus
chemotherapy; RIT, radioimmunotherapy.
For many patients with lowtumor burden FL who either

need or desire treatment, several groups have tested relatively low-intensity therapies with variable objectives.
Among these, rituximab plus lenalidomide is emerging as an
extremely promising biologic doublet that could be a novel
therapeutic backbone. In contrast to most antineoplastic
agents in FL, lenalidomide appears to have pleiotropic
effects on both the malignant and nonmalignant components
of lymphoma (reviewed in Press et al22). The importance of
the microenvironment in FL is underscored by studies
showing a nine-fold survival difference on the basis of the
gene expression profile of background, nonmalignant
immune-response cells.23,24 As a single agent in relapsed
FL, lenalidomide has a response rate of approximately 30%.
When combined with rituximab, there appears to be clinical
synergy, prompting this doublet to be tested in front-line
settings. The Cancer and Leukemia Group B (CALGB)/
Alliance recently completed accrual to a front-line study of
lenalidomide and rituximab (NCT01145495), with promising preliminary results. The M. D. Anderson Cancer Center
also tested lenalidomide and rituximab, and reported an
impressive 79% complete remission rate in 30 patients with
FL.10 The Swiss cooperative group is testing rituximab with
or without lenalidomide in a randomized phase II setting
(NCT01307605). Clearly, this is a regimen of great interest,
and has the potential to replace chemotherapy (which is
being prospectively tested in patients with high tumor
burden by the Groupe dEtude des Lymphomes de lAdulte
[GELA]).
Targeted agents, including monoclonal antibodies and
radioimmunoconjugates have also been tested in treatmentnave patients with low tumor burden. Single agent I131
tositumumab results in a 75% complete remission rate and
median progression-free survival (PFS) of 6.1 years in newly
diagnosed patients.11 Similarly, 90Y-ibritumomab tiuxetan
appears most efficacious when used in earlier disease
states.12 To date, despite this encouraging data, radioimmunotherapy as a single agent has not been approved for newly
diagnosed FL, and there are no comparative studies against
single-agent rituximab. More recently, a new generation of
monoclonal antibodies with possible advantages over ritux-
imab are being tested in front-line settings, including ofatumumab and GA101.
There are several quite promising agents being tested in
the relapsed setting (discussed later herein) that should
make their way to this population, but the biggest challenge
to incorporating these and other new agents in the front-line
setting for patients with asymptomatic or lowtumor burden
FL is agreeing on the best clinical and trial end points.
Overall survival benefit requires years, perhaps decades, of
observation, and progression-free survival does not reflect
the fact that many low tumor burden patients do not require
therapy even when there is radiographic or clinical proof of
mild progression. Time to cytotoxic intervention, as selected
in the United Kingdom trial and the ECOG RESORT trial,
may be more reflective of real-world dilemmas, but the
definition of this end point needs to be refined and agreed on
in the broader research community.
HighTumor Burden FL (Treatment Nave)
For patients with symptoms, or for those meeting Groupe

dEtude des Lymphomes Folliculaires (GELF) criteria, effective treatment has immediate benefit in terms of symptom
relief and reduction in tumor burden. The most common
approach has been chemoimmunotherapy induction with or
without rituximab maintenance or radioimmunotherapy
consolidation. There are at least three important broad
questions for patients in which new agents could make an
impact in this clinical disease state: 1) What is the best
induction chemoimmunotherapy combination; 2) What is
the best strategy for prolonging remission; and 3) Can a
non chemotherapy-based regimen supplant classic cytotoxics?
The recent success of bendamustine is an instructive
example of how an agent, even if chemotherapy based, can
dramatically affect the standard of care, and perhaps serve
as a backbone for other agents. Bendamustine is a unique
chemotherapeutic agent with bifunctional properties of both
alkylating agents and purine analogs and strong singleagent activity despite prior alkylator exposure.13-15 When
combined with rituximab, again in the relapsed setting, it
showed high response rates with promising durability.16
483
SONALI M. SMITH
On the basis of encouraging data in relapsed patients, the

German Low-Grade Lymphoma Study Group compared bendamustine and rituximab (BR) against rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (RCHOP) in 549 treatment-nave indolent lymphomas, with
FL comprising more than half of patients. All patients
required therapy as per predefined criteria. Initially designed as a noninferiority study, BR instead showed superior complete response rates (40.1% vs. 30.8%; p 0.03) and
PFS (54.8 vs. 34.8 months; p 0.0002); with a median
follow-up time of approximately 3 years, there is a trend
toward improved overall survival.17 The option to avoid
anthracyclines is appealing, particularly in a disease in
which the median age is in the sixth decade of life and
comorbidities often influence therapeutic choices. A confirmatory trial of BR compared with rituximab, cyclophosphamide, vincristine, prednisone (R-CVP) or R-CHOP recently
completed accrual (NCT00877006). For now, although it is
not clear that BR should be the front-line standard for all
patients in this disease state, it clearly provides a new
standard in terms of efficacy and tolerability for patients
with high tumor burden/symptomatic FL and is a platform
for new regimens. An Eastern Cooperative Oncology Group
(ECOG)-led study is evaluating BR with or without bortezomib followed by rituximab with or without lenalidomide
as one example (NCT01216683).
Despite the high response rates to most chemoimmunotherapy regimens, relapse is inevitable and has prompted
evaluation of maintenance or consolidation strategies. There
are currently two approved postremission strategies: maintenance rituximab (on the basis of the PRIMA trial) and
consolidative radioimmunotherapy (on the basis of the FIT
trial). The PRIMA (Primary Rituximab and Maintenance)
trial provided clear evidence that maintenance rituximab,
even after a rituximab-containing chemotherapy induction,
could improve PFS (74.9% vs. 57.6%; p 0.0001). However,
there was no improvement in overall survival. The FIT
(First-Line Indolent) trial delivered induction chemotherapy
(with or without rituximab) and then randomly assigned
patients to either 90Y-ibritumomab tiuxetan or observation.
Patients in the consolidation arm had a significant conversion of partial to complete remissions and a near-tripling of
PFS (13.3 vs. 36.5 months; HR 0.465; p 0.0001).18 A
recent update of the FIT trial showed that PFS advantage
persists with 66-month median follow-up.19 However, only
15% of patients received rituximab as part of their initial
induction. Two phase II studies support the addition of
consolidative 90Y-ibritumumab tiuxetan after chemoimmunotherapy, but a definitive phase III trial is lacking.20,21 I131
tositumumab consolidation has also been tested, with
clearly promising results after either CHOP or CVP
chemotherapy.22-24 Mature follow-up of Southwest Oncology
Group (SWOG) protocol S9911 shows a 5-year PFS and
overall survival of 67% and 87%, respectively, which significantly exceeds results for patients receiving treatment
without immunotherapy in prior SWOG studies. This promising phase II trial prompted a randomized phase III trial of
CHOP with either rituximab or consolidative I131 tositumumab, which showed equivalence of both arms and overall
excellent outcomes.25
Should all patients receive maintenance or consolidation
after chemoimmunotherapy? An unplanned but intriguing
subanalysis of the PRIMA trial evaluated the role of func-
484
tional imaging via fluorodeoxyglucose positron emission

tomography (FDG-PET) as a reflection of minimal residual
disease (MRD).26 Thirty-two (26%) of 122 patients remained
PET positive at the end of induction. PET positivity was
strongly predictive of PFS (20 months vs. not reached; p
0.001) and was more sensitive than standard computed
tomography (CT) criteria at identifying patients with an
inferior outcome. As a result of small numbers, it was not
possible to determine whether maintenance rituximab was
able to overcome post-induction PET positivity. However,
the use of PET in the PRIMA study reflects the general
interest in determining MRD status in FL as a predictor of
adverse outcome, and this might identify a population in
which new agents could be evaluated.
Several considerations regarding maintenance and consolidation should be raised. First, if a front-line regimen has
greater relative efficacy, it is more difficult to demonstrate
the impact of a maintenance or consolidation strategy. As an
example, to date, there are no data showing that maintenance or consolidation after BR is superior to observation
alone. Second, there are currently no data guiding the
selection of maintenance rituximab compared with consolidative radioimmunotherapy. In addition, there are other
agents, such as lenalidomide or others (discussed later
herein), that are orally bioavailable, have encouraging
safety profiles, and show preliminary activity that should be
tested in this setting.
First and Second Relapse of FL
Perhaps one of the most heterogeneous time points of FL

is the time of first or second relapse. At this point, disease
can be rituximab sensitive or resistant, and chemotherapy
sensitive or resistant. Patients with either an asymptomatic
or low-volume relapse generally receive treatment similar to
that of patients who are newly diagnosed, using chemoimmunotherapy or biologic agents, whereas patients with a
quick relapse or aggressive course receive treatment similar
to that of patients with multiple relapses, as discussed in the
following pages. Similar to the front-line setting, there are
no readily available clinical or biologic markers to predict
outcome or select therapeutic regimens.
For many patients, the first or second relapse, particularly
if the duration of response to previous regimens has been
short, is the ideal time to consider autologous or allogeneic
hematopoietic stem-cell transplantation. Although this is
reviewed in detail elsewhere, it is important to note that
dose intensification with autologous stem-cell transplantation has been a helpful tool for many patients in first or
second relapse. However, it is equally critical to note that
the vast majority of data were generated before the advent of
rituximab and that there are no randomized trials in the
modern era comparing autologous stem-cell transplantion
with chemoimmunotherapy regimens. Allogeneic stem-cell
transplantation is the only known curative modality for
relapsed advanced-stage FL, and a careful discussion of
potential risks and benefits with an experienced transplant
center should be considered for patients at early relapse,
particularly if there has been limited benefit to chemoimmunotherapeutic regimens.
Similar to front-line settings, bendamustine is increasingly used as a therapeutic backbone to which new agents
are added. One example is the addition of the proteasome
inhibitor bortezomib. The ubiquitin-proteasome pathway
mediates the ubiquitation and degradation of proteins, and

is frequently deregulated in lymphomas, including FL. Although suppression of nuclear factor-kappaB (NFB) appears to be the major mechanism of action of bortezomib,
there is also upregulation of proapoptotic factors such as
NOXA and downregulation of antiapoptotic factors. The
increased activity of proapoptotic factors in a disease that
is characterized by universal bcl2 overexpression may be
mechanistically important. The VERTICAL trial (Velcade
in Relapsed or Refractory Follicular Lymphoma) added
four weekly doses of bortezomib to a 35-day cycle of
escalating doses of bendamustine plus rituximab for five
consecutive cycles.27 Among 63 patients receiving the
highest bendamustine dose level, the overall response
rate was 88% and complete response rate was 53%. Another
trial testing this combination had a slightly different
treatment schedule that used the more standard bortezomib
scheduling of twice-weekly doses.28 Again, there was an
impressive overall response rate of 83%, and this regimen
is currently undergoing further testing in both relapsed
and front-line settings to determine the incremental benefit of bortezomib to a bendamustine and rituximab backbone.
Given high patient heterogeneity, it is clear that a subset
of patients stand to benefit a great deal from new agents
used at this point in the overall disease course. However,
short of randomized multiarm trials, the challenge will be to
standardize patients in some way to allow fair comparisons
of different regimens. It is not sufficient to use the number of
prior regimens, because, as discussed earlier, the front-line
regimens vary greatly, and time to relapse and determination of relative rituximab sensitivity compared with resistance all influence the likelihood of response to subsequent
therapy. A second-line FLIPI or other biologic tool is greatly
needed. Until these tools are available, this heterogeneous
disease state will remain the most challenging setting in
which to assess relative efficacy of a new therapy, and
incorporation of new agents will be difficult to apply in a
controlled fashion.
Multiply Relapsed and/or Refractory FL: Promising
Agents and Approaches for Chemoresistant and
Rituximab-Resistant Disease
Despite the major positive impact of chemoimmunotherapy, nearly all patients eventually demonstrate both
chemotherapy and rituximab resistance, making multiply
relapsed FL a disease state of great unmet need. An agent
capable of demonstrating activity in this clinical setting
would quickly have impact and could be moved earlier into
the disease course, as reflected by lenalidomide and bendamustine, both of which first showed activity in relapsed/
refractory FL. This is currently the disease state in which
gains in knowledge regarding FL pathogenesis have spurred
new and targeted agents with promising clinical validation
ongoing.
Among B-cell malignancies, few potential targets are as
ubiquitous as the B-cell receptor (BCR) and its downstream
signaling cascade. BCR normally responds to antigens by
triggering an internal signal leading to gene transcription
and B-cell activation and proliferation.29 Malignant B-cell
transformation usually retains the need for an intact and
tonically active BCR, and agents that block its signaling
have shown promising preclinical effects.30-33 The Tec-
kinase Brutons tyrosine kinase (BTK) appears to be required to form immunoglobulin and to allow B-cell survival
as part of BCR signaling.34 PCI-32765 is a novel, orally
available, irreversible covalent inhibitor of BTK. Preclinical
studies confirm its selectivity for its target, the ability to
completely halt BCR signaling, and potent activity in B-cell
lymphomas, chronic lymphocytic leukemia models, and autoimmune models.35-38 A phase I trial in B-cell non-Hodgkin
lymphoma showed an impressive response rate of 54% in an
intent-to-treat population that included both aggressive and
indolent histologies.39 Among 16 patients with FL, one-third
had an objective response including three complete responses. Early results are promising, but it is also clear that
single-agent PCI-32765 primarily leads to partial responses
in FL and its impact on response durability is yet to be
determined. Using this agent in earlier disease states is
worthy of investigation.
Just downstream of BCR signaling is the phosphoinositide
3-kinase (PI3K)/Akt/mammalian target of rapamycin
(mTOR; PAM) axis, which is also emerging as a major
pathogenetic mechanism in FL. The natural function of
PI3K is to transduce external growth signals and modulate downstream targets that control cellular proliferation, motility, metabolism and cell growth vs. survival
(reviewed in Courtney, Corcoran, and Engelman40). There
are four PI3K isoformsalpha, beta, gamma, deltawith
the delta isoform having restricted expression in human
leukocytes. CAL-101 is an orally bioavailable PI3K
inhibitor that is highly selective for the p110 delta isoform.41
In vitro models show that CAL-101 can block tonic PI3K
signaling with decreased activation of downstream targets, including Akt. Preliminary data from phase I studies
show promising activity of CAL-101 in B-cell malignancies, including FL.42-44 Among 30 patients with indolent
lymphomas, including 17 patients with FL, the single-agent
overall response rate was 63%, with a median PFS exceeding 1 year. The most common grade 3 or 4 event was
transient increase of hepatic enzymes, which occurred in
27% of patients. On the basis of the safety profile and
efficacy, CAL-101 has been added to the backbone of
bendamustine and rituximab in a phase I combination
study (NCT01088048).45 Preliminary results show excellent
tolerability and an overall response rate of more than 65%
in a group of heavily pretreated patients with indolent
lymphomas.
An important downstream substrate of PI3K and Akt is
the serine/threonine kinase mTOR. The natural functions of
mTOR are to integrate growth signals and nutrient availability and influence cell growth via control over mRNA
translation. Several preclinical investigations support
the central role of mTOR in FL (reviewed in Smith46).
Single-agent temsirolimus, a rapamycin analog, was tested
in 39 patients with relapsed or refractory FL; more than
half of patients had an objective response, including 25%
complete responses.47 Nonhematologic toxicities included
metabolic abnormalities (hyperglycemia, hypertriglyceridemia, hypercholesterolemia), stomatitis, and rash. Combination
studies
of
temsirolimus
plus
lenalidomide
(NCT01076543) and everolimus plus lenalidomide
(NCT01075321) are ongoing.
A near-universal feature of FL is BCL2 overexpression,
primarily as a result of the hallmark translocation, t(14;18).
The constitutive expression of the antiapoptotic BCL2 pro-
485
SONALI M. SMITH
tein is not only diagnostic, but also probably underlies the

characteristic disease persistence after standard therapy.
Downregulation of BCL2 in preclinical models improves
chemosensitivity to other agents, and could be an important
therapeutic adjunct. G3139, an antisense oligonucleotide
against BCL2 mRNA, showed preliminary activity in indolent lymphomas but is no longer in development.48,49 Currently, several small-molecule inhibitors are in early phase
trials, including navitoclax (ABT-263) and obatoclax mesylate (GX15 070).
Finally, unique B-cell surface molecules provide opportunity for targeted agents to have a major impact. Novel
monoclonal antibodies and antibody-drug conjugates targeting CD20 and non-CD20 surface molecules are under active
investigation including ofatumumab, GA101, veltuzumab,
90
AME-133, inotuzumab ozogamicin,

Y-epratuzumab,
SAR3419 and others (reviewed in Leonard and Martin50).
Conclusion
There is a plethora of new agents, and the optimal

incorporation of these agents is highly dependent on the
clinical setting and goals of therapy. The ideal agent should
reflect the underlying biology, with preclinically validated
targets and better predictive tools to help individualize
therapy. Many of these new agents are being added to or are
supplanting standard cytotoxics as a result of their favorable toxicity profile and promising efficacy. Although we are
short of a cure for most patients, it is clear that incremental
benefit has been achieved and collaborative efforts should
continue.
Author
Sonali M. Smith
Employment or
Leadership
Positions
Consultant or
Advisory Role
Biogen Idec;
Celgene;
Cephalon;
Genentech;
GlaxoSmithKline;
Spectrum
Pharmaceuticals
Stock
Ownership
Honoraria
Biogen Idec;
Celgene;
Genentech
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Swenson WT, Wooldridge JE, Lynch CF, et al. Improved survival of
follicular lymphoma patients in the United States. J Clin Oncol. 2005;23:
5019-5026.
2. Fisher RI, LeBlanc M, Press OW, et al. New treatment options have
changed the survival of patients with follicular lymphoma. J Clin Oncol.
2005;23:8447-8452.
3. Federico M, Bellei M, Marcheselli L, et al. Follicular lymphoma international prognostic index 2: a new prognostic index for follicular lymphoma
developed by the international follicular lymphoma prognostic factor project.
J Clin Oncol. 2009;27:4555-4562.
4. Solal-Celigny P, Roy P, Colombat P, et al. Follicular lymphoma international prognostic index. Blood. 2004;104:1258-1265.
5. Friedberg JW, Taylor MD, Cerhan JR, et al. Follicular lymphoma in the
United States: first report of the national LymphoCare study. J Clin Oncol.
2009;27:1202-128.
6. Ardeshna KM, Smith P, Norton A, et al. Long-term effect of a watch and
wait policy versus immediate systemic treatment for asymptomatic advancedstage non-Hodgkin lymphoma: a randomised controlled trial. Lancet. 2003;
362:516-522.
7. Brice P, Bastion Y, Lepage E, et al. Comparison in low-tumor-burden
follicular lymphomas between an initial no-treatment policy, prednimustine,
or interferon alfa: a randomized study from the Groupe dEtude des Lymphomes Folliculaires. Groupe dEtude des Lymphomes de lAdulte. J Clin
Oncol. 1997;15:1110-1117.
8. Ardeshna KM, Qian W, Smith P, et al. An Intergroup randomised trial
of rituximab versus a watch and wait strategy in patients with stage II, III,
IV, asymptomatic, non-bulky follicular lymphoma (grades 1, 2 and 3a): a
preliminary analysis. Blood. ASH Annual Meeting Abstracts. 2010;116:
abstract 6.
9. Kahl BS, Hong F, Williams ME, et al. Results of Eastern Cooperative
Oncology Group Protocol E4402 (RESORT): a randomized phase III study
comparing two different rituximab dosing strategies for low tumor burden
follicular lymphoma. Blood. ASH Annual Meeting Abstracts. 2011;118; LBA
6.
10. Samaniego F, Hagemeister F, McLaughlin P, et al. High response rates
with lenalidomide plus rituximab for untreated indolent B-cell non-Hodgkin
lymphoma, including those meeting GELF criteria. J Clin Oncol. 2011;29
(suppl; abtr 8030).
11. Kaminski MS, Tuck M, Estes J, et al. 131I-tositumomab therapy as
initial treatment for follicular lymphoma. N Engl J Med. 2005;352:441-449.
12. Scholz CW, Pinto A, Linkesch W, et al. 90Yttrium ibritumomab
486
tiuxetan as first line treatment for follicular lymphoma: first results from an
international phase II clinical trial. Blood. 2010;116:593A.
13. Cheson BD, Rummel MJ. Bendamustine: rebirth of an old drug. J Clin
Oncol. 2009;27:1492-1501.
14. Kahl BS, Bartlett NL, Leonard JP, et al. Bendamustine is effective
therapy in patients with rituximab-refractory, indolent B-cell non-Hodgkin
lymphoma: results from a Multicenter Study. Cancer. 2010;116:106-114.
15. Friedberg JW, Cohen P, Chen L, et al. Bendamustine in patients with
rituximab-refractory indolent and transformed non-Hodgkins lymphoma:
results from a phase II multicenter, single-agent study. J Clin Oncol.
2008;26:204-210.
16. Rummel MJ, Al-Batran SE, Kim SZ, et al. Bendamustine plus rituximab is effective and has a favorable toxicity profile in the treatment of mantle
cell and low-grade non-Hodgkins lymphoma. J Clin Oncol. 2005;23:33833389.
17. Rummel MJ, Niederle N, Maschmeyer G, et al. Bendamustine plus
rituximab is superior in respect of progression free survival and CR rate when
compared to CHOP plus rituximab as first-line treatment of patients with
advanced follicular, indolent, and mantle cell lymphomas: final results of a
randomized phase III study of the StiL (Study Group Indolent Lymphomas,
Germany). Blood. 2009;114:405A.
18. Morschhauser F, Radford J, Van Hoof A, et al. Phase III trial of
consolidation therapy with yttrium-90-ibritumomab tiuxetan compared with
no additional therapy after first remission in advanced follicular lymphoma.
J Clin Oncol. 2008;26:5156-5164.
19. Hagenbeek A, Radford J, Van Hoof A, et al. 90Y-ibritumomab tiuxetan
(Zevalin) consolidation of first remission in advanced-stage follicular nonHodgkins lymphoma: updated results after a median follow-up of 66.2
months from the international, randomized, phase III First-Line Indolent
Trial (FIT) in 414 patients. Blood. 2010;116:594A.
20. Hainsworth JD, Spigel DR, Markus TM, et al. Rituximab plus shortduration chemotherapy followed by Yttrium-90 Ibritumomab tiuxetan as
first-line treatment for patients with follicular non-Hodgkin lymphoma: a
phase II trial of the Sarah Cannon Oncology Research Consortium. Clin
Lymphoma Myeloma. 2009;9:223-228.
21. Jacobs SA, Swerdlow SH, Kant J, et al. Phase II trial of short-course
CHOP-R followed by 90Y-ibritumomab tiuxetan and extended rituximab in
previously untreated follicular lymphoma. Clin Cancer Res. 2008;14:70887094.
22. Press OW, Unger JM, Braziel RM, et al. Phase II trial of CHOP
chemotherapy followed by tositumomab/iodine I-131 tositumomab for previ-

ously untreated follicular non-Hodgkins lymphoma: five-year follow-up of
Southwest Oncology Group Protocol S9911. J Clin Oncol. 2006;24:4143-4149.
23. Press OW, Unger JM, Braziel RM, et al. A phase 2 trial of CHOP
chemotherapy followed by tositumomab/iodine I 131 tositumomab for previously untreated follicular non-Hodgkin lymphoma: Southwest Oncology
Group Protocol S9911. Blood. 2003;102:1606-1612.
24. Link BK, Martin P, Kaminski MS, et al. Cyclophosphamide, vincristine, and prednisone followed by tositumomab and iodine-131-tositumomab in
patients with untreated low-grade follicular lymphoma: eight-year follow-up
of a multicenter phase II study. J Clin Oncol. 2010;28:3035-3041.
25. Press OW, Unger JM, Rimsza L, et al. A phase III randomized
intergroup trial (SWOG S0016) of CHOP chemotherapy plus rituximab vs.
CHOP chemotherapy plus iodine-131-tositumomab for the treatment of newly
diagnosed follicular non-Hodgkins lymphoma. Blood. 2011;118:98A.
26. Trotman J, Fournier M, Lamy T, et al. Positron emission tomographycomputed tomography (PET-CT) after induction therapy is highly predictive
of patient outcome in follicular lymphoma: analysis of PET-CT in a subset of
PRIMA trial participants. J Clin Oncol. 2011;29:3194-3200.
27. Fowler N, Kahl BS, Lee P, et al. Bortezomib, bendamustine, and
rituximab in patients with relapsed or refractory follicular lymphoma: the
phase II VERTICAL study. J Clin Oncol. 2011;29:3389-3395.
28. Friedberg JW, Vose JM, Kelly JL, et al. The combination of bendamustine, bortezomib, and rituximab for patients with relapsed/refractory indolent
and mantle cell non-Hodgkin lymphoma. Blood. 2011;117:2807-2812.
29. Lenz G, Staudt LM. Aggressive lymphomas. N Engl J Med. 2010;362:
1417-1429.
30. Turner M, Schweighoffer E, Colucci F, et al. Tyrosine kinase SYK:
essential functions for immunoreceptor signalling. Immunol Today. 2000;21:
148-154.
31. Chen L, Monti S, Juszczynski P, et al. SYK-dependent tonic B-cell
receptor signaling is a rational treatment target in diffuse large B-cell
lymphoma. Blood. 2008;111:2230-2237.
32. Gururajan M, Jennings CD, Bondada S. Cutting edge: constitutive B
cell receptor signaling is critical for basal growth of B lymphoma. J Immunol.
2006;176:5715-5719.
33. Friedberg JW, Sharman J, Sweetenham J, et al. Inhibition of Syk with
fostamatinib disodium has significant clinical activity in non-Hodgkin lymphoma and chronic lymphocytic leukemia. Blood. 2008;115:2578-2585.
34. Kuppers R. Mechanisms of B-cell lymphoma pathogenesis. Nat Rev
Cancer. 2005;5:251-262.
35. Ponader S, Chen SS, Buggy JJ, et al. Brutons tyrosine kinase inhibitor
PCI-32765 thwarts chronic lymphocytic leukemia cell survival and tissue
homing in vitro and in vivo. Blood. 2012;119:1182-1189.
36. Chang BY, Huang MM, Francesco M, et al. The Bruton tyrosine kinase
inhibitor PCI-32765 ameliorates autoimmune arthritis by inhibition of multiple effector cells. Arthritis Res Ther. 2011;13:R115.
37. Honigberg LA, Smith AM, Sirisawad M, et al. The Bruton tyrosine
kinase inhibitor PCI-32765 blocks B-cell activation and is efficacious in
models of autoimmune disease and B-cell malignancy. Proc Natl Acad Sci U
S A. 2010;107:13075-13080
38. Herman SE, Gordon AL, Hertlein E, et al. Bruton tyrosine kinase
represents a promising therapeutic target for treatment of chronic lymphocytic leukemia and is effectively targeted by PCI-32765. Blood. 2011117:62876296.
39. Fowler N, Sharman J, Smith SM, et al. The Btk inhibitor, PCI-32765,
induces durable responses with minimal toxicity in patients with relapsed/
refractory B-cell malignancies: results from a phase I study. Blood. 2010;116:
964A.
40. Courtney KD, Corcoran RB, Engelman JA. The PI3K pathway as drug
target in human cancer. J Clin Oncol. 2010;28:1075-1083.
41. Lannutti BJ, Meadows SA, Herman SE, et al. CAL-101, a p110delta
selective phosphatidylinositol-3-kinase inhibitor for the treatment of B-cell
malignancies, inhibits PI3K signaling and cellular viability. Blood. 2011;117:
591-594.
42. Furman R, Byrd JC, Brown JR, et al. CAL-101, An isoform-selective
inhibitor of phosphatidylinositol 3-kinase p110, demonstrates clinical activity
and pharmacodynamic effects in patients with relapsed or refractory chronic
lymphocytic leukemia. Blood. 2011;116:55A.
43. Kahl BS, Byrd JC, Flinn IW, et al. Clinical safety and activity in a
phase 1 study of CAL-101, an isoform-selective inhibitor of phosphatidylinositol 3-kinase P110, in patients with relapsed or refractory non-Hodgkin
lymphoma. Blood. 2010;116:1777A.
44. Webb HK, Chen H, Yu AS, et al. Clinical pharmacokinetics of CAL-101,
a p110 isoform-selective PI3K inhibitor, following single- and multiple-dose
administration in healthy volunteers and patients with hematological malignancies. Blood. 2010;116:1774A.
45. De Vos S, Schreeder MT, Flinn IW, et al. A phase 1 study of the
selective phosphatidylinositol 3-kinase-delta (PI3K) inhibitor, Cal-101 (GS1101), in combination with rituximab and/or bendamustine in patients with
previously treated, indolent non-Hodgkin lymphoma (iNHL). Blood. 2011;
118:2699A.
46. Smith SM. Clinical development of mTOR inhibitors: a focus on
lymphoma. Rev Recent Clin Trials. 2007;2:103-110.
47. Smith SM, van Besien K, Karrison T, et al. Temsirolimus has activity
in non-mantle cell non-Hodgkins lymphoma subtypes: The University of
Chicago phase II consortium. J Clin Oncol. 2010;28:4740-4746.
48. Pro B, Leber B, Smith M, et al. Phase II multicenter study of
oblimersen sodium, a Bcl-2 antisense oligonucleotide, in combination with
rituximab in patients with recurrent B-cell non-Hodgkin lymphoma. Br J
Haematol. 2008;143:355-360.
49. Waters JS, Webb A, Cunningham D, et al. Phase I clinical and
pharmacokinetic study of bcl-2 antisense oligonucleotide therapy in patients
with non-Hodgkins lymphoma. J Clin Oncol. 2000;18:1812-1823.
50. Leonard JP, Martin P. Novel agents for follicular lymphoma. Hematology Am Soc Hematol Educ Program. 2010;2010:259-264.
487
What Is the Best First-Line Treatment Strategy

for Patients with Indolent Lymphomas?
By Gilles Salles, MD, PhD, Herve Ghesquie` res, MD, and Emmanuel Bachy, MD
Overview: Although advanced follicular lymphoma is considered incurable, patient outcomes have improved over the last
decade with the use of anti-CD20 monoclonal antibodies.
Multiple treatment options are available and their use depends
on clinical presentation (i.e., Ann Arbor stage, tumor burden,
symptoms) and patient condition and age. Radiation therapy
for patients with limited stage disease remains useful, although its use in the era of anti-CD20 antibodies should be
re-evaluated. Single-agent rituximab has been tested in multiple studies with patients with low tumor burden. Short
treatment duration provides a response lasting 2 to 3 years,
although the benefit of maintenance therapy with rituximab
after induction therapy with rituximab remains unproven.
When watchful waiting is not an option, a combination of
HE CLINICAL outcome of patients with indolent lymphoma has markedly improved over the last decade.
Recent epidemiologic data have estimated that the 5- and
10-year overall survival (OS) rates for patients with indolent
lymphoma older than age 60 are close to 85% and 73%,
respectively.1 For patients with follicular lymphoma, several
comparisons from single-center and cooperative-group studies indicate that the median OS has increased from 8 to 10
years to 12 to 15 years2,3
This progress has been achieved, at least partly, because
of the introduction of anti-CD20 monoclonal antibodies.
However, the lack of clinical or biologic criteria either to plan
the optimal time to initiate therapy or to select between
using anti-CD20 monoclonal antibodies as single agents or
in combination with chemotherapy likely explain the heterogeneity of first-line treatment decisions observed in the
LymphoCare study.4 Furthermore, several anti-CD20 antibodies have been developed, including naked antibodies,
such as rituximab and ofatumumab, and radiolabeled antibodies, such as tositumomab and ibritumomab tiutexan,
yet the optimal chemotherapy regimen remains undefined.
With these multiple treatment options, it is worth examining the recent and follow-up results of studies performed
to help guide clinical decisions in the management of patients with follicular lymphoma. Because indolent lymphomas remain incurable and most patients experience disease
recurrence, patient quality of life, the ability to deliver
subsequent treatments, and potential long-term adverse
effects also need to be taken into account when considering
first-line treatment strategies.
Although treatment algorithms used for patients with
disseminated forms of mucosa-associated lymphoid tissue
(MALT) or nonsplenic marginal zone lymphomas can be
similar to those used for patients with follicular lymphoma,
first-line management of localized MALT5 and patients with
lymphocytic and lymphoplamasmacytic lymphomas6 are
quite distinct and will not be addressed in this manuscript.
We will consider several questions on first-line treatment for
patients with follicular lymphoma in light of the most recent
studies.
488
rituximab with chemotherapy is the standard of care: alkylating agents with anthracycline or bendamustine appear to be
the most widely used regimens, but alkylating agents alone
may still be used in selected patients subgroups. The toxicity
of regimens containing fludarabine appears to limit their
indication as first-line treatment. In patients responding to
one of these combinations, consolidation therapy with rituximab maintenance has been shown to prolong progressionfree survival with acceptable toxicity. The benefit of radioimmunotherapy in first-line treatment is still uncertain. With
patients surviving for many years, the therapeutic strategy of
first-line management should weigh the quality and duration of
response against the risk of long-term toxicities.
Is Radiation Therapy for Patients with Limited-Stage

Follicular Lymphoma Still an Option in 2012?
Radiation therapy has long been considered the treatment

of choice for patients with follicular (or indolent) lymphoma
with Ann Arbor stage I or stage II disease. This option has
been promoted as potentially curative, although the disease
might recur in areas outside the radiation fields in most
patients.7,8 Given the potential toxicity associated with
radiation therapy in specific areas, a watchful waiting approach has been also proposed.9 A large epidemiologic study
supports the use of radiation therapy in patients with
follicular lymphoma, with a significant improvement in
long-term OS for patients with stages I and II follicular
lymphoma treated with radiation therapy compared with
those not receiving radiation therapy (p 0.0001).10 Yet,
this study has several limitations, including the lack of
details about certain prognostic factors (e.g., lactate dehydrogenase [LDH], tumor bulk) and the observation period
(1973 to 2004) when monoclonal antibodies where not an
option for treatment. Another recent retrospective study11
including patients with stage I disease reported no difference in the progression-free survival (PFS) for patients
treated with radiation therapy compared with those that
were untreated (i.e., watchful waiting), but chemotherapy
with rituximab or combined modalities were found to provide the best outcome. This suggests that radiation therapy
use should be limited to those patients that have localized
follicular lymphoma (i.e., stage I or confluent stage II)
without adverse features (e.g., grade 3, tumor bulk, elevated
LDH).12,13 For localized follicular lymphoma, a dose of 24 Gy
appears sufficient.14
From the Hospices Civils de Lyon & Universite Lyon 1, Pierre-Benite, France; Centre Leon
Berard, Lyon, France.
Address reprint requests to Gilles Salles, MD, PhD, Centre Hospitalier Lyon-Sud, 69495
Pierre Benite, France; email: gilles.salles@chu-lyon.fr.
1092-9118/10/1-10
TREATMENT OPTIONS IN INDOLENT LYMPHOMA

Is Early Intervention Preferable to Watchful Waiting
in 2012?
The safety of delaying the start of systemic therapy in

patients with advanced disease but a limited tumor burden
(based on Groupe dEtude des Lymphomes Folliculaires
[GELF] or British National Lymphoma Investigation
[BNLI] criteria, Table 1) was well documented in the 1990s,
and three randomized trials reported comparable survival
outcomes for patients treated immediately or followed
through watchful waiting.15-17
The low toxicity profile of rituximab has prompted the
investigation of short treatments of this antibody as a
single-agent treatment for patients with follicular lymphoma without aggressive features.18-20 In phase II studies,
median times to disease progression have ranged between 2
and 3 years. The long-term follow-up in one study where
patients received 4 weekly infusions followed by 4 additional
infusions administered every 2 months suggested potential
disease control in 45% of patients at 8 years.21
These promising results prompted the design of a large
phase III study in the United Kingdom that compared
rituximab with watchful waiting.22 At 3 years, compared
with the control arm, patients receiving 4 weekly courses of
rituximab induction therapy followed by 2 years of rituximab maintenance therapy every 2 months showed significant improvement regarding the time to initiation of the
next treatment (primary endpoint, 91% compared with 48%,
p 0.001) and for time to progression (81% compared with
33%, p 0.001).22 However, no major difference in quality of
life was observed and OS was identical in each study arm.23
Of note, treatment with single-agent rituximab was not
available for those patients (even in the control arm) who
then received cytotoxic therapy when their disease pro-
KEY POINTS
Decisions to initiate treatment and to opt for a

specific therapeutic strategy should take into account
tumor burden, symptoms, and patient wishes and
condition as well as the potential long-term effects of
the treatment.
New studies should investigate the role of radiation
therapy for localized disease in the context of antiCD20 antibodies.
Watchful waiting remains an option for selected patients with low tumor burden; the sustained benefit
of prolonged rituximab maintenance therapy after a
short-term induction is not established in these patients.
In patients in need of cytotoxic therapy, the different
chemotherapy regimens available in combination
with rituximab may have distinct patterns of efficacy,
in term of response and progression-free survival,
and short- and long-term toxicities.
In patients responding to first-line immunochemotherapy, administration of maintenance therapy with
rituximab improves progression-free survival, but the
role of radioimmunotherapy in this setting remains
uncertain.
Table 1. Tumor Burden Criteria Used to Initiate a Cytotoxic

Treatment for Patients with Follicular Lymphoma
GELF criteria (FL2000 and PRIMA studies)
(any one of these criteria) (29)
BNLI criteria
(any one of these criteria) (17)
High tumor bulk defined by either:

- a tumor 7 cm
- 3 nodes in 3 distinct areas each
3 cm
- symptomatic splenic enlargement
- organ compression
- ascites or pleural effusion
Presence of systemic symptoms
ECOG Performance status 1
Serum LDH or beta2-microglobulin
above normal values
Rapid generalized disease progression in

the preceding 3 months
Life threatening organ involvement
Renal or macroscopic liver infiltration
Bone lesions
Presence of systemic symptoms or pruritus

Hemoglobin 10 g/dL or
WBC 3.0 109/L or
platelet counts 100 109/L;
related to marrow involvement
Abbreviations: BNLI, British National Lymphoma Investigation; ECOG, Eastern

Cooperative Oncology Group; GELF, Groupe dEtude des Lymphomes Folliculaires; LDH, lactate dehydrogenase; WBC, white blood cells.
gressed. Results from the recent ECOG E4402 RESORT

study have also challenged the benefit of prolonged rituximab treatment in a similar group of patients.24 In this trial,
the administration of maintenance therapy with rituximab
(one infusion every 3 months) for patients responding to 4
weekly rituximab infusions was compared to rituximab
retreatment at the time of progression. The primary endpoint, defined as time to treatment failure (i.e., initiation of
cytotoxic therapy or resistance to rituximab) was not different in the two study arms, although the costs and toxicities
associated with rituximab maintenance therapy appear to
be higher. Investigators in the United Kingdom study prematurely closed an experimental arm with only the 4 weekly
infusions, and the 84 patients receiving this limited intervention appeared to have an intermediate outcome (80% not
requiring treatment at 3 years, and 60% PFS) between those
of the observation arm and those receiving rituximab maintenance therapy.22
From the most current data, the strategy of delaying the
initiation of systemic therapy in patients with follicular
lymphoma still remains a justified option in 2012. Close
follow-up both allows the identification of patients that are
experiencing symptoms or rapid disease progression, and
the ability to start delivering a combination of anti-CD20
and chemotherapy, which has been shown to improve OS in
patients requiring immediate intervention. If physicians
and patients choose an earlier intervention with rituximab
as a single agent, the duration of treatment should be
limited (e.g., 4 weekly infusions, which could be eventually
completed with 4 additional infusions).
Other forms of short-treatment interventions, such as
short courses of chemotherapy with rituximab combined
with radioimmunotherapy, have also been investigated in
phase II studies.25-27 These studies included a variable
proportion of patients with characteristics similar to those
usually selected for watchful waiting. Given the heterogeneity of the patient population, the noncomparative nature of
these studies, and the lack of information about long-term
toxicities (see below), such abbreviated treatment approaches cannot be regarded as standard first-line therapy
today.
489
SALLES, GHESQUIE`RES, AND BACHY

Table 2. Studies Combining Chemotherapy and Monoclonal Antibodies in Patients with Follicular Lymphoma
Reference
Marcus, et al. (28)

Salles, et al. (29)
Rummel, et al. (33)
Chemotherapy
regimen
Median age,
y
Median follow-up,
months
PFS
R-CVP
R-CVP
R-CHOP
R-CHOP
R-Bendamustine
162
268
881
140
139
52
57.5
56
60
60
53
42
42
34
34
34 mo (median)
53%a (3.5-y from registration)
66.5%a (3.5-y after randomization)
48 mo (median) 55% (estimated PFS at 3 y)
median not reached 68% (estimated PFS at 3 y)
Abbreviations: mo, months; PFS, progression-free survival; R-bendamustine, regimen of rituximab/bendamustine; R-CHOP, regimen of rituximab/cyclophosphamide/
doxorubicin/vincristine/prednisone; R-CVP, regimen of rituximab/cyclophosphamide/vincristine/prednisone.
a
Only responders to induction therapy were randomized.
Is There a Superior Chemotherapy Induction Regimen

for Patients Requiring Immediate Intervention?
In patients requiring treatment because of disease characteristics at diagnosis or because of disease progression
after watchful waiting, the combination of rituximab plus
chemotherapy has become the usual standard of care (Table
2). Indeed, several randomized studies have demonstrated
that this combination improves the OS of patients with
follicular lymphoma, and a meta-analysis has indicated a
significant reduction in the risk of death (HR for mortality
0.63; 95% CI, 0.51 to 0.79). Different chemotherapy regimens (e.g., CVP [cyclophosphamide, vincristine, prednisone], CHOP [cyclophosphamide, doxorubicin, vincristine,
prednisone], fludarabine combinations, bendamustine) associated with rituximab are available (e.g., R-CVP, R-CHOP,
R-bendamustine, etc.) and the debate about their potential
benefit remains unsettled.
The CVP regimen has resulted in lower response rates
(around 85%) and shorter times to disease progression,28
even when maintenance therapy with rituximab is delivered
to patients that have responded to treatment.29 The only
randomized study comparing these regimens did not show a
difference in OS between R-CVP and R-CHOP.30 But in
patients with adverse features (such as a high score on the
Follicular Lymphoma International Prognostic Index
[FLIPI]), the median times to progression were rather short
(26 months),28 and recent survey data suggest that patients
with high-risk disease receiving R-CVP might have reduced
survival compared with those receiving R-CHOP.31 However, given the potential long-term cardiac toxicity of anthracyclines and their eventual benefit as second-line
treatment, some physicians might defer the use of R-CHOP
in first-line therapy.
Although developed in the last 15 years, a regimen containing fludarabine in combination with mitoxantrone was
recently compared with R-CVP and R-CHOP in a randomized study.30 If PFS appeared favorable, lower OS was
observed in the fludarabine arm, which was in line with data
from the PRIMA study with R-FCM (rituximab plus fludarabine, cyclophosphamide, and mitoxantrone).32 Of note, this
lower survival appeared to be partly related to the toxicities
of regimens containing fludarabine, underscoring the need
to consider short- and long-term adverse events.
Regarding bendamustine, one randomized study demonstrated a better response rate and prolonged time to progression with the use of R-bendamustine as compared to
R-CHOP (HR 0.63, p 0.028), but OS was not significantly
different between groups.33 In this study, the short-term
toxicity profile of R-bendamustine was found to be much
better than that observed with R-CHOP. However, detailed
and long-term data of this study are still unavailable.
490
Another randomized study (NCT00877006) comparing

R-bendamustine to R-CHOP or R-CVP has recently completed accrual.
When considering the choice of regimen, both short- and
long-term toxicities should be considered, as well as patient
age and comorbidities. The incidence of patients with
changes in histology at progression may also be worth
investigating. In patients with a median survival exceeding
10 years, it becomes essential to develop comparative studies examining long-term outcome, to ensure that the use of a
given regimen does not preclude or increase toxicities associated with efficient second-line therapy.
What Is the Role of Anti-CD20 Maintenance Therapy
after Induction Chemotherapy?
Because most patients experience disease recurrence

within 3 to 5 years after chemotherapy with rituximab,
attempts to prolong efficacy have been investigated. The
PRIMA studies evaluated the addition of 2 years of rituximab maintenance therapy (i.e., one infusion every 2
months) in patients with disease that responded to R-CVP,
R-CHOP, or R-FCM.29 The interim analysis demonstrated a
significant reduction in the risk of lymphoma progression for
those patients receiving maintenance therapy (HR 0.5;
95% CI, 0.39 to 0.64, p 0.0001). At 4 years, PFS was still
significantly better in the experimental arm than in the
control arm (68% vs. 50%, respectively, p 0.001) (Salles G,
personal results, Table 3). The benefit of rituximab maintenance therapy was observed regardless of patient age,
the quality of response after induction therapy (complete
response/unconfirmed complete response compared with
partial response), or FLIPI category (Fig. 1). During maintenance therapy with rituximab, patients experienced more
frequent adverse events, especially grade 2 infections, but
very few patients withdrew from the study for treatmentrelated toxicities. No differences in OS were observed between the two study arms.29 A meta-analysis including
2,586 patients from nine trials indicated that the use of
rituximab maintenance therapy could improve OS, although
this benefit was not significant for patients receiving rituximab maintenance therapy after first-line treatment (HR
0.86; 95% CI, 0.60 to 1.25).34 A large study is currently
comparing the optimal duration of rituximab maintenance
(2 years compared with 4 years) administered after
R-bendamustine (NCT00877214).
What Is the Role of Radioimmunotherapy
in the First-Line Treatment of Patients
with Follicular Lymphoma?
Radioimmunotherapy has been investigated as singleagent and consolidation therapy after chemotherapy. Tosi-

Table 3. Studies Investigating an Anti-CD20 Antibody Consolidation or Maintenance Therapy after Chemotherapy with Rituximab
Reference
Induction txa
Hagenbeek et al (37)
Salles et al (29)
Salles et al (29)
Press et al (38)
Chemotherapyb
R-CVP
R-CHOP
R-CHOP
CHOP
Intervention
90
Y ibritumomab tiutexan
rituximab maintenance
(12 infusions/2 y)
Observation
131
I tositumomab
PFS
n per arm
(control/intervention)
Median
follow-up
HR (95% CI)
Control
Intervention
202/207
113/109
386/382
265/267
66 mo
48 mo
48 mo
4.9 y
0.51 (0.40,66)
0.71 (0.481.04)
0.49 (0.390,62)
NA
15 mo (median)
47% (4-y)
50% (4-y)
76% (2-y)
49 mo (median)
60% (4-y)
70% (4-y)
80% (2-y)
Abbreviations: CHOP, regimen of cyclophosphamide/doxorubicin/vincristine/prednisone CI, confidence interval; HR, hazard ratio; mo, months; NA, not available; PFS,
progression-free survival; R-CHOP, regimen of rituximab/cyclophosphamide/doxorubicin/vincristine/prednisone; R-CVP, regimen of rituximab/cyclophosphamide/
vincristine/prednisone; tx, treatment; y, year.
a
Before randomization.
b
Chemotherapy included anthracycline in 43% of patients and rituximab in 15%.
tumomab administered as single-agent therapy in 76

patients provided a high response rate (95%) and the 5-year
PFS was 59%.35 Unfortunately, this approach was never
evaluated in a randomized study against other forms of
therapy. Two randomized studies (Table 3) addressed the
role of radioimmunotherapy as consolidation therapy in
patient responders to chemotherapy. In the FIT study,36,37
administration of a single course of ibritumomab tiutexan
resulted in high rates of complete response (87.4% with
either complete or unconfirmed complete response) and
significant improvements in PFS (median 15 months in the
observation arm compared with 49 months in the experimental arm). OS was not different between the two groups
but six myeloid malignancies were observed in patients that
had received radioimmunotherapy, compared with one in
the control arm (p 0.0001).37 The major pitfall of this
study was the fact that only 15% of patients had received
rituximab with induction chemotherapy. The results of the
SWOG0016 Intergroup study comparing R-CHOP (267 patients) with CHOP followed by tositumomab (265 patients)
have been recently presented.38 With a median follow-up of
4.9 years, 2-year PFS and OS were not significantly different
between the groups (80% and 93%, respectively, in the
radioimmunotherapy arm, 76% and 97%, respectively, in the
R-CHOP arm). A slight but nonsignificant excess of secondary myeloid malignancies was observed in patients in the
radioimmunotherapy arm.38
Based on these results, it is therefore difficult to establish
the role of radioimmunotherapy in the first-line treatment of
patients with follicular lymphoma. Studies assessing the
role of consolidation therapy with radiolabeled antibodies
against rituximab maintenance in the context of induction

chemotherapy with rituximab should be conducted.
Is There a Role for High-Dose Therapy Supported
with Stem Cell Transplantation?
Although this strategy was investigated before the rituximab era with some indication of prolonged PFS in at least
two studies,39 autologous stem cell transplantation does not
seem to improve OS.40 This strategy therefore does not
represent a standard option as consolidation therapy for
patients with responsive disease in first-line settings.
Future Prospects for the Management of Indolent
Lymphoma in First-Line Therapy
Although the results achieved with the combination of

chemotherapy and maintenance or consolidation therapies
are favorable, with 5-year OS estimates exceeding 90%,
several questions remain. The first question regards the
stratification of patients, which is currently performed with
clinical criteria. It is our hope that new biologic tools will
translate recent advances in the understanding of the biology of follicular lymphoma into clinical practice and provide
useful predictors of patient outcome that will help tailor
therapy.41 A second question lies in the evaluation of treatment efficacy and the benefit of achieving complete response
to first-line therapy.42 To improve the accuracy of this
evaluation, minimal residual disease has not gained a wide
acceptance outside clinical trials given its technical limitations. But recent studies underline the value of 18Ffluorodeoxy-D-glucose positron emission tomography in
deciphering the quality of response in patients that achieved
Fig 1. Outcome of patients randomly assigned to observation or rituximab maintenance therapy in the PRIMA trial according to the Follicular
Lymphoma Prognostic Index (FLIPI).
Abbreviations: PFS, progression free survival; NA, not applicable.
491
SALLES, GHESQUIE`RES, AND BACHY
partial or incomplete response.43,44 Whether modifying

treatment for patients with disease that does not achieved
complete metabolic response will improve outcomes has yet
to be determined. Finally, the emergence of new biologic
therapies, such engineered monoclonal antibodies,45 immunomodulatory agents,46 or kinase inhibitors,47 represent
innovative options that might change the current treatment
landscape.
Author
Gilles Salles
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Calistoga
Pharmaceuticals;
Celgene;
Janssen-Cilag;
Roche
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Celgene;
Janssen-Cilag;
Pfizer; Roche
Herve Ghesquie`res*
Emmanuel Bachy*
REFERENCES
1. Pulte D, Gondos A, Brenner H. Expected long-term survival of older
patients diagnosed with non-Hodgkin lymphoma in 2008-2012. Cancer Epidemiol. 2012;36:e19-25.
2. Fisher RI, LeBlanc M, Press OW, et al. New treatment options have
changed the survival of patients with follicular lymphoma. J Clin Oncol.
2005;23:8447-8452.
3. Sebban C, Brice P, Delarue R, et al. Impact of rituximab and/or
high-dose therapy with autotransplant at time of relapse in patients with
follicular lymphoma: A GELA study. J Clin Oncol. 2008;26:3614-3620.
4. Friedberg JW, Taylor MD, Cerhan JR, et al. Follicular lymphoma in the
United States: first report of the national LymphoCare study. J Clin Oncol.
2009;27:1202-1208.
5. Bertoni F, Coiffier B, Salles G, et al. MALT lymphomas: pathogenesis
can drive treatment. Oncology. 2011;25:1134-1142, 1147.
6. Peinert S, Seymour JF. Indolent lymphomas other than follicular and
marginal zone lymphomas. Hematol Oncol Clin North Am. 2008;22:903-940.
7. Vaughan Hudson B, Vaughan Hudson G, MacLennan KA, et al. Clinical
stage 1 non-Hodgkins lymphoma: long-term follow-up of patients treated by
the British National Lymphoma Investigation with radiotherapy alone as
initial therapy. Br J Cancer. 1994;69:1088-1093.
8. Mac Manus MP, Hoppe RT. Is radiotherapy curative for stage I and II
low-grade follicular lymphoma? Results of a long-term follow-up study of
patients treated at Stanford University. J Clin Oncol. 1996;14:1282-1290.
9. Advani R, Rosenberg SA, Horning SJ. Stage I and II follicular nonHodgkins lymphoma: long-term follow-up of no initial therapy. J Clin Oncol.
2004;22:1454-1459.
10. Pugh TJ, Ballonoff A, Newman F, et al. Improved survival in patients
with early stage low-grade follicular lymphoma treated with radiation: a
Surveillance, Epidemiology, and End Results database analysis. Cancer.
2010;116:3843-3851.
11. Friedberg J, Byrtek M, Link B, et al. Should radiation therapy (Xrt) be
the standard therapeutic approach for stage I follicular lymphoma (FL)? A
comparative effectiveness analysis of the National Lymphocare Study (Nlcs).
In: Internationational Conference on Malignat Lymphoma; 2011; Lugano;
2011;p. iv91.
12. Dreyling M, Ghielmini M, Marcus R, et al. Newly diagnosed and
relapsed follicular lymphoma: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2011;22:vi59-63 (suppl 6).
13. Zelenetz AD, Abramson JS, Advani RH, et al. NCCN Clinical Practice
Guidelines in Oncology: Non-Hodgkins lymphomas. JNCCN. 2010;8:288-334.
14. Lowry L, Smith P, Qian W, et al. Reduced dose radiotherapy for local
control in non-Hodgkin lymphoma: a randomised phase III trial. Radiother
Oncol. 2011;100:86-92.
15. Young RC, Longo DL, Glatstein E, et al. The treatment of indolent
lymphomas: watchful waiting v aggressive combined modality treatment.
Semin Hematol. 1988;25:11-16 (suppl 2).
16. Brice P, Bastion Y, Lepage E, et al. Comparison in low-tumor-burden
follicular lymphomas between an initial no-treatment policy, prednimustine,
or interferon alfa: a randomized study from the Groupe dEtude des Lymphomes Folliculaires. Groupe dEtude des Lymphomes de lAdulte. J Clin
Oncol. 1997;15:1110-1117.
17. Ardeshna KM, Smith P, Norton A, al. E. Long-term effect of a watch
and wait policy versus immediate systemic treatment for asymptomatic
492
advanced-stage non-Hodgkin lymphoma: a randomised controlled trial. Lancet. 2003;362:516-522.

18. Hainsworth JD, Burris HA, 3rd, Morrissey LH, et al. Rituximab
monoclonal antibody as initial systemic therapy for patients with low-grade
non-Hodgkin lymphoma. Blood. 2000;95:3052-3056.
19. Colombat P, Salles G, Brousse N, et al. Rituximab (anti-CD20 monoclonal antibody) as single first-line therapy for patients with follicular
lymphoma with a low tumor burden: clinical and molecular evaluation. Blood.
2001;97:101-106.
20. Ghielmini M, Schmitz SF, Cogliatti S, et al. Effect of single-agent
rituximab given at the standard schedule or as prolonged treatment in
patients with mantle cell lymphoma: a study of the Swiss Group for Clinical
Cancer Res (SAKK). J Clin Oncol. 2005;23:705-711.
21. Martinelli G, Schmitz SF, Utiger U, et al. Long-term follow-up of
patients with follicular lymphoma receiving single-agent rituximab at two
different schedules in trial SAKK 35/98. J Clin Oncol. 2010;28:4480-4484.
22. Ardeshna KM, Smith P, Qian W, et al. An intergroup randomised trial
of rituximab versus a watch and wait strategy in patients with stage II, III,
IV, asymptomatic, non-bulky follicular lymphoma (grades 1, 2 and 3a). A
preliminary analysis. ASH Annual Meeting Abstracts. 2011;116 (abstr 6).
23. Ardeshna KM, Qian W, Stephens R, et al. Preliminary Results Of
Quality Of Life (Qol) Analyses From The Intergroup Phase III Randomised
Trial Of Rituximab Vs A Watch And Wait Approach In Patients With
Advanced Stage, Asymptomatic, Non-Bulky Follicular Lymphoma (FL). Ann
Oncol. 2011;22:iv88 (suppl 4).
24. Kahl BS, Hong F, Williams ME, et al. Results of Eastern Cooperative
Oncology Group Protocol E4402 (RESORT): a randomized phase III study
comparing two different rituximab dosing strategies for low tumor burden
follicular lymphoma. ASH Annual Meeting Abstracts. 2011:118 (suppl LBA6).
25. Leonard JP, Coleman M, Kostakoglu L, et al. Abbreviated chemotherapy with fludarabine followed by tositumomab and iodine I 131 tositumomab
for untreated follicular lymphoma. J Clin Oncol. 2005;23:5696-5704.
26. Zinzani PL, Tani M, Pulsoni A, et al. Fludarabine and mitoxantrone
followed by yttrium-90 ibritumomab tiuxetan in previously untreated patients with follicular non-Hodgkin lymphoma trial: a phase II nonrandomised trial (FLUMIZ). Lancet Oncol. 2008;9:352-358.
27. Link BK, Martin P, Kaminski MS, et al. Cyclophosphamide, vincristine, and prednisone followed by tositumomab and iodine-131-tositumomab in
patients with untreated low-grade follicular lymphoma: eight-year follow-up
of a multicenter phase II study. J Clin Oncol. 2010;28:3035-3041.
28. Marcus R, Imrie K, Solal-Celigny P, et al. Phase III study of R-CVP
compared with cyclophosphamide, vincristine, and prednisone alone in patients with previously untreated advanced follicular lymphoma. J Clin Oncol.
2008;26:4579-4586.
29. Salles G, Seymour JF, Offner F, et al. Rituximab maintenance for 2
years in patients with high tumour burden follicular lymphoma responding to
rituximab plus chemotherapy (PRIMA): a phase 3, randomised controlled
trial. Lancet. 2011;377:42-51.
30. Federico M, Luminari S, Dondi A, et al. R-CVP vs R-CHOP vs R-FM for
the initial treatment of patients with advanced stage follicular lymphoma.
Preliminary results of Foll05 Iil Trial. Ann Oncol. 2011;22:iv128 (suppl 4).
31. Nastoupil L, Sinha R, Byrtek M, et al. A comparison of the effectiveness

of first-line chemoimmunotherapy regimens for follicular lymphoma (FL)
used in the United States. ASH Annual Meeting Abstracts. 2011;118 (abstr
97).
32. Morschhauser F, Seymour JF, Feugier P, et al. Impact of induction
chemotherapy regimen on response, safety and outcome in the Prima study.
Ann Oncol. 2011;22:iv88 (suppl 4).
33. Rummel MJ, Niederle N, Maschmeyer G, et al. Bendamustine plus
rituximab is superior in respect of progression free survival and CR rate when
compared to CHOP plus rituximab as first-line treatment of patients with
advanced follicular, indolent, and mantle cell lymphomas: final results of a
randomized phase III study of the StiL (Study Group Indolent Lymphomas,
Germany). ASH Annual Meeting Abstracts. 2009;114 (abstr 405).
34. Vidal L, Gafter-Gvili A, Salles G, et al. Rituximab maintenance for the
treatment of patients with follicular lymphoma: an updated systematic
review and meta-analysis of randomized trials. J Natl Cancer Inst. 2011;103:
1799-1806.
35. Kaminski MS, Tuck M, Estes J, et al. 131I-tositumomab therapy as
initial treatment for follicular lymphoma. N Engl J Med. 2005;352:441-449.
36. Morschhauser F, Radford J, Van Hoof A, et al. Phase III trial of
consolidation therapy with yttrium-90-ibritumomab tiuxetan compared with
no additional therapy after first remission in advanced follicular lymphoma.
J Clin Oncol. 2008;26:5156-5164.
37. Hagenbeek A, Radford J, Van Hoof A, et al. 90Y-Ibritumomab tiuxetan
(Zevalin(R)) consolidation of first remission in advanced-stage follicular
non-Hodgkins lymphoma: updated results after a median follow-up of 66.2
months from the international, randomized, phase III First-Line Indolent
Trial (FIT) in 414 Patients. ASH Annual Meeting Abstracts. 2010;116 (abstr
594).
38. Press OW, Unger JM, Rimsza LM, et al. A phase III randomized
intergroup trial (SWOG S0016) of CHOP chemotherapy plus rituximab vs.
CHOP chemotherapy plus iodine-131-tositumomab for the treatment of newly
diagnosed follicular non-Hodgkins lymphoma. ASH Annual Meeting Abstracts. 2011;118 (abstr 98).
39. Bachy E, Salles G. Marrow-ablative treatment and autologous stem cell
transplantation in follicular NHL. Best Pract Res Clin Haematol. 2011;24:

257-270.
40. Al Khabori M, de Almeida JR, Guyatt GH, et al. Autologous stem cell
transplantation in follicular lymphoma: a systematic review and metaanalysis. J Natl Cancer Inst. 2012;104:18-28.
41. Relander T, Johnson NA, Farinha P, et al Prognostic factors in
follicular lymphoma. J Clin Oncol. 2010;28:2902-2913.
42. Bachy E, Brice P, Fournier M, et al. Long term follow-up of the GELF86
French and Belgian trials: complete remission after first line treatment with
conventional chemotherapy in newly diagnosed follicular lymphoma patients
correlates with prolonged overall survival compared with partial remission.
ASH Annual Meeting Abstracts. 2008;112 (abstr 2590).
43. Trotman J, Fournier M, Lamy T, et al. Positron emission tomographycomputed tomography (PET-CT) after induction therapy is highly predictive
of patient outcome in follicular lymphoma: analysis of PET-CT in a subset of
PRIMA trial participants. J Clin Oncol. 2011;29:3194-200.
44. Dupuis J, Meignan M, Julian A, et al. Significant prognostic impact of
[18F] fluorodeoxyglucose-PET scan performed during and at the end of
treatment with R-CHOP in high-tumor mass follicular lymphoma patients: a
GELA-GOELAMS study. ASH Annual Meeting Abstracts. 2011;118 (abstr
877).
45. Salles GA, Morschhauser F, Thieblemont C, et al. Efficacy and safety of
obinutuzumab (GA101) monotherapy in relapsed/refractory indolent nonHodgkins lymphoma: results from a phase I/II study (BO20999). ASH Annual
Meeting Abstracts. 2011;118 (abstr 268).
46. Samaniego F, Hagemeister F, Mclaughlin P, et al. High response rates
with lenalidomide plus rituximab for untreated indolent B-cell non-Hodgkin
lymphoma, including those meeting GELF criteria. J Clin Oncol. 2011;29
47. Furman RR, Byrd JC, Flinn IW, et al. Interim results from a phase I
study of CAL-101, a selective oral inhibitor of phosphatidylinositol 3-kinase
p110d isoform, in patients with relapsed or refractory hematologic malignancies. J Clin Oncol.2010; 28 (suppl; abstr 3032).
493
What Is the Role of Transplantation for

Indolent Lymphoma?
By Ryan D. Cassaday, MD, and Ajay K. Gopal, MD
Overview: Despite advances in chemoimmunotherapy, indolent B-cell non-Hodgkin lymphomas (B-NHLs) are generally
not considered curable with this approach. Much attention has
been paid to the prospect of hematopoietic cell transplantation (HCT) as a way to improve long-term outcomes for this
group of diseases. Autologous (auto) HCT provides intensive
conditioning therapy followed by rescue of hematopoiesis,
and this has been shown in randomized studies to prolong
survival compared with more standard chemotherapy, albeit
with increased short-term toxicity and the potential for higher
rates of secondary malignancies. Allogeneic (allo) HCT can
provide anticancer effects beyond the conditioning therapy
through the immune-mediated graft-versus-lymphoma (GVL)
-NHLs are often classified into indolent and aggressive

subtypes. Indolent lymphomas typically include several different specific histologies: follicular lymphoma (FL),
lymphoplasmacytic lymphoma (LPL), small lymphocytic
lymphoma (SLL), and marginal zone lymphoma (MZL).1 The
vast majority of patients present with disseminated disease
requiring systemic therapy, which can improve overall survival (OS) but is not thought to be curative. Over time,
indolent B-NHL becomes increasingly refractory to chemotherapy, with shallower, shorter remissions the norm.
HCT is a strategy that aims to overcome the limitations of
standard chemoimmunotherapy in two general forms. Auto
HCT relies on the steep dose-response curve observed in
hematopoietic malignancies by using MA doses of chemoradiotherapy followed by auto hematopoietic stem cell rescue.
In contrast, allo HCT also employs an immune-mediated
GVL effect to eradicate disease following preparative regimens ranging in intensity from high-dose MA combinations
to very low-dose (2 Gy) radiation. The less intensive conditioning strategies are designed to suppress the patients
immune system sufficiently to allow engraftment of the
donor hematopoietic cells and relies almost exclusively on
the GVL effect.2,3 However, this same immunologic phenomenon can affect normal healthy tissues in the patient,
inducing GVHDthe major source of morbidity and mortality of this approach.4 Nevertheless, allo HCT remains a
viable option for select patients as it can offer durable
remissions and the potential for cure in patients with
relapsed or refractory indolent lymphomas.
Indications for Auto HCT
Initial Consolidation
Auto HCT is generally not recommended as initial consolidation of response for any indolent B-NHL outside the
context of a clinical trial. This conclusion is supported by at
least four prospective randomized controlled trials (RCTs)
that investigated this approach as part of first-line management (Table 1).5-8 Although most of these studies show
improved progression-free survival (PFS) or event-free survival (EFS) when compared with standard chemotherapy,
overall survival (OS) has not been shown to be significantly
improved (p 0.5 in the 3 studies where this comparison
was reported). Moreover, there appeared to be an increased
494
effect. It can be administered following myeloablative (MA)

conditioning or reduced-intensity (RI) regimens aimed at sufficiently suppressing the patients immune system to allow
engraftment of donor hematopoiesis. However, this same
potentially curative alloreactivity of the engrafted immune
system can lead to graft-versus-host disease (GVHD), a significant cause of morbidity and mortality following allo HCT.
This article will discuss the current role of both auto HCT and
allo HCT in the management of indolent lymphoma as well as
the relative risks and benefits of each approach such that the
reader can place this in context of the multitude of options
available for patients with indolent B-NHL.
risk of secondary malignancies (both myeloid and solid

tumors) associated with auto HCT, which is becoming increasingly relevant in indolent lymphoma as survival has
continued to improve with modern chemoimmunotherapy
regimens. Though select retrospective analyses have implicated etoposide, alkylating agents, and total body irradiation as increasing this risk, the study by Sebban and others
used these treatments as part of the transplant arm and saw
no increased incidence of secondary cancers compared to
the non-transplant arm.8 In the trial that noted the highest
rates of secondary malignancies, Gyan and colleagues postulated that the practice of in vitro purging of B lymphocytes may have been responsible for this finding, perhaps by
affecting post-transplant immunosurveillance.5 This process
is used rarely today with the application of in vivo B-cell
purging by rituximab.
Relapsed Disease
The major role of auto HCT remains its use in patients

with relapsed but chemotherapy-sensitive indolent B-NHL.
Unfortunately, most of the data supporting this approach
come from single-arm studies suggesting that a subset of
patients can again achieve long-term remissions with
this treatment. For example, in the previously mentioned
study by Ladetto et al, patients initially treated in the
chemotherapy-only arm were allowed to cross over to auto
HCT at the time of relapse.6 Among the 28 patients salvaged
with this approach, the 3-year projected EFS and OS were
68% and 81%, respectively. Likewise, a retrospective study
from the United Kingdom in which patients received auto
HCT for FL in second or later remission showed a plateau in
the freedom from progression curve at 48% at 12 years
(median follow-up of 13.5 years), offering further evidence
that relapsed FL can be salvaged with auto HCT.9 To date,
there has been a single small, but important, prospective
From the Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle,
WA; Division of Medical Oncology, Department of Medicine, University of Washington,
Seattle, WA.
Address reprint requests to Ajay Gopal, MD, 1100 Fairview Ave. N., Mailstop G3200,
Seattle, WA 98109; email: agopal@uw.edu.
1092-9118/10/1-10
ROLE OF HCT FOR INDOLENT LYMPHOMA

Table 1. Summary of Results from Selected Randomized Controlled Trials of Autologous Hematopoietic Cell Transplantation for
Follicular Lymphoma
Study
(Sample Size)
Induction Therapy^
Consolidation Therapy^
EFS/PFS
Disease Setting: Initial Consolidation

Dexa-BEAM, then Cy/TBI-Auto HCT
CHOP or MCP 2, then IFN
Lenz et al 20047
(n 307)
CHOP or MCP 46
Sebban et al8 2006

(n 401)
CHOP 4
CHVP IFN 6
CE, then CE/TBI-Auto HCT

CHVP IFN 6
Ladetto et al6 2008

(n 136)
APO 2 DHAP 2*
CHOP 6, then R 4
CE R, then Mito/Mel-Auto HCT

None
Gyan et al5 2009

(n 166)
VCAP 23 IMV 1 or DHAP 23*

CHVP 6
TBI/Cy-Auto HCT
CHVP 6, IFN 18 mo
p
Schouten et al10 2003
(n 89)
CHOP or MIME 3
Disease Setting: Relapsed Disease

Cy/TBI-Auto HCT
CHOP or MIME 3
65%
33%
0.001
38%
28%
0.11
61%
28%
0.001
56%
39%
0.03
57%
17%
p 0.001
Secondary
Malignancies
Median
F/U
NR
NR
NR
76%
71%
p 0.53
81%
80%
p 0.96
76%
80%
p 0.55
NR
NR
NR
6%
7%
NR
8%
3%
NR
14%
1%
NR
4.2 yrs
70%
42%
p 0.026
NR
NR
NR
OS
7.7 yrs
4.2 yrs
9 yrs
5.8 yrs
Abbreviations: EFS, event-free survival; PFS, progression-free survival; OS, overall survival; F/U, follow-up; CHOP, cyclophosphamide, doxorubicin, vincristine, and
prednisone; MCP, mitoxantrone, chlorambucil, and prednisone; Dexa-BEAM, dexamethasone, carmustine, etoposide, cytarabine, and melphalan; Cy, cyclophosphamide; TBI, total body irradiation; auto, autologous; HCT, hematopoietic cell transplantation; IFN, interferon-alpha; NR, not reported; yrs, years; CE, cyclophosphamide
and etoposide; CHVP, cyclophosphamide, doxorubicin, teniposide, and prednisolone; APO, doxorubicin, prednisone, and vincristine; DHAP, dexamethasone,
cytarabine, and cisplatin; R, rituximab; mito, mitoxantrone; mel, melphalan; VCAP, vindesine, cyclophosphamide, doxorubicin, and prednisone; IMV, ifosfamide,
methotrexate, and etoposide; MIME, mesna, ifosfamide, mitoxantrone, and etoposide.
^
The numbers following the chemotherapy regimen abbreviations denote the number of cycles administered.
* In these studies, DHAP was given only to subjects who did not achieve a sufficient response to the initial chemotherapy (in the study by Gyan et al 5 , DHAP was given
instead of IMV).
RCT that evaluated auto HCT for relapsed, chemotherapysensitive FL showing a significant benefit over standard
chemotherapy in both PFS (p 0.001) and OS (p 0.026)
with nearly 6 years of follow-up (Table 1).10
Therefore, these data suggest that for most patients with
FL, auto HCT should be considered part of the treatment of
relapse. Outside of the previously mentioned studies focusing on FL, there are little data describing the utility of auto
KEY POINTS
Indolent lymphomas are not considered curable with

standard chemotherapy approaches, but hematopoietic cell transplantation (HCT) can be considered for
selected patients with relapsed or refractory indolent
lymphoma.
In randomized studies compared with standard nontransplant approaches, autologous HCT as initial
consolidation has shown improved progression-free
(PFS) survival but not significantly better overall
survival (OS).
Autologous transplantation for relapsed chemotherapy-sensitive indolent lymphoma has been
shown to improve both PFS and OS in a single phase
III trial.
Allogeneic HCT can offer durable remissions even for
patients who have relapsed after autologous HCT.
The use of reduced-intensity conditioning before allogeneic HCT may mitigate some of the treatmentrelated toxicity, but graft-versus-host disease
remains a significant cause of morbidity and mortality with this approach.
HCT in other indolent B-NHL. Approximately 11% of the

patients enrolled in the RCT from the German Low Grade
Lymphoma Study Group had LPL: this subgroup had a
5-year PFS of 65% after auto HCT compared with 73% with
maintenance interferon-alpha (p 0.98).7 The National
Comprehensive Cancer Network (NCCN) recommends considering the use of auto HCT to consolidate a second or later
remission for FL and MZL and for salvage treatment of
LPL.11 Beyond the NCCN guidelines, the Fourth International Workshop on Waldenstroms Macroglobulinemia reviewed the available data and generated treatment
recommendations for LPL.12 These guidelines go into
slightly more detail regarding the potential role for HCT,
but this groups conclusions were similar to that of NCCN,
stating that prospective trials are needed. Open questions
in the field of auto HCT for indolent B-NHL include the
optimal conditioning regimen (chemotherapy alone compared with radiation plus chemotherapy) and the role of
high-dose therapy in patients with rituximab-refractory
disease. This latter point is particularly interesting, given
that this agent was not widely available at the time that
most of the above-noted prospective studies began. The
study by Ladetto and others showed no improvement in OS
with initial consolidative auto HCT when both groups received rituximab, but did not address the issue of postrituximab relapse.6 One could hypothesize that intensive
chemoradiotherapy may have a greater relative benefit in
rituximab-refractory, yet chemosensitive indolent lymphoma, though prospective randomized trials evaluating
this specific scenario are needed to confirm this premise.
Indications for Allo HCT
Although allo HCT has an established role in the management of myeloid neoplasms, its exact place in the treatment
of lymphoid malignancy is less clear. As stated above, the
495
CASSADAY AND GOPAL
potential for GVL-mediated disease eradication does make it

an attractive option for patients with increasingly shorter
remission durations or young patients with relapsed disease
where the likelihood of decades of disease control with
currently available standard therapies is low. The associated toxicities (particularly GVHD) and the limitations
imposed by finding human leukocyte antigen (HLA)matched donors reduce the number of potential candidates
for this treatment. For this reason, the NCCN guidelines
reserve allo HCT for highly selected patients beyond second
relapse.11 This section will discuss the literature available
that compares auto HCT with allo HCT, the use of MA or RI
conditioning, and options for treating relapse after allo HCT.
Auto HCT Compared with Allo HCT
There have been no sufficiently powered RCTs addressing

the relative benefits of auto compared with allo HCT for
indolent lymphoma. There have been several analyses comparing outcomes of these two approaches, but they are
obviously limited by their nonrandomized and typically
retrospective design. For example, allo HCT is reserved for
later in the treatment course (often having relapsed after
auto HCT), so this generally selects for more advanced or
chemorefractory disease. Alternatively, because it is understood that allo HCT is a higher-risk treatment, it may not be
offered to patients felt to be more frail. It is important to
appreciate such biases when assessing studies of this type,
but several potentially useful themes do emerge when reviewing the results.
The largest auto HCT/MA allo HCT comparison included
904 patients with FL from the International Bone Marrow
Transplant Registry/Autologous Bone Marrow Transplant
Registry (IBMTR/ABMTR).13 These were patients who received transplantation during the 1990s, before the widespread use of rituximab and RI allo HCT. The patients
receiving allo HCT were more likely to have worse performance status, abnormal lactate dehydrogenase (LDH) levels, advanced disease, and chemotherapy-resistant disease.
As mentioned above, the relative risk of TRM with allo HCT
was significantly higher compared with auto HCT (p
0.001), while the relative risk of relapse was significantly
lower with allo HCT (p 0.03). This yielded similar OS
between the groups, though there were more long-term
disease-free survivors with few late relapses in the allo HCT
cohort. Similar findings were reported in a smaller retrospective analysis from investigators in the United Kingdom.14 Interestingly, both studies suggested lower rates of
secondary malignancies in the allo HCT group potentially
because of the allogeneic graft replacing damaged host
hematopoiesis; though confounding factorssuch as the use
total-body irradiation conditioning regimens could have
contributed to these differences.13,14
A prospective cohort study of auto HCT compared with
allo HCT in 216 patients with FL from the NCCN Outcomes
Database Project was recently presented in abstract form.15
Importantly, these patients were all treated in the postrituximab era. Additionally, the patients receiving allo HCT
received a mix of MA and RI conditioning, though the groups
were reported in aggregate. These investigators found that
allo HCT recipients were generally younger and more heavily pretreated than those in the auto HCT cohort. With a
median follow-up of 2.9 years, they noted an overall nonrelapse mortality (NRM) rate of 10% with auto HCT compared
496
with 33% with allo HCT (p 0.0001). The 3-year estimate of

failure-free survival (with failure defined as relapse, transformation, disease progression, or death) was not different
between the two groups (p 0.3), with a rate of 55% for auto
HCT compared with 56% for allo HCT. However, the 3-year
estimate of OS was significantly different (p 0.001), with
a rate of 85% for auto HCT compared with 64% for allo HCT.
Using a multivariate analysis to adjust for age, number of
prior therapies, and disease status at the time of HCT, allo
HCT was still associated with an increased risk of death
(hazard ratio [HR] 2.2, p 0.01). Taken together, these
studies demonstrate the potential to control relapsed or
refractory disease with allo HCT, but the relative toxicities
compared with that of auto HCT make it a less attractive
option earlier in the disease course.
Lastly, the Bone Marrow Transplant Clinical Trials Network performed a prospective study comparing auto HCT
with RI allo HCT in patients with relapsed but
chemotherapy-sensitive FL, but it closed early because of
slow accrual.16 This study used a biologic treatment assignment, where patients with an HLA-identical sibling received
RI allo HCT, and those without received auto HCT. Because
it closed early, statistical comparisons between the groups
could not be performed. Nevertheless, they did find that the
3-year OS and PFS for auto HCT was 73% and 63% (respectively; 20 patients), compared with 100% and 86% (respectively; 7 patients) for RI allo HCT. Although firm conclusions
cannot be drawn from this study, these data do suggest at
least comparable short-term efficacy between auto HCT and
RI allo HCT for chemotherapy-sensitive relapsed FL.
MA Allo HCT Compared with RI Allo HCT
Given the relative risks and benefits of allo HCT, it stands

to reason that this treatment approach would offer more
benefit if the toxicity could somehow be mitigated. With the
advent of RI conditioning regimens, this has become an
appealing modality for management of relapsed or refractory indolent lymphoma.2,3 Biologically, the growth kinetics
of this group of diseases would seem to make them amenable
to an approach in which the bulk of antimalignancy effects
are mediated by GVL. In particular, diseases that progress
more slowly may be less likely to outpace the engraftment
of the donor immune system, which is required to see GVL,
and less likely to require high-dose conditioning to induce
early disease control. Although limited prospective data
exist to support this rationale, again there are retrospective
analyses that can offer some evidence that this hypothesis is
correct.
The largest of these studies for indolent lymphoma comparing RI with MA conditioning was a review of the IBMTR/
ABMTR registry.17 They compiled 120 MA and 88 RI cases,
all of whom had FL. Patients in the RI cohort were noted to
be significantly older and later in the disease course, while
the MA cohort more frequently included patients with primary refractory disease and bone marrow involvement.
Even though the type of conditioning did not correlate with
risk of TRM, PFS, or OS, recipients of RI allo HCT did have
a significantly higher risk of progression on multivariate
analysis (p 0.044). These results are in contrast to a cohort
of 220 patients (84 of whom had indolent disease including
chronic lymphocytic leukemia [CLL] and T-cell lymphomas)
treated with either MA or RI allo HCT at Fred Hutchinson
Cancer Research Center (FHCRC).18 Among patients with

Table 2. Summary of Results from Selected Prospective Trials of Reduced-Intensity Allogeneic Hematopoietic Cell Transplantation
for Indolent Lymphoma
Study
(Sample Size)
Morris et al23 2004

(n 41 low-grade^; 88 total)*
Rezvani et al22 2008
(n 62; 16 with HT)
Khouri et al21 2008
(n 47, all FL)
Piana et al24 2010
(n 37, all FL)
Thomson et al26 2010
(n 82, all FL)*
Shea et al25 2011
(n 16 FL; 47 total)
Conditioning
Regimen
GVHD
Prophylaxis
Prior Tx: Median

(Range)
Prior
Auto HCT
aGVHD, cGVHD
NRM/TRM
EFS/PFS
OS
Median
F/U
Flu/Mel
Alem CSP
3 (16)#
37%
15%, 5%#
11%
65%
73%
3 yrs
TBI (18%),
Flu/TBI (82%)#
FCR
CSP MMF
6 (119)#
44%#
63%, 47%#
42%#
43%
52%
3 yrs
TAC MTX
3 (27)
19%
15%
83%
85%
5 yrs
Flu/Mel
CSP MTX
3 (NR)
46%
11%
36%
51%, 53%
37%
55%
57%
4.3 yrs
Flu/Mel
Alem CSP
4 (18)
26%
13%, 18%
15%
76%
76%
3.6 yrs
FC
TAC
2 (13)
0%
29%, 18%#
14%#
75%
81%
4.6 yrs
Abbreviations: GVHD, graft-versus-host disease (a acute Grades 2 4, c chronic extensive); Tx, treatment; auto HCT, autologous hematopoietic cell
transplantation; aGVHD, acute GVHD (grades 2 4); cGVHD, chronic GVHD (extensive); NRM, nonrelapse mortality; TRM, treatment-related mortality; EFS, event-free
survival; PFS, progression-free survival; OS, overall survival; F/U, follow-up; Flu, fludarabine; Mel, melphalan; Alem, alemtuzumab; CSP, cyclosporine; yrs, years; HT,
histologic transformation; TBI, total body irradiation; MMF, mycophenolate mofetil; FL, follicular lymphoma; FCR, fludarabine, cyclophosphamide, and rituximab; TAC,
tacrolimus; MTX, methotrexate; NR, not reported; FC, fludarabine and cyclophosphamide.
^
Specific histologies comprised within this group are as follows: 29 follicular lymphomas, three lymphoplasmacytic lymphomas, and 9 chronic lymphocytic or
prolymphocytic leukemias.
* Nineteen of the patients treated in the study by Thomson et al from 2010 were also described in the report by Morris et al from 2004 but with more than 5 years
of additional follow-up. Since this comprised a minority of the patients in the Morris study, it was included as its own reference.
# The data reported were not segregated by histology and therefore represent the result for the entire study cohort, not exclusively for patients with low-grade
lymphoma
indolent histologies, there was a significantly higher NRM

(p 0.02), a trend toward higher overall mortality (p
0.07), and a nonsignificantly lower risk of relapse after MA
(p 0.33) compared with RI allo HCT. These authors also
used the Hematopoietic Cell Transplant-Comorbidity Index
(HCT-CI) to estimate risk based on pretransplant comorbidities, but this analysis was performed for the entire study
cohort, irrespective of histology.19 In patients with a lowrisk HCT-CI score, there was no difference in NRM, relapse
rate, or OS at 3 years between patients receiving RI or MA
allo HCT. However, in patients with an intermediate- or
high-risk HCT-CI score, the RI patients had significantly
lower NRM (p 0.001) and significantly higher OS (p
0.007) compared with MA patients, but relapse rates were
not statistically different (p 0.26). Another relatively large
registry study from the European Group for Blood and
Marrow Transplantation came to comparable conclusions,
though this study was unique in that it restricted the
analysis to patients with HLA-matched unrelated donors.20
In 131 patients with FL, MA allo HCT was associated with
significantly worse outcomes for NRM, PFS, and OS by
multivariate analysis (p 0.01 for all three endpoints).
Although the data are not conclusive, it would appear that
RI allo HCT may offer long-term disease control with less
risk of serious toxicity compared with MA allo, perhaps most
relevant in patients with significant comorbidities.
Focus on RI Allo HCT
There have been several prospective studies focusing on

the outcome of indolent lymphomas following RI allo HCT,
but all have been single-arm trials (Table 2).21-26 One
important finding seen in each of these studies is that
relapses of indolent lymphoma following RI allo HCT were
rarely seen past 3 years. The study by Khouri et al was
recently updated and presented in abstract form. They have
seen only one additional recurrence of FL with an added
3 years of follow-up, yielding 10-year estimates of OS of
78% and PFS of 72%.27 In the study by Thomson et al, the
latest relapse seen was at 43 months, with no other relapses

seen past 3 years.26 A few other salient features from these
studies are worth mentioning.
First, patients included in these studies were generally
heavily pretreated, with a substantial portion in some of the
studies having received prior auto HCT. This is in keeping
with the NCCN guidelines that allo HCT should generally
be reserved for patients with relapsed for refractory disease.11 There were varying degrees of chemotherapy sensitivity across the trials. The studies by Khouri et al and Shea
et al exclusively enrolled patients who were in either complete remission (CR) or partial remission (PR), which may in
part explain why these studies had some of the best survival
outcomes.21,25 The report by Pinana et al noted a nonsignificant trend in 4-year OS according to disease status (71% for
those in CR, 48% for those in PR, and 29% for refractory or
progressive disease [PD]; p 0.09).24 In the largest of these
studies, Thomson et al noted that the absence of prior auto
HCT predicted for improved OS and PFS in a univariate
analysis, but this did not persist following multivariate
analysis.26 The studies by Morris et al and Rezvani et al
supported the concept that chemotherapy sensitivity and
disease status at the time of transplantation are predictive
of outcome.22,23 However, even in these seemingly higherrisk patients, salvage and long-term remission is possible
with this approach.
Another issue in these studies worth addressing is the
effect of GVHD on outcome. Classically, GVHD is an indication of an immunologically active allograft, which should
also be capable of mediating GVL effects. However, GVHD
remains the major cause of morbidity and NRM in allo HCT.
Although none of these studies were designed to determine
the best method of GVHD prophylaxis, two of these studies
from the same group used a strategy of partial in vivo T-cell
depletion using the anti-CD52 antibody alemtuzumab.23,26
In these studies, the authors noted a relatively low incidence
of GVHD. However, relapse rates with this approach were
higher than that seen in the other prospective studies
497
CASSADAY AND GOPAL
Fig. 1. Proposed algorithm to incorporate the use of hematopoietic cell transplantation in patients with indolent B-cell non-Hodgkin
lymphoma. ^ As the best treatment for relapsed indolent lymphoma remains unclear, these patients should be treated in the context of a clinical
trial if feasible. * If a patient has not responded to second-line chemotherapy or has significant disease burden, enrollment in a clinical trial
should be strongly considered before use of HCT. Either auto or allo HCT could then be considered to consolidate a response to this intervention.
Abbreviations: HCT, hematopoietic cell transplantation; CR1, first complete remission; HCT-CI, hematopoietic cell transplant-comorbidity
index; auto, autologous; allo, allogeneic; RI, reduced-intensity; MA, myeloablative; BSC, best supportive care; WIS, withdrawal of immunosuppression; DLI, donor lymphocyte infusion.
reviewed in this article, though these relapses were often

salvageable with further immunomanipulation (see additional details below). It is noteworthy that none of these
prospective studies showed that the presence of either acute
or chronic GVHD was associated with decreased relapse or
improved disease-free survival.
With these data suggesting that relapse rates are higher
in patients not in CR, investigators hypothesized that radioimmunotherapy (RIT) as part of the preparative regimen
could deliver targeted radiation to tumor sites with minimal
nonhematologic toxicity to improve early post-transplant
disease control and potentially lead to improved PFS.
Bethge et al performed a phase II study of yttrium-90
conjugated to the anti-CD20 antibody ibritumomab tiuxetan
(90-YIT) plus RI allo HCT.28 A total of 40 patients were
accrued (17 with FL, 1 with MZL, 1 with LPL, 13 with CLL,
and 8 with mantle cell lymphoma). All had received at least
two prior chemotherapy regimens, and 82% had active
disease at the time of HCT. This treatment approach yielded
2-year estimated EFS of 43% and OS of 51% for the entire
cohort. NRM at 2 years was relatively high compared with
other studies at 45%, which the authors attributed to a
relatively high-risk patient population and high frequency
(68%) receiving grafts from unrelated donors. Additionally, a
cohort of 40 patients with high-risk B-cell lymphomas was
treated at FHCRC with 90-YIT along with RI allo HCT, 45%
498
of whom had indolent histologies.29 For the indolent subset,

only 17% (3 patients) had chemotherapy-sensitive disease at
the time of HCT, but the 3-month response rate was still
83%, with over 75% alive and approximately 50% alive and
progression-free at 2.5 years. The cumulative incidence
estimate of NRM at 2.5 years was 16%, more in line with
other studies reported here. These studies suggest the
feasibility and potential efficacy of the use of RIT added to RI
preparative regimens for patients with indolent B-NHL not
in CR.
Management of Relapse after Allo HCT
Relapse of hematologic malignancies after allo HCT poses

a particularly daunting challenge, and indolent lymphomas
are no exception. Apart from additional chemoimmunotherapy or radiation, manipulation of the allografted immune system is an intervention unique to this particular
disease state. This can be attempted using withdrawal of
immunosuppression (WIS) or donor lymphocyte infusion
(DLI). Many of the studies mentioned above include a small
population of patients where such interventions were performed with or without additional anticancer therapy, occasionally with dramatic and durable responses.21,26 DLI in
particular is advocated in patients who received alemtuzumab as part of their transplant conditioning. A recent
publication from FHCRC specifically analyzed their experi-
ence with managing lymphoma relapsing after allo HCT,

revealing that those with indolent histologies had significantly less mortality compared to aggressive NHL in a
multivariate model (p 0.008).30 This analysis also demonstrated potential benefit of WIS and DLI. In the absence of
significant GVHD, WIS and/or DLI should be considered in
patients with indolent lymphoma who have relapsed after
allo HCT and desire further treatment of their disease.
Special Circumstance: Histologic Transformation
One of the most challenging aspects of managing indolent

lymphoma is the occurrence of histologic transformation
(HT) to an aggressive large-cell lymphoma. The general
treatment approach is similar to that for diffuse large B-cell
lymphoma (DLBCL), with one notable exception. Since HT
is thought to have a worse prognosis than de novo DLBCL,
the NCCN guidelines recommend considering consolidation
with either auto or allo HCT in first remission (either
complete or partial).11 This may be particularly true for
patients who are unable to receive anthracycline-based
therapy. Few of the studies mentioned above included patients with HT, as this manifestation takes on the characteristics of an intermediate- or high-grade lymphoma. A
recent prospective study in Norway evaluated 47 patients
with indolent B-NHL that had relapsed with HT following
chemotherapy.31 All patients treated in this trial received
salvage chemotherapy with a plan to consolidate responders
with auto HCT; however, only 30 met this criterion and
received sufficient treatment to be included in the analysis.
Among these transplanted patients, the 5-year PFS and OS
rates were 32% and 47%, respectively, with a plateau in the
PFS curve at 30% beyond 3.5 years. It is noteworthy that, in
the 17 patients from the initial cohort who did not undergo
auto HCT, there were three patients (18%)referred to by
the doctors as long-term survivorswho received salvage
chemotherapy and/or radiation alone. The adverse prognosis
of HT before RI allo HCT was evaluated by Rezvani et al,
noting that the risk of relapse was nearly 5 times higher for
HT than for those with nontransformed relapsed/refractory
indolent lymphoma (p 0.001).22 Moreover, OS at 3 years
for relapsed/refractory indolent disease was 52% compared

with 18% for HT (p 0.02). Although these data underscore
the poor prognosis following HT, long-term remissions are
possible in a subset of patients if aggressive therapy can be
tolerated.
Conclusion
Although the long-term prognosis of indolent lymphomas

has improved with standard chemoimmunotherapy approaches alone, HCT does offer the potential for durable
remissions for a subset of patients who are eligible for such
treatment. Figure 1 shows a proposed algorithm for the
application of HCT in the management of these diseases,
conceding that rigorous prospective trials are not readily
available for all decision points. It is important to remember
that both auto and allo HCT require a large amount of
preparation (both for the provider and for the patient), so
their application should be considered early in the course of
treating patients with indolent lymphoma.
In summary, auto HCT for chemotherapy-sensitive relapse can provide improved outcomes, though concerns regarding secondary malignancies remain. The development
of RI allo HCT has allowed the benefits of GVL and a
potential cure to be offered to patients that would have been
deemed too unfit for traditional MA conditioning. However,
despite this advance, GVHD remains a significant cause of
morbidity and mortality. Challenges also remain in improving disease control for those with chemotherapy-resistant or
transformed disease as well placing HCT in the context of
ever-expanding nontransplant options for patients with indolent B-NHL. Despite these hurdles, both auto and allo
HCT remain viable, effective options for many patients with
indolent lymphoma.
Acknowledgment
Dr. Cassaday is supported by the National Institutes of
Health T32 training grant number T32CA00951527. Dr. Gopal
is supported by the following: P01 CA044991, Leukemia &
Lymphoma Society SCOR grant 7040, and a gift from Frank and
Betty Vandermeer. Dr. Gopal is a Scholar in Clinical Research
of the Leukemia & Lymphoma Society.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Ryan D. Cassaday*
Ajay K. Gopal
Seattle Genetics
Millennium;
Seattle Genetics
Abbott
Laboratories;
Biomarin;
Cephalon;
GlaxoSmithKline;
Merck; Pfizer;
Piramal; Seattle
Genetics;
Spectrum
Pharmaceuticals
REFERENCES
1. Swerdlow SH, Campo E, Harris NL, et al (eds). WHO Classification of
Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France; International Agency for Research on Cancer: 2008.
2. Khouri IF, Keating M, Korbling M, et al. Transplant-lite: Induction of
graft-versus-malignancy using fludarabine-based nonablative chemotherapy
and allogeneic blood progenitor-cell transplantation as treatment for lymphoid malignancies. J Clin Oncol. 1998;16:2817-2824.
3. McSweeney PA, Niederwieser D, Shizuru JA, et al. Hematopoietic cell

transplantation in older patients with hematologic malignancies: Replacing
high-dose cytotoxic therapy with graft-versus-tumor effects. Blood. 2001;97:
3390-3400.
4. Sun Y, Tawara I, Toubai T, et al. Pathophysiology of acute graft-versushost disease: Recent advances. Transl Res. 2007;150:197-214.
5. Gyan E, Foussard C, Bertrand P, et al. High-dose therapy followed by
499
CASSADAY AND GOPAL

autologous purged stem cell transplantation and doxorubicin-based chemotherapy in patients with advanced follicular lymphoma: A randomized multicenter study by the GOELAMS with final results after a median follow-up of
9 years. Blood. 2009;113:995-1001.
6. Ladetto M, De Marco F, Benedetti F, et al. Prospective, multicenter
randomized GITMO/IIL trial comparing intensive (R-HDS) versus conventional (CHOP-R) chemoimmunotherapy in high-risk follicular lymphoma at
diagnosis: The superior disease control of R-HDS does not translate into an
overall survival advantage. Blood. 2008;111:4004-4013.
7. Lenz G, Dreyling M, Schiegnitz E, et al. Myeloablative radiochemotherapy followed by autologous stem cell transplantation in first remission
prolongs progression-free survival in follicular lymphoma: Results of a prospective, randomized trial of the German Low-Grade Lymphoma Study
Group. Blood. 2004;104:2667-2674.
8. Sebban C, Mounier N, Brousse N, et al. Standard chemotherapy with
interferon compared with CHOP followed by high-dose therapy with autologous stem cell transplantation in untreated patients with advanced follicular
lymphoma: The GELF-94 randomized study from the Groupe dEtude des
Lymphomes de lAdulte (GELA). Blood. 2006;108:2540-2544.
9. Rohatiner AZ, Nadler L, Davies AJ, et al. Myeloablative therapy with
autologous bone marrow transplantation for follicular lymphoma at the time
of second or subsequent remission: Long-term follow-up. J Clin Oncol.
2007;25:2554-2559.
10. Schouten HC, Qian W, Kvaloy S, et al. High-dose therapy improves
progression-free survival and survival in relapsed follicular non-Hodgkins
lymphoma: Results from the randomized European CUP trial. J Clin Oncol.
2003;21:3918-3927.
11. Zelentz AD, Abramson JA, Advani RA, et al. NCCN Clinical Practice
Guidelines in Oncology: Non-Hodgkins Lymphomas. http://www.nccn.org/
professionals/physician_gls/f_guidelines.asp. Accessed January 11, 2012.
12. Dimopoulos MA, Gertz MA, Kastritis E, et al. Update on treatment
recommendations from the Fourth International Workshop on Waldenstroms
Macroglobulinemia. J Clin Oncol. 2009;27:120-126.
13. van Besien K, Loberiza FR, Jr., Bajorunaite R, et al. Comparison of
autologous and allogeneic hematopoietic stem cell transplantation for follicular lymphoma. Blood. 2003;102:3521-3529.
14. Ingram W, Devereux S, Das-Gupta EP, et al. Outcome of BEAMautologous and BEAM-alemtuzumab allogeneic transplantation in relapsed
advanced stage follicular lymphoma. Br J Haematol. 2008;141:235-243.
15. Evens AM, Vanderplas A, LaCasce AS, et al. Outcomes with autologous
(auto) and allogeneic (allo) stem cell transplantation (SCT) for relapsed/
refractory follicular lymphoma (FL) in the post-rituximab era: A comparative
analysis from the National Comprehensive Cancer Network (NCCN) NonHodgkins Lymphoma (NHL) Outcomes Database Project. Blood. 2011;118
(abstr 4112).
16. Tomblyn MR, Ewell M, Bredeson C, et al. Autologous versus reducedintensity allogeneic hematopoietic cell transplantation for patients with
chemosensitive follicular non-Hodgkin lymphoma beyond first complete response or first partial response. Biol Blood Marrow Transplant. 2011;17:
1051-1057.
17. Hari P, Carreras J, Zhang MJ, et al. Allogeneic transplants in follicular
lymphoma: Higher risk of disease progression after reduced-intensity compared to myeloablative conditioning. Biol Blood Marrow Transplant. 2008;
14:236-245.
18. Sorror ML, Storer BE, Maloney DG, et al. Outcomes after allogeneic
500
hematopoietic cell transplantation with nonmyeloablative or myeloablative

conditioning regimens for treatment of lymphoma and chronic lymphocytic
leukemia. Blood. 2008;111:446-452.
19. Sorror ML, Maris MB, Storb R, et al. Hematopoietic cell transplantation (HCT)-specific comorbidity index: A new tool for risk assessment before
allogeneic HCT. Blood. 2005;106:2912-2919.
20. Avivi I, Montoto S, Canals C, et al. Matched unrelated donor stem cell
transplant in 131 patients with follicular lymphoma: An analysis from the
Lymphoma Working Party of the European Group for Blood and Marrow
Transplantation. Br J Haematol. 2009;147:719-728.
21. Khouri IF, McLaughlin P, Saliba RM, et al. Eight-year experience with
allogeneic stem cell transplantation for relapsed follicular lymphoma after
nonmyeloablative conditioning with fludarabine, cyclophosphamide, and
rituximab. Blood. 2008;111:5530-5536.
22. Rezvani AR, Storer B, Maris M, et al. Nonmyeloablative allogeneic
hematopoietic cell transplantation in relapsed, refractory, and transformed
indolent non-Hodgkins lymphoma. J Clin Oncol. 2008;26:211-217.
23. Morris E, Thomson K, Craddock C, et al. Outcomes after alemtuzumabcontaining reduced-intensity allogeneic transplantation regimen for relapsed
and refractory non-Hodgkin lymphoma. Blood. 2004;104:3865-3871.
24. Pinana JL, Martino R, Gayoso J, et al. Reduced intensity conditioning
HLA identical sibling donor allogeneic stem cell transplantation for patients
with follicular lymphoma: Long-term follow-up from two prospective multicenter trials. Haematologica. 2010;95:1176-1182.
25. Shea T, Johnson J, Westervelt P, et al. Reduced-intensity allogeneic
transplantation provides high event-free and overall survival in patients with
advanced indolent B cell malignancies: CALGB 109901. Biol Blood Marrow
Transplant. 2011;17:1395-1403.
26. Thomson KJ, Morris EC, Milligan D, et al. T-cell-depleted reducedintensity transplantation followed by donor leukocyte infusions to promote
graft-versus-lymphoma activity results in excellent long-term survival in
patients with multiply relapsed follicular lymphoma. J Clin Oncol. 2010;28:
3695-3700.
27. Khouri IF, Saliba RM, Valverde R, et al. Nonmyeloablative allogeneic
stem cell transplantation with or without 90-yttrium ibritumomab tiuxetan
(90YIT) is curative for relapsed follicular lymphoma: Median 9-year follow-up
results. Blood. 2011;118 (abstr 662).
28. Bethge WA, Lange T, Meisner C, et al. Radioimmunotherapy with
yttrium-90-ibritumomab tiuxetan as part of a reduced-intensity conditioning
regimen for allogeneic hematopoietic cell transplantation in patients with
advanced non-Hodgkin lymphoma: Results of a phase 2 study. Blood. 2010;
116:1795-1802.
29. Gopal AK, Guthrie KA, Rajendran J, et al. Y-Ibritumomab tiuxetan,
fludarabine, and TBI-based nonmyeloablative allogeneic transplantation
conditioning for patients with persistent high-risk B-cell lymphoma. Blood.
2011;118:1132-1139.
30. Ram R, Gooley TA, Maloney DG, et al. Histology and time to progression predict survival for lymphoma recurring after reduced-intensity conditioning and allogeneic hematopoietic cell transplantation. Biol Blood Marrow
Transplant. 2011;17:1537-1545.
31. Eide MB, Lauritzsen GF, Kvalheim G, et al. High-dose chemotherapy
with autologous stem cell support for patients with histologically transformed
B-cell non-Hodgkin lymphomas. A Norwegian multi centre phase II study.
Br J Haematol. 2011;152:600-610.
CONTROVERSIES IN MYELOMA: INDUCTION,

TRANSPLANT, AND MAINTENANCE
CHAIR
Amrita Krishnan, MD
City of Hope National Medical Center
Duarte, CA
SPEAKERS
S. Vincent Rajkumar, MD
Mayo Clinic
Rochester, MN
Michel Attal, MD
CHU Purpan
Toulouse, France
Stem Cell Transplantation for Multiple

Myeloma: Who, When, and What Type?
By Amrita Krishnan, MD
Overview: Early randomized trials of high-dose chemotherapy with autologous stem cell rescue showed improved
progression-free survival (PFS) over conventional chemotherapy. However, in the era of novel agents for myeloma in
conjunction with the evolution of hematopoietic stem cell
transplantation, many new questions arise. First, how can
novel agents be incorporated into the transplant paradigm?
Given the efficacy of new induction regimens, should transplant be delayed until relapse? Also, in the era of individualized medicine, chronologic age alone should not drive
decisions regarding transplantation. Therefore, the feasibility
and role of transplantation in older patients with myeloma is
HE CONCEPT of dose-intensive chemotherapy to treat

a malignant disease is a paradigm that has been
explored for over a quarter century. A group from The Royal
Marsden Hospital1 demonstrated that increased doses of
melphalan could induce responses in nine patients with
myeloma; however, this was at the cost of significant hematologic toxicity with a median of 46 days of neutropenia. In
later trials with the use of autologous hematopoietic peripheral blood stem cell support, this approach became safer.
Ultimately, a randomized trial by the Intergroupe Francophone du Myelome (IFM) demonstrated the superiority of
high-dose chemotherapy in terms of both OS and diseasefree survival over conventional chemotherapy.2
The IFM trial is now more than a decade old, and in this
era of novel agents, many new questions remain unanswered. For example, the original IFM trial included patients 65 years and younger. In this modern era, should
there be an age limit for myeloma transplant? Also, what is
the optimal induction regimen, and how many cycles of
therapy should be administered (i.e., when do you transplant a patient)? Lastly, what is the best way to incorporate
novel agents into the post-transplant setting?
When to Transplant
It is common knowledge that the use of melphalancontaining induction regimens should be avoided in patients
who could potentially undergo autologous transplantation.3
In contrast, there is little consensus on the optimal nonmelphalan-containing induction regimen. The summary by
S. Vincent Rajkumar, MD, will discuss in depth the choice of
an optimal induction regimen. In regard to the optimal
number of cycles before transplant, there remains no consensus. There are numerous phase I and II clinical trials of
two-, three-, and even four-drug induction regimens. All
incorporate at least one novel agent and traditionally still
include a steroid backbone, although with an intent to be
more steroid-sparing by using lower doses of steroids. The
lenalidomide, bortezomib, and dexamethasone (RVD) regimen was initially used in a phase I trial in the relapsed
setting and demonstrated good tolerability and high response rates.3 In the upfront setting, response rates are
extremely high (100%), as is depth of response (complete
remission [CR]/near CR [nCR] 40%).4 In that initial trial,
patients had the option to proceed to transplant after four
502
being studied. The controversy of transplant type (i.e., autologous compared with reduced intensity allogeneic transplant)
remains unresolved. Several large international trials have
demonstrated conflicting results in regard to an overall survival (OS) benefit with the allogeneic approach. The role of
allogeneic transplant remains under study especially in the
high-risk population, which has high relapse rates with traditional autologous approaches. Future directions to reduce
relapse include post-transplantation consolidation and maintenance therapy with either approved agents or new agents
and immunotherapy, either vaccine based or natural killer (NK)
and T-cell based.
cycles, and the response rate was reported after four cycles.
However, the investigators report 75% of patients had a
further upgrade of responses at six or eight cycles. It is
unknown whether this would ultimately affect posttransplant PFS.
A new three-drug regimen that may soon gain traction is
the combination of the new proteasome inhibitor carfilzomib, lenalidomide, and dexamethasone (CRD).5 Preliminary results showed very high response rates (100%
partial response [PR]) in phase I and II trials, as well as the
highest depth of response seen outside a transplant setting
(79% CR/nCR). Twenty-four of 49 patients in this trial had
stem cells collected. However, too few patients have gone on
to transplant to be able to assess the effect of this induction
on post-transplant survival either with early or delayed
transplant.
Indirectly, one can extrapolate from phase III trials that
improved responses before transplant can mean further
improvements post-transplant, at least in terms of PFS.
There have been several large phase III trials in Europe
comparing traditional regimens to newer three-drug regimens incorporating novel agents. For example, Cavo et al
compared bortezomib, thalidomide, and dexamethasone
(VTD) with thalidomide and dexamethasone (TD).6 Both
arms received three courses of induction before transplant,
followed by two cycles of consolidation with VTD or TD. Not
surprisingly, the CR/nCR rate pretransplant was higher in
the VTD arm (31% vs. 11%; p 0.0001 for TD). Posttransplant this translated into an improved PFS at three
years (69% vs. 37%) but not an OS benefit. Though it
remains unknown whether this was due to induction or
consolidation or both.
In the MRCIX trial of CTD versus CVAD, the post-SCT
CR rate was 50% versus 37% in the CVAD arm. PFS was
greater in patients who received a CR post-transplant.
Further statistical modeling suggested that with longer
follow up this translated into a small PFS benefit.
From the City of Hope Cancer Center, Duarte, CA.

Address reprint requests Amrita Krishnan, MD, 1500 E. Duarte Road, Duarte, CA,
91010; email: akrishnan@coh.org.
1092-9118/10/1-10
TRANSPLANTATION FOR MYELOMA
Alternately for less aggressive disease, many experts in

the field advocate the use of a two-drug regimen of either
bortezomib and dexamethasone or lenalidomide and dexamethasone depending on the patients comorbidities. The
lenalidomide plus low-dose dexamethasone approach has
good tolerability and 2-year OS data in both young and
elderly patients.7 Although the point at which patients
should receive transplant on this induction regimen remains
controversial, it is clear that prolonged exposure to lenalidomide may impair stem cell yields. Therefore, if a transplant
is planned, transplant physicians recommend that stem cell
collection should occur before the six months of lenalidomide
exposure.8
Who Should Undergo a Transplant?
Age Limits for Transplantation?
Myeloma is a disease of elderly patients with a median of

72 years at presentation. Therefore, if arbitrary age cutoffs
are used, a large portion of patients with myeloma would be
ineligible for a transplant. In earlier eras, autologous transplantation was generally reserved for patients younger than
60 or 65 years. The original IFM trialwhich served as the
platform for high-dose therapy in multiple myeloma enrolled patients 65 years and younger. However, since that
time, there have been several trials suggesting that age is
not a prognostic variable to transplant outcome in myeloma,
but rather it is disease-related indices. A group from M. D.
Anderson Cancer Center published their experience with 84
patients with the median age 72. All patients met the
standard organ function criteria for transplant and received
melphalan doses between 140 mg/m2 and 200 mg/m2.
Nonrelapse-related mortality was 3%, and 5-year OS was
67%.9 A larger registry-based study comparing patients
older than 60 years and younger than 60 years showed no
difference in transplant-related mortality or PFS between
older and younger patients. Though the median age in the
older than age 60 group was still on the younger side at 63
years.10
Specific issues about elderly patients to consider begin
with the choice of induction therapy for a patient that one
KEY POINTS
The use of novel agents as induction therapy for

myeloma can improve post-transplant progressionfree survival.
Studies are ongoing to determine the optimal timing
of transplantation after an effective induction strategy (i.e., early compared with delayed transplant).
Patients with high-risk myeloma especially those
with 17p deletionremain a challenge and may be a
subgroup for whom allogeneic transplantation should
be considered.
Post-transplant strategies to reduce relapse, such as
consolidation or immunotherapy, are being studied.
Age should not be considered a barrier to transplantation, but appropriate assessment of the older patient should include geriatric assessments and
traditional functional testing.
would consider transplant eligible. As previously mentioned, melphalan-based regimens are to be avoided because
of potential stem cell toxicity. High-dose dexamethasone is
also particularly difficult for the older patient because of
toxicity. In addition, increasing age has been correlated
inversely with CD34 counts.11 Therefore, stem cell collection
should be considered early in the disease course, and the use
of agents such as plerixafor may be necessary.
Before transplant, Karnofsky performance status is the
most traditional evaluation in addition to organ function
testing. However, in the older patient, performance status as
the only marker of functional status may be inadequate. The
concept of frailty is important to consider. For example, the
very fit older person who exercises and is fully independent
has different needs than the mildly frail individual who may
need help for household tasks. Frailty in and of itself is a
predictor of outcome in elderly patients.12 The geriatric
assessment tool may be helpful as part of pretransplant
assessment. Although it has not yet been validated in the
transplant setting, it has been a predictor of chemotherapy
toxicity with conventional chemotherapy. This tool, in addition to comorbidities, encompasses cognition, psychologic
status, social functioning and support, and nutritional status.13 Optimal assessment of all these factors could help
predict the older patient who may pass all the functional
testing but may have the potential for significant debility
with the transplant. In addition, it may help with needs
assessment for post-transplant care.
Patients at High Risk
Risk stratification is an important part of the initial

assessment of a myeloma patient and obviously influences
the choices of induction therapy, but what role should it play
in the transplant paradigm? Generally, risk stratification
encompasses International Staging System (ISS) staging,
cytogenetics, and fluorescence in situ hybridization (FISH)
analysis and possibly gene expression profiling (GEP). The
early transplant studies did not differentiate between patients at standard risk and those at high risk. Also, although
later trials did attempt to stratify patients and as the
definition of high-risk disease evolves, the applicability of
those older trials is limited. Recent trials tend to define
high risk as t(4,14), t(14,16), deletion 13, or deletion 17p. A
retrospective study reviewed approximately 500 patients
who received bortezomib plus dexamethasone before highdose melphalan compared with a group treated with vincristine, doxorubicin, dexamethasone (VAD) induction before
high-dose melphalan.14 The use of bortezomib could abrogate some of the high-risk prognosis conferred by t(4,14).
Event-free survival (EFS) was 28 months for patients
treated with bortezomib compared with 16 months for the
VAD group (p 0.001). However, this still remained a poor
prognostic factor compared with patients without t(4,14). In
contrast, no improvement in either EFS or OS was seen for
patients with 17p deletion treated with the bortezomibbased induction. However, in the HOVON-65 GMMG-HD4
trial, bortezomib-based treatment before and after stem cell
transplant markedly improved DFS and OS in patients with
17p deletion compared with those treated with VAD induction and thalidomide maintenance.15
GEP has also been studied as a prognostic indicator in the
era of novel agents plus high-dose therapy as part of the
Arkansas Total Therapy 3 protocol.16 For patients with
503
AMRITA KRISHNAN
Fig. 1. BMT CTN 012 trial: tandem autologous transplant compared with autologous plus reduced intensity allogeneic transplant.
Abbreviations: PFS, progression-free survival; OS, overall survival.
low-risk gene expression profile myeloma, the addition of

bortezomib to the Total Therapy regimen improved clinical
outcomes. However, for GEP-defined high-risk disease, EFS
was uniformly poor at 27% for patients with ISS stage III
disease and 36% for ISS stage I. Therefore, different treatment approaches are still needed for this group of patients.
Whether the optimal approach is Total Therapy 5 (increased
dose density) or allogeneic approaches, which are discussed
below, remains to be seen.
Plasma cell leukemia is also generally considered a poor
prognostic indicator. The role of transplantation for this
group of patients is controversial. However, novel agents for
induction may improve the prognosis and make transplant a
viable option. A Center for International Bone Marrow
Transplant Research retrospective review of 160 patients
with plasma cell leukemia who received either autologous or
allogeneic transplantation, demonstrated surprisingly good
survival in the autologous arm: 64% at a median follow-up of
38 months.17
In summary, novel agents as part of induction and possibly consolidation and maintenance may improve autologous
transplant outcomes of certain subgroups of patients with
high-risk disease. However, for the group with 17p deletion,
improvement may be demonstrated, with bortezomib alternate strategies are still needed.
Allogeneic Transplantation
Should allogeneic transplantation be considered for patients with poor-risk disease, including those with 17p
deletion? Traditional allogeneic transplant with myeloablative conditioning was associated with high transplantrelated mortality but did demonstrate a higher rate of
molecular remission and lower rates of relapse than autologous transplantation.18 Reduced intensity allogeneic transplant may harness the immunologic benefits of the
allogeneic approach but with reduced transplant-related
mortality. It remains controversial whether this ultimately
will lead to improved OS, especially for patients with highrisk disease.
504
Three large trials have shown conflicting results. The

French Group (IFM) ran two parallel phase II trials in
patients with high-risk myeloma, defined by deletion of
chromosome 13 or beta-2 microglobulin greater than 3 mg/L.
IFM99 03 analyzed 65 patients with available human leukocyte antigen (HLA)matched sibling donors who received
an allogeneic hematopoietic cell transplant (HCT) with reduced intensity conditioning after an autologous HCT. Outcomes of these patients were compared with 219 patients
enrolled in IFM99 04, an auto-auto trial. EFS and OS were
not significantly different for patients receiving auto-allo
and auto-auto in the IFM trial.19
In contrast, Bruno et al in an Italian biologic assignment
trial demonstrated superior EFS and OS in 58 and 46
recipients of auto-allo transplant compared with tandem
autologous transplant. However, they did not stratify patients as high risk.20 The United States BMT CTN 0102
trial did stratify patients as high risk and standard risk. The
definition of high risk was the same as during the era when
the trial started and, therefore, was defined as deletion 13 by
classic karyotyping and beta-2 microglobulin greater than
3.0. The trials primary endpoint was 3-year PFS in patients
with standard-risk disease, and for this group there was no
PFS or OS benefit to auto-allo transplant (Fig. 1). Subgroup
analysis of patients with high-risk, protocol-defined disease
also did not show any benefit for the allogeneic route.
However, since deletion 17p was not protocol defined, it
remains unclear whether this group of patients could benefit
from the allogeneic approach.21 It also remains unclear
whether with longer follow-up the benefit of an allogeneic
approach may become apparent as more autologous patients
continue to relapse. Indeed, in the trial reported by Bjorkstrand et al, 5-year follow-up was an OS benefit of 65% in the
auto-allo arm compared with 58% in auto-auto arm.22
Early Compared with Delayed Transplant
This remains the most difficult and, as yet, not wholly

answered question. One randomized trial demonstrated
comparable efficacy in terms of OS for early transplant
compared with transplant delayed until first relapse, although early transplant improved quality of life scores.23
However, this was in the era predating novel agents. The
novel agents clearly have improved tolerability and efficacy
over the historic conventional agents. On the converse side,
it is also unknown whether we can effectively use high-dose
melphalan to salvage patients who relapse after initial
induction with the novel agents. Therefore, are patients
missing the optimal window to benefit from high-dose therapy if they delay transplant?
The Mayo Clinic studied 290 patients with newly diagnosed myeloma treated with an IMiD-based induction regimen. Patients who underwent stem cell harvest and
transplant within 12 months of diagnosis were considered
early transplant, and those who underwent these procedures more than 12 months after diagnosis were considered
delayed transplant. The overall response to transplant was
the same in the early and delayed transplant groups. The
time to progression after transplant also was not significantly different between the early and delayed transplant
groups (20 months vs. 16 months). However, the retrospective nature of the trial precludes accurate assessment of why
some patients opted for early rather than delayed transplant. In addition, the patients in the delayed transplant
group primarily had transplant at first relapse, so it is
unknown whether comparable results for deferred transplant would be seen after multiple relapses.
The current US-French Intergroup trial would be the most
relevant trial in the era of novel agents to answer the early
or delayed transplant question. However, it too cannot
address the question of the comparability of transplant after
multiple relapses. In this trial, patients with newly diagnosed myeloma will receive one cycle of RVD and then be
randomly selected to receive either additional RVD followed
by stem cell apheresis and autologous transplant or further
RVD and stem cell collection and no transplant until relapse (Fig. 2). This large trial will also look at important
surrogate and prognostic features, such as cytogenetics and
GEP.
What Is the Optimal Post-transplant Therapy?
Despite the increased response rates seen with autologous

transplant after induction therapy using novel agents, most
Fig. 2. Abbreviations: ASCT, autologous stem cell transplant; MEL,

melphalan; RVD, lenalidomide, bortezomib, and dexamethasone.
Fig. 3. Abbreviations: MM, multiple myeloma; SCT, stem cell transplantation; RVD, lenalidomide, bortezomib, and dexamethasone;
ASCT, autologous stem cell transplant.
patients will ultimately relapse. Options to reduce or delay

post-transplant relapse include consolidation therapy and/or
maintenance therapy. This article will focus on consolidation therapy.
Consolidation Therapy
This approach is akin to the leukemia therapy (i.e.,

further intensive chemotherapy after initial induction therapy). However, in contrast to leukemia consolidation, there
is no standard in myeloma in regard to either the agents or
the number of cycles administered. Before the approval of
novel agents, one consolidative approach was in fact a
second transplant as part of a tandem transplant in patients
who failed to achieve a very good partial remission (VGPR)
or better. The Arkansas Total Therapy 3 approach uses
tandem transplant but with novel agents as consolidation:
VTD PACE; bortezomib, thalidomide, dexamethasone, cisplatin, adriamycin, cyclophosphamide, and etoposide as induction and post-tandem transplant.24
Novel agents used as consolidation may serve to both
improve depth of response and prolong responses. A proof of
principle was the work by Ladetto et al, who treated 39
patients in a VGPR postsingle autologous transplant. These
selected patients had a detectable quantifiable molecular
marker based on immunoglobulin heavy chain rearrangement that could be followed. They were treated with four
courses of VTD, and the molecular markers were followed by
polymerase chain reaction (PCR) using tumor clone specific
primers. CR increased from 15% post-transplant to 49%
post-VTD, and even more striking was the increase in
molecular remissions from 3% to 18%.25 In addition, patients who had a quantitative depletion in tumor burden
above the median as measured by PCR had an improved
PFS.
The Nordic study group used a different post-transplant
consolidative strategy in a randomized phase III trial. In one
arm, patients received bortezomib in the traditional schedule for three cycles followed by weekly dosing for four cycles,
and the observation arm had no therapy. The treatment arm
had a higher response rate and PFS but no benefit in OS.26
The ongoing BMT CTN 0702 STAMINA trial may be illuminating regarding the benefit of consolidation. In this threearmed trial, all patients will receive a single autologous
transplant, one arm will receive a subsequent second transplant, one arm will receive RVD for four cycles, and one arm
505
AMRITA KRISHNAN
will go straight to maintenance. All three arms will receive

lenalidomide maintenance for 3 years or until disease progression (Fig. 3).
Immunotherapy
The post-transplant period is the ideal time point for

immunotherapy, as theoretically the disease burden is low.
Immune function remains depressed post-high-dose therapy
for many months. Ex vivo expansion and subsequent transfer of autologous stimulated T cells may enhance host
antitumor immunity and may also allow for enhancement
of a post-transplant vaccination strategy against tumordirected antigens. A phase I/II trial of this strategy in 54
patients has been conducted.27 At day 2, patients who were
post-transplant received an infusion of ex vivo stimulated
antiCD3/antiCD28 T cells. Patients positive for HLA-A2
antigen received pneumococcal vaccination and a multipeptide tumor antigen vaccine. Patients who were negative
received the pneumococcal vaccine only. Significant T-cell
recovery was seen at day 14. A subset of patients developed
immune responses to tumor antigens, but this did not
translate into improved EFS. However, the potential of this

approach is suggested by the demonstration of the rapid
recovery of cellular and humoral immunity as well as
immune responses to the cancer vaccine. Future trials will
focus on vaccines for other possibly more potent tumor
antigens and the use of adjuvant therapy, such as immunomodulatory agents, in conjunction with T-cell infusions.
Conclusion
Where does this leave the treating physician in 2012 as he

or she eyes the future? This future will likely include a more
risk-adapted approach, with risk stratification determined
by clinical features, cytogenetics, and GEP. The goal of
therapy would be to minimize toxicity in the patient with
low-risk disease and minimize risk of relapse in the patient
with poor-risk disease. Post-transplant disease assessment
by detection of minimal residual disease may also help guide
therapy. Options would include post-transplant consolidation with either current or new agents, reduced intensity
allogeneic transplant, NK- or T-cell-directed immunotherapy, and long-term maintenance with new agents.
Author
Amrita Krishnan
Employment or
Leadership
Positions
Consultant or
Advisory Role
Merck
Stock
Ownership
Celgene
Honoraria
Celgene;
Genentech;
Millennium
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. McElwain TJ, Powles RL. High-dose intravenous melphalan for plasma
cell leukemia and myeloma. The Lancet. 1983;322:822-823.
2. Attal M, Harousseau JL, Stooa AM, et al. A prospective randomized trial
of autologous bone marrow transplantation chemotherapy in multiple myeloma Intergroupe Francasi du Myelome. N Engl J Med. 1996;335:91-97.
3. Rajkumar SV. Multiple Myeloma 2012 update on diagnosis, risk stratification and management. Am J Hematology. 2012;87:78-88.
4. Richardson PG, Weller E, Lonial S, et al. Lenalidomide, bortezomib, and
dexamethasone combination therapy in patients with newly diagnosed myeloma. Blood. 2010;16:679-686.
5. Jakubowiak A, Dytfield D, Jagannath S, et al. Final results of a frontline
phase 1/2 study of carfilzomib, lenalidomide, and low-dose dexamethasone
(CRD) in multiple myeloma. Blood. 2011;118 (abstr 631).
6. Cavo M, Tacchetti P, Patriarca F, et al. Bortezomib with thalidomide
plus dexamethasone compared with thalidomide plus dexamethasone as
induction therapy before and consolidation therapy after double autologous
stem cell transplantation in newly diagnosed multiple myeloma. A randomized phase 3 study. Lancet. 2010;376:2075-2085.
7. Rajkumar SV, Jacobus S, Callander NS, et al. Lenalidomide plus
high-dose dexamethasone versus lenalidomide plus low-dose dexamethasone
as initial therapy for newly diagnosed multiple myeloma; An open-label
randomized controlled trial. Lancet Oncol. 2010;11:29-37.
8. Giralt S, Stadtmauer EA, Harousseau JL, et al. International Myeloma
Working Group (IMWG) consensus statement and guidelines regarding the
current status of stem cell collection and high-dose therapy for multiple
myeloma and the role of plerixafor (AMD3100). Leukemia. 2009;23:19041912.
9. Bashir Q, Shah N, Parmar S, et al. Feasibility of autologous stem cell
transplantation in patients 70 years with multiple myeloma. Leuk Lymphoma. 2012;53:118-122.
10. Reece DE, Bredeson C, Perez WS, et al. Autologous stem cell transplantation in multiple myeloma patients 60 vs. 60 years of age. Bone
Marrow Transplant. 2003;32:1135-1143.
11. Morris CL, Siegel E, Barlogie B, et al. Mobilization of CD34 cells in
elderly patients with multiple myeloma: influence of age, prior therapy,
platelet count and mobilization regimen. Br J Haematol. 2003;120:413-423.
506
12. Fried LP, Tangen CM, Watson J, et al. Fraility in older adults: evidence
for a phenotype. J Gerontol Biol Sci Med Sci. 2001;56:M146-M156.
13. Hurria A, Togawa Kayo, Supriya G, et al. Predicting chemotherapy
toxicity in older adults with cancer: a prospective multicenter study. J Clin
Oncol. 2011;29:3457-3465.
14. Avet-Loiseau H, Leleu X, Roussel M, et al. Bortezomib plus dexamethasone induction improves outcome of patients with t(4,14) myeloma but not
outcome of patients with del(17p). J Clin Oncol. 2010;28:4630-4634.
15. Neben K, Lokhorst H, Jauch A. et al. Administration of bortezomib
before and after autologous stem cell transplantation improves outcome in
multiple myeloma patients with 17p deletion. Blood. 2011;119:3-19.
16. Waheed S, Shaugnessy JD, van Rhee F, et al. International staging
system and metaphase cytogenetic abnormalities in the era of gene expression profiling data in multiple myeloma treated with total therapy 2 and 3
protocols. Cancer. 2011;117:1001-1009.
17. Mahindra A, Vesole D, Kalaycio M, et al. Autologous hematopoietic
stem cell transplantation is a safe and effective treatment for primary plasma
cell leukemia. Blood. 2009;114 (abstr 532).
18. Bensinger WI, Buckner CD, Anasetti C, et al. Allogeneic marrow
transplantation for multiple myeloma: An analysis of risk factors on outcome.
Blood. 1996;88:2787-2793.
19. Garban F, Attal M, Michallet M, et al. Prospective comparison of
autologous stem cell transplantation followed by dose-reduced allograft
(IFM99-03 trial) with tandem autologous stem cell transplantation
(IFM99-04 trial) in high-risk de novo multiple myeloma. Blood. 2006;107:
3474-3480.
20. Bruno B, Storer B, Patriarca F, et al. Long-term follow up of a
comparison of non-myeloablative allografting with autografting for newly
diagnosed myeloma. Blood. 2010;116 (abstr 525).
21. Krishnan A, Pasquini M, Logan B, et al. Autologous hematopoietic stem
cell transplantation followed by allogeneic or autologous hematopoietic stem
cell transplantation in patients with multiple myeloma (BMTCTN0102): a
phase 3 biological assignment trial. Lancet Oncol. 2011;12:1195-1203.
22. Bjorkstrand B, Iacobelli S, Hegenbart U, et al. Tandem autologous/
reduced-intensity conditioning allogeneic stem-cell transplantation versus

autologous transplantation in myeloma: long-term follow-up. J Clin Oncol.
2011;29:3016-3022.
23. Fermand JP, Ravaud P, Chevret S, et al. High-dose therapy and
autologous peripheral blood stem cell transplantation in multiple myeloma;
up-front or rescue treatment? Results of a multicenter sequential randomized
clinical trial. Blood. 1998;92:3131-3136.
24. Van Rhee F, Szymonika J, Anaissie E, et al. Total Therapy Three for
multiple myeloma: prognostic implications of cumulative dosing and premature discontinuation of VTD maintenance components bortezomib, thalidomide, and dexamethasone relevant to all phases of therapy. Blood. 2010;115:
1220-1227.
25. Ladetto M, Pagliano G, Ferrero S, et al. Major tumor shrinking and

persistent molecular remissions after consolidation with bortezomib, thalidomide, and dexamethasone in patients with autograft for myeloma. J Clin
Oncol. 2010;28:2077-2084.
26. Mellqvist UH, Ginsing P, Hjertner O, et al. Improved progression
free survival with bortezomib consolidation after high-dose melphalan; results of a randomized phase III trial. Haematologica. 2011;96:S31 (suppl;
abstr 11).
27. Rapoport A, Aqui N, Stadtmauer E, et al. Combination immunotherapy
using adoptive T cell transfer and tumor antigen vaccination on the basis of
HTERT and surviving after ASCT for myeloma. Blood. 2011;117:788-797.
507
Upfront Therapy for Myeloma: Tailoring

Therapy across the Disease Spectrum
By S. Vincent Rajkumar, MD
Overview: The treatment of multiple myeloma is evolving

rapidly. Despite the number of regimens and combinations
available, there is lack of data from phase III trials demonstrating superiority of one regimen over the other in terms of
overall survival and/or quality of life. The only clear survival
signals have come from studies that compared newer regimens with historic ones such as melphalan-prednisone (MP)
or vincristine-doxorubicin hydrochloride-thalidomide (VAD).
Thus, the choice of therapy at present is often made based on
physician discretion, bias, and limited data from phase II
studies. Further, the regimens available have considerably
ULTIPLE MYELOMA is a malignant plasma-cell

proliferative disorder that accounts for approximately 10% of all hematologic malignancies.1,2 Over 20,000
new cases are diagnosed annually in the United States, and
most patients continue to die of the disease. The risk of
multiple myeloma is two-fold higher in blacks compared
with whites. Multiple myeloma evolves from an asymptomatic premalignant stage termed monoclonal gammopathy of
undetermined significance (MGUS). In the spectrum of
plasma cell disorders, there is an additional intermediate
arbitrary clinical stagetermed smoldering multiple myeloma (SMM) that comprises of a mixture of patients with
premalignancy (i.e., MGUS) and malignancy (i.e., early
stage multiple myeloma). The SMM category is clinically
important because it is not possible in most instances to
determine which patient with SMM has premalignancy
compared with malignancy; therefore, these patients are
currently observed without therapy until overt signs or
malignancy develop. They are also candidates for clinical
trials, testing the value of early intervention. The diagnostic
criteria for multiple myeloma, MGUS, and SMM are given in
Table 1.3
Prognosis and Risk-Stratification
Prognosis in myeloma depends on four major factors: host

characteristics (age, performance status, comorbidities),
stage, disease aggressiveness, and response to therapy.4
Each of these four factors result in poor prognosis through
different mechanisms, and therefore the strategy to overcome each of these factors will need to be quite different. For
example, it is not possible to overcome the poor prognosis
associated with host factors such as advanced age or multiple comorbidities by increasing the aggressiveness of the
therapy used. Similarly, it is not possible to overcome the
poor prognostic effect conferred by acute renal failure without taking into account the specific drugs than can be used
safely, and the specific dose modifications that may be
required. In other words, one needs to know precisely the
mechanism by which a given factor confers its adverse
prognostic value before we can develop strategies to overcome such an effect.
Staging of myeloma by using the Durie-Salmon Staging
(DSS) or the International Staging System (ISS) both provide prognostic information, but are typically not helpful in
making therapeutic choices because the risk groups are
508
different profiles in terms of safety, convenience, and cost.

Given the dramatic variations in expected outcome depending
on the various known prognostic factors, a risk-adapted
strategy is required to provide the best available therapy to
each patient based on host factors as well as prognostic
markers of disease aggressiveness. This article reviews the
current status of myeloma therapy and risk stratification.
Results from major phase III trials are reviewed, and a
risk-adapted individualized approach to therapy is presented
and discussed.
heterogeneous and defy any rational uniform approach to

risk-adapted therapy. For instance, stage III ISS (defined on
the basis of a high beta-2 microglobulin level) consists of a
heterogeneous mix of patients, some of whom derive the
stage III label because of renal failure, whereas some meet
criteria for stage III because of a high tumor burden; both
renal failure and high tumor burden have the same effect on
beta-2 microglobulin. Thus staging, whereas useful for prognosis, is not particularly useful in deciding choice of therapy
in multiple myeloma; the one exception to this rule is stage
I Durie-Salmon disease, which is a heterogeneous group of
patients with either SMM (no therapy needed) or solitary
plasmacytoma (typically treated with radiation alone).
Response to therapy is also not useful in determining
choice of therapy since this information is only available
post-hoc, and there are no good reliable tests that allow us
to predict response to specific drugs ahead of time. Thus
risk-adapted, individualized therapy for myeloma is currently done primarily on the basis of two factors: host factors
and markers of disease aggressiveness.
Host factors are currently used in multiple myeloma
primarily to determine eligibility for autologous stem cell
transplantation (ASCT). In general, patients need to meet
minimum requirements for age, organ function, and performance status to be considered eligible for ASCT. The initial
therapy varies according to eligibility for transplantation.
Disease aggressiveness has a major effect on prognosis and
can also be used to individualize therapy. Table 2 provides a
risk stratification model based on markers of disease aggressiveness (mSMART).5 This requires fluorescent in situ hybridization (FISH) or similar studies be done on the bone
marrow at the time of initial diagnosis to detect t(11;14),
t(4;14), t(14;16), t(6;14), t(14;20), and deletion of 17p.5 Gene
expression profiling (GEP), if available, can provide additional prognostic value. Patients with standard-risk myeloma (75% of myeloma) have a median overall survival (OS)
of an excess of 7 years whereas those with high-risk disease
From the Division of Hematology, Mayo Clinic, Rochester, MN.

Address reprint requests to S. Vincent Rajkumar, MD, Division of Hematology, Mayo
Clinic, 200 First Street SW, Rochester, MN 55905; email: rajkumar.vincent@mayo.edu.
1092-9118/10/1-10
CURRENT STATUS OF MYELOMA THERAPY

Table 1. Diagnostic Criteria for Multiple Myeloma and Related Disorders
Disorder
Monoclonal gammopathy of undetermined

significance (MGUS)
Light chain MGUS
Smoldering multiple myeloma (also referred

to as asymptomatic multiple myeloma)
Solitary Plasmacytoma
Multiple myeloma
Disease Definition
All three criteria must be met:

Serum monoclonal protein 3gm/dL
Clonal bone marrow plasma cells 10%, and
Absence of end-organ damage such as hypercalcemia, renal insufficiency, anemia, and bone lesions (CRAB) that can be
attributed to the plasma cell proliferative disorder; or in the case of IgM MGUS, no evidence of anemia, constitutional
symptoms, hyperviscosity, lymphadenopathy, or hepatosplenomegaly that can be attributed to the underlying
lymphoproliferative disorder.
All criteria must be met:
Abnormal FLC ratio (0.26 or 1.65)
Increased level of the appropriate involved light chain (increased kappa FLC in patients with ratio 1.65 and increased
lambda FLC in patients with ratio 0.26)
Absence of end-organ damage such as hypercalcemia, renal insufficiency, anemia, and bone lesions (CRAB) that can be
attributed to the plasma cell proliferative disorder
No immunoglobulin heavy chain expression on immunofixation
Both criteria must be met:
Serum monoclonal protein (IgG or IgA) 3gm/dL and/or clonal bone marrow plasma cells 10%, and
Absence of end-organ damage such as lytic bone lesions, anemia, hypercalcemia, or renal failure that can be attributed
to a plasma cell proliferative disorder
All four criteria must be met
Biopsy-proven solitary lesion of bone or soft tissue with evidence of clonal plasma cells
Normal bone marrow with no evidence of clonal plasma cells
Normal skeletal survey and MRI of spine and pelvis (except for the primary solitary lesion)
Absence of end-organ damage such as hypercalcemia, renal insufficiency, anemia, or other bone lesions (CRAB) that
can be attributed to a lympho-plasma cell proliferative disorder
All three criteria must be met except as noted:
Clonal bone marrow plasma cells 10% and/or biopsy proven plasmacytoma
Presence of serum and/or urinary monoclonal protein (except in patients with true non-secretory multiple myeloma), and
Evidence of end organ damage that can be attributed to the underlying plasma cell proliferative disorder, specifically
o Hypercalcemia: Serum calcium 11.5 mg/dL or
o Renal insufficiency: Serum creatinine 1.73 mmol/L (or 2 mg/dL) or estimated creatinine clearance less than 40
mL/min
o Anemia: Normochromic, normocytic with a hemoglobin value of 2 g/dL below the lower limit of normal or a
hemoglobin value 10 g/dL
o Bone lesions: Lytic lesions, severe osteopenia or pathologic fractures
Modified from Kyle and Rajkumar. 1
KEY POINTS
The diagnosis of multiple myeloma requires 10% or

more clonal plasma cells on bone marrow examination or a biopsy-proven plasmacytoma plus evidence
of end-organ damage felt to be related to the underlying plasma cell disorder.
Patients with 17p deletion, t(14;16), t(14;20), or highrisk gene expression profiling signature have highrisk myeloma; patients with t(4;14) translocation
have intermediate-risk disease; all others are considered to have standard-risk myeloma.
A risk-adapted strategy to multiple myeloma is essential because prognosis varies dramatically across
risk groups, the various modern regimens in use have
different safety and cost profiles, and there are no
clear survival or quality-of-life data from randomized
trials showing a clear superiority of one strategy over
the other.
Risk-stratified therapy relies on both host factors as
well as prognostic markers of disease aggressiveness.
Bortezomib-based regimens appear to be able to overcome the poor prognostic effect of t4;14 translocation.
(15% of myeloma) have a median OS of 2 to 3 years, despite

initial therapy containing bortezomib, tandem ASCT, and
routine use of maintenance therapy.1
Risk-Adapted Therapy
With the number of regimens and combinations in myeloma available, and the lack of phase III trials demonstrating superiority in the only two endpoints that matter
(overall survival and/or quality of life), the choice of therapy
is often made based on physician discretion, bias, and
limited data from phase II studies. Further, the regimens
available have considerably different profiles in terms of
safety, convenience, and cost, thereby dictating the need for
Table 2. Risk-Stratification of Myeloma
A. Standard Risk
Hyperdiploidy
t (11;14)
t (6;14)
B. Intermediate Risk
t (4;14)
C. High Risk
17p deletion
t (14;16)
t (14;20)
High-risk gene expression profiling signature
Modified from Rajkumar. 2
509
S. VINCENT RAJKUMAR
Table 3. Rationale for Individualized, Risk-Adapted Therapy in Multiple Myeloma
Factor
Individualized Approach to Initial Therapy
Age, performance status, co-morbidities
Decision on Transplant Eligible versus Transplant Ineligible

ASCT offered to patients age 65 (or up to age 75 in
USA) who have adequate performance status and
acceptable organ function
Nature and duration of initial therapy varies
according to eligibility for ASCT
Regimens with the least impact on quality of life
Rd
MP
Rapidly acting regimens using drugs that are safe to use
in renal failure
VTD
VCD
Possible plasmapheresis
Possible hemodialysis
Goal of therapy is rapid reduction of involved serum-free
light chains to less than 50 mg/dL
Wide choice of induction regimens based on patient
preference, cost, toxicity, availability, etc.
Rd
VD
VTD
VCD
Other
Advanced age (80 yr)
Acute renal failure due to light chain

cast-nephropathy
Standard-risk myeloma
Intermediate-risk myeloma
High-risk myeloma
Plasma cell leukemia or multiple sites of

extramedullary disease at
presentation
Bortezomib-based induction required

VD
VTD
VCD
ASCT
Bortezomib-based maintenance therapy
Experimental approaches
Regimens with promising phase II data
VRD
Initial therapy with multi-agent chemotherapy such as
VDT-PACE
ASCT
Maintenance therapy
Rationale
The value of ASCT has been demonstrated in

randomized trials only in this patient population
Patient safety
Although phase III data are not available, oral

regimens such as Rd or MP are well tolerated and
effective in patients with advanced age
Rapid reversal of renal failure is possible with this
approach
Persistent renal failure carries an adverse prognosis
Patient safety dictates that in the presence of acute
renal failure, drugs that are renally excreted, such as
lenalidomide, are best avoided particularly when
reversal of renal failure is a stated goal
Little or no phase III data showing overall survival is
affected based on which of these induction regimens
are used as initial therapy
Cost, toxicity, and availability concerns do matter and
should be taken into consideration in the absence of
compelling overall survival data dictating the use of
one specific regimen over the other, given the
estimated median survival of 710 yr
Almost complete reversal of the poor prognostic impact
of t4;14 has been noted with this strategy
The median OS with approaches used in either

standard- or intermediate-risk disease produce a
median survival of only 23 yr
The median OS is poor with approaches used in either
standard- or intermediaterisk disease
VDT-PACE can control disease rapidly and predictably
Abbreviations: ASCT, autologous stem cell transplantation; MP, melphalan plus prednisone; Rd, lenalidomide plus low dose dexamethasone; VCD, bortezomib,
cyclophosphamide, dexamethasone; VD, bortezomib plus dexamethasone; VDT-PACE, bortezomib, dexamethasone, thalidomide, cisplatin, doxorubicin, cyclophosphamide, etoposide; VRD, bortezomib, lenalidomide, dexamethasone; VTD, bortezomib, thalidomide, dexamethasone.
a judicious approach to the use of the available options.

Thus, given the dramatic variations in prognosis depending
on the various known prognostic factors, a one-size-fits-all
approach is both unreasonable and obsolete in multiple
myeloma. Even skeptics of risk-adapted or individualized
therapy do in reality practice such an approach, albeit
subconsciously. Table 3 shows how therapy can be tailored
by using host factors and markers of disease aggressiveness
and the rationale for such a risk-adapted strategy.
A risk-adapted strategy does not and should not mean
giving suboptimal (weak) therapy to patients with standard
risk and reserving the best therapy for patients at high risk.
To the contrary, it means not subjecting (or recommending
to) patients with standard risk regimens that have not been
proven to be useful in randomized trials. For example, VRD
has shown promise in newly diagnosed myeloma in a phase
II trial. That is sufficient to design phase III trials with this
regimen, but insufficient to recommend as an option outside
a clinical trial setting to patients with newly diagnosed
myeloma, especially to patients with standard-risk disease
in whom the value of this regimen, given its cost and
toxicity, requires data from randomized trials. One could
510
argue, however, that the outcome of patients with high-risk

myeloma is so poor with current available (phase III-vetted)
regimens that a new regimen with high activity such as VRD
is reasonable despite the added cost and toxicity. In other
words, the principle of risk-adapted therapy is to individualize the treatment options for each patient in a manner
that carefully balances the risks and benefits depending on
prognosis and patient expectations. Patients with high risk
may be willing to take the added risks of unproven therapy
more often than patients with standard risk. Patients with
standard risk should be treated by using regimens that have
been tested and found effective in phase III trials.
More importantly, there is a critical ongoing cure compared with control debate in the myeloma community on
whether we should treat the disease with an aggressive
multidrug strategy targeting complete response (CR) or a
sequential disease control approach that emphasizes quality
of life as well as OS.6 A risk-adapted strategy allows patients with standard-risk myeloma to have the choice
between pursuing either approach since there are no randomized data demonstrating the clear superiority of one
approach over the other. It also highlights the need for such

Initial Treatment in Patients Eligible for ASCT
Fig 1. Approach to the treatment of newly diagnosed myeloma (A)

and in newly diagnosed patients with myeloma with special circumstances (B).
Abbreviations: ASCT, autologous stem cell transplantation; CR,
complete response; Dex, dexamethasone; Rd, lenalidomide plus
low-dose dexamethasone; VCD, bortezomib, cyclophosphamide,
dexamethasone; VDT-PACE, bortezomib, dexamethasone, thalidomide, cisplatin, doxorubicin, cyclophosphamide, etoposide; VGPR,
very good partial response; VRD, bortezomib, lenalidomide, dexamethasone; VTD, bortezomib, thalidomide, dexamethasone.
* For patients who choose delayed ASCT, continue Rd.
When continuing Rd, dexamethasone usually discontinued after
12 months, and continued long-term lenalidomide is an option for
patients who are tolerating treatment well.
trials. On the other hand, patients with high-risk require a

complete response for long-term survival and hence clearly
need an aggressive strategy designed to eradicate all plasma
cells, if possible.
Approach to Initial Therapy
The approach to treatment of symptomatic newly diagnosed multiple myeloma is outlined in Fig. 1 and is dictated
by host factors that govern eligibility for ASCT and markers
of disease aggressiveness that determine risk-stratification.1 The major regimens used for therapy and the data to
support their use are listed in Tables 4 and 5. In general,
all patients are probably best served by participation in
clinical trials, and every effort should be made to encourage
participation in such studies. The algorithms outlined in
this review are to be considered primarily when a suitable
clinical trial is unavailable or impractical.
Typically, patients eligible for ASCT are treated with

approximately two to four cycles of induction therapy before
stem cell harvest. In general, regardless of the regimen
used, once weekly subcutaneous bortezomib and dexamethasone at a 40-mg once-a-week schedule (low-dose dexamethasone) is preferred in most patients for initial therapy,
unless there is an urgent need for rapid disease control.
Several other regimens besides the ones discussed in this
review have been tested in newly diagnosed multiple myeloma, but there are no clear data from randomized controlled trials that they have an effect on long-term
endpoints.
Standard risk. Patients with standard risk who are eligible for ASCT can be treated with a variety of active regimens. The main options for initial therapy are lenalidomide
plus low-dose dexamethasone (Rd) or one of several
bortezomib-based regimens. Lenalidomide plus dexamethasone is active in newly diagnosed myeloma. Rd, which uses
low-dose dexamethasone, has less toxicity and better overall
survival than lenalidomide plus high-dose dexamethasone.7
One caveat is that Rd may impair collection of peripheral
blood stem cells for transplant in some patients when
mobilized with granulocyte colony stimulating factor (GCSF) alone. Thus, patients over the age of 65 and those who
have received more than four cycles of Rd) stem cells must be
mobilized with either cyclophosphamide plus G-CSF or with
plerixafor. All patients require antithrombosis prophylaxis
with aspirin; low molecularweight heparin or coumadin is
needed in patients at high risk of DVT.8
Several bortezomib-containing regimens can also be used
as initial therapy instead of Rd. Harousseau and colleagues
compared bortezomib plus dexamethasone (VD) with vincristine, doxorubicin hydrochloride, dexamethasone (VAD)
as pre-transplantinduction therapy.9 Post-induction verygood-partial-response (VGPR) was superior with VD compared with VAD, 38% compared with 15%, respectively.
However, progression-free survival (PFS) improvement was
modest, 36 months compared with 30 months, respectively,
and did not reach statistical significance. No overall survival
benefit was noted. Three-drug regimens such as bortezomibcyclophosphamide-dexamethasone (VCD), and bortezomibthalidomide-dexamethasone (VTD), and bortezomiblenalidomide-dexamethasone (VRD) are in various stages
of development.10 In randomized trials, VTD has shown
better response rates and PFS compared with TD,11 as well
as VD.12 VCD has impressive activity in newly diagnosed
multiple myeloma and is less expensive than either VTD or
VRD. Preliminary studies indicate that VCD is well tolerated and has similar activity compared with VRD, making
it an excellent choice when considering a bortezomibcontaining regimen for frontline use.13
There are no data so far comparing Rd with any of the
bortezomib-containing regimens. Results from a Southwest
Oncology Group (SWOG) randomized trial that compared
VRD to Rd are awaited. One drawback of bortezomibcontaining regimens is the risk of neurotoxicity early in the
disease course. However, recent studies show that the neurotoxicity of bortezomib can be greatly diminished by administering bortezomib using a once-weekly schedule,14,15 and
by administering the drug subcutaneously.16 Unlike lenalidomide, bortezomib does not appear to have any adverse
511
S. VINCENT RAJKUMAR
Table 4. Major Treatment Regimens in Multiple Myeloma
Regimen
Melphalan-prednisone (7-d schedule)
Thalidomide-dexamethasone**
Lenalidomide-dexamethasone
Bortezomib-dexamethasone**
Melphalan-prednisone-thalidomide
Bortezomib-melphalan-prednisone**
Bortezomib-thalidomide-dexamethasone**
Bortezomib-cyclophosphamide-dexamethasone** (VCD)
Bortezomib-lenalidomide-dexamethasone**
Usual Dosing Schedule*
Melphalan 810 mg oral days 17

Prednisone 60 mg/d oral days 17
Repeated every 6 wk
Thalidomide 200 mg oral days 128
Dexamethasone 40 mg oral days 1, 8, 15, 22
Repeated every 4 wk
Lenalidomide 25 mg oral days 121 every 28 d
Dexamethasone 40 mg oral days 1, 8, 15, 22 every 28 d
Repeated every 4 wk
Bortezomib 1.3 mg/m2 subcutaneous or intravenous days 1, 8, 15, 22
Dexamethasone 20 mg on day of and day after bortezomib (or 40 mg days 1, 8, 15, 22)
Repeated every 4 wk
Melphalan 0.25 mg/kg oral days 14 (use 0.20 mg/kg/d oral days 14 in patients over the age of 75)
Prednisone 2 mg/kg oral days 14
Thalidomide 100200 mg oral days 128 (use 100 mg dose in patients 75)
Repeated every 6 wk
Melphalan 9 mg/m2 oral days 14
Prednisone 60 mg/m2 oral days 1 to 4
Repeated every 35 d
Thalidomide 100200 mg oral days 121
Repeated every 4 wk 4 cycles as pre-transplant induction therapy
Cyclophosphamide 300 mg/m2 orally on days 1, 8, 15 and 22
Bortezomib 1.3 mg/m2 subcutaneous or intravenously on days 1, 8, 15, 22
Dexamethasone 40 mg orally on days on days 1, 8, 15, 22
Repeated every 4 wk
Bortezomib 1.3 mg/m2 subcutaneous or intravenous days 1, 8, 15
Lenalidomide 25 mg oral days 114
Repeated every 3 wk
Reproduced from Rajkumar. 2

* All doses need to be adjusted for performance status, renal function, blood counts, and other toxicities.
** Doses of dexamethasone and/or bortezomib reduced based on subsequent data showing lower toxicity and similar efficacy with reduced doses.
Omit day 22 dose if counts are low or when the regimen is used as maintenance therapy; when used as maintenance therapy for patients at high risk, delays can
be instituted between cycles.
Omit day 15 dose if counts are low or when the regimen is used as maintenance therapy; when used as maintenance therapy for patients at high risk , lenalidomide
dose may be decreased to 10 15 mg per day, and delays can be instituted between cycles as done in total therapy protocols. 17,28
effect on stem cell mobilization and does not cause DVT.

Since there are no data from randomized trials, the choice of
therapy is driven primarily by expert opinion, feasibility,
cost, and patient preference.
Based on cost and toxicity considerations, Rd or VCD are
probably the best options for this patient population at
present. After four cycles of therapy with Rd or VCD, stem
cells should be harvested, and patients can either undergo
frontline ASCT (preferred) or resume induction therapy,
delaying ASCT until first relapse.
Intermediate risk. In patients with standard risk, there
are no data from randomized trials that a bortezomibcontaining regimen provides superior OS compared with Rd,
or vice versa. By contrast, so far only bortezomib-containing
regimens appear to be able overcome the poor prognosis
associated with the t4;14 translocation.11 There are also
data that the best results in the t4;14 population come from
studies that used post-transplant bortezomib-based maintenance therapy, and double ASCT.
High risk. Even regimens as aggressive as total therapy 3
(bortezomib-based induction therapy, tandem autologous
transplantation, and bortezomib-based maintenance therapy) have not made any meaningful improvement in the
outcome of patients with high-risk disease. Some of these
512
patients may be content with the best quality of life possible,

given the grave prognosis, and choose a gentle approach
such as Rd. However, it would be reasonable to consider
expensive, albeit unproven, approaches such as VRD in this
patient population. It is not clear whether a full myeloablative allogeneic transplantation may be of value, but some
patients are willing to take on the high risk of transplantrelated mortality in exchange for a chance of long-term
survival.
Plasma cell leukemia or multiple extramedullary plasmacytomas. Multiagent combination chemotherapy should be
considered for these patients as initial therapy since rapid
and reliable disease control is essential. VDT-PACE (bortezomib, dexamethasone, thalidomide, cisplatin, doxorubicin,
cyclophosphamide, and etoposide) has been tested extensively as part of the total therapy 3 trials and is a valuable
option.17 Following two cycles of therapy, these patients
should be considered candidates for ASCT and subsequently
for maintenance therapy with a bortezomib-based regimen.
Options for Initial Treatment in Patients Not Eligible
for ASCT
In patients with newly diagnosed multiple myeloma who

are considered ineligible for ASCT because of age or other

Table 5. Results of Recent Randomized Studies in Newly Diagnosed Myeloma
Trial
Rajkumar et al
Harousseau et al9
Cavo et al11
Moreau et al12
Facon et al18
Hulin et al19
Wijermans et al22
Palumbo et al29
Waage et al21
San Miguel et al**26,30
Regimen
No. of
Patients
Overall
Response
Rate (%)
CR plus
VGPR (%)
Progression-free
Survival
(Median in Months)
RD
Rd
VAD
VD
TD
VTD
VD
VTD
MP
Mel 100
MPT
MP Placebo
MPT
MP
MPT
MP
MPT
MP Placebo
MPT
MP
VMP
223
222
242
240
238
236
99
100
196
126
125
116
113
168
165
164
167
175
182
331
337
81
70
63
79
79
93
81
90
35
65
76
31
62
45
66
48
69
33
34
35
71
50
40
15
38
28
62
35
51
7
43
47
7
21
10
27
11
29
7
23
8
41
19.1
25.3
30
36
40
NR
N/A
N/A
17.8
19.4
27.5
18.5
24.1
9
13
14.5
21.8
14
15
16.6
24
P for Progression
Free Survival
0.026
0.06
0.006
0.001
0.001
0.001
0.004
NS
0.001
3 yr Overall
Survival
Rate (%)*
Overall Survival
(Median in Months)
75
74
77
81
84
86
N/A
N/A
48
52
66
40
55
43
55
65
65
43
43
54
69
NR
NR
NR
NR
NR
NR
N/A
N/A
33.2
38.3
51.6
29.1
44
31
40
47.6
45
32
29
43
NR
P for
Overall
Survival
0.47
0.46
0.3
0.001
0.028
0.05
0.79
0.16
0.001
Reproduced from Rajkumar. 2

Abbreviations: CR, complete response; MP, melphalan plus prednisone; MPT, melphalan plus prednisone plus thalidomide; N/A, not available; NS, not significant; Rd,
lenalidomide plus dexamethasone; TD, thalidomide plus dexamethasone; VGPR, very good partial response; VMP, bortezomib plus melphalan plus prednisone; VTD,
bortezomib, thalidomide, dexamethasone.
* Estimated from survival curves when not reported.
** Progression-free survival not reported; numbers indicate time to progression.
comorbidities, the major options at present are either

alkylator- or bortezomib-based combination therapies given
for 9 to 18 months, or Rd.1 With Rd, the optimum duration
of therapy is unclear; most patients continue therapy until
progression, provided the treatment is well tolerated. In
such patients, dexamethasone is usually lowered to a minimal dose or discontinued after the first year. Results of a
randomized trial comparing Rd for 18 months compared
with Rd until progression are awaited. Thalidomide plus
dexamethasone is inferior to melphalan plus prednisone
(MP) in terms of overall survival and is not recommended.
The addition of lenalidomide to MP (MPR) does not improve
PFS compared with MP alone, and is also not recommended.
Standard risk. Six randomized studies have found that
melphalan, prednisone, thalidomide (MPT) improves response rates with MP.18-23 Four of these trials have shown a
prolongation of PFS with MPT,18-20,22 and an OS advantage
has been observed in the two Intergroupe Francophone
Myelome (IFM) trials and in the trial by Wijermans and
colleagues.18-22 Two meta-analyses of these randomized trials have been conducted and they show a clear superiority
of MPT over MP.24,25 Grade 3 4 adverse events occur in
approximately 55% of patients treated with MPT, compared
to 22% with MP.20
Bortezomib-containing combinations serve as an alternative to MPT as initial therapy in this patient population. In
a large phase III trial, bortezomib, melphalan, prednisone
(VMP) led to improved OS compared with MP.26 Neuropathy
is an important risk with VMP therapy; grade 3 neuropathy
occurred in 13% of patients compared to 0% with MP.26
Thus, as with transplant-eligible patients, the once-weekly,
subcutaneous schedule is preferred. There is no significant
advantage of either bortezomib, thalidomide, prednisone

(VTP) or VTD over VMP. VCD is a minor variation of the
VMP combination that can be used in place of VMP to
minimize side effects.
Rd is also an attractive option for the treatment of elderly
patients with newly diagnosed myeloma because of its
excellent tolerability, convenience, and efficacy. The threeyear OS rate with Rd in patients age 70 and older who did
not receive ASCT is 70%, and is comparable to results with
MPT and VMP. An ongoing phase III trial is currently
comparing MPT with Rd for 18 months compared with Rd
until progression.
No results from randomized trials are available yet comparing Rd, MPT, VMP or VCD. Although MPT and VMP
have the strongest data, they are also less well-tolerated
than Rd and VCD, respectively. Based on cost and toxicity
considerations, Rd or VCD are probably the best options for
this patient population at present.
Intermediate risk. Bortezomib-containing regimens can
overcome to some extent the poor prognosis associated with
the t4;14 translocation.11,17,27 As a result, in patients with
intermediate risk, bortezomib-containing initial therapy
such as VCD for approximately 24 months is preferred.
High risk, plasma cell leukemia, multiple extramedullary
plasmacytomas. Patients ineligible for ASCT with high-risk
disease, plasma cell leukemia or multiple extramedullary
plasmacytomas have poor prognosis. In these patients, depending on tolerability, it would be reasonable to consider
VCD or VRD as initial therapy. The duration of therapy
will be dependent on the tolerability of these regimens in a
given patient. If possible, some form of continuous therapy is
probably needed.
513
S. VINCENT RAJKUMAR
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
S. Vincent Rajkumar*
REFERENCES
1. Rajkumar SV. Treatment of Multiple Myeloma. Nature Rev Clin Oncol.
2011;8:479-491.
2. Rajkumar SV. Multiple myeloma: 2012 update on diagnosis, riskstratification, and management. American Journal of Hematology. 2012;87:
78-88.
3. Kyle RA, Rajkumar SV. Criteria for diagnosis, staging, risk stratification
and response assessment of multiple myeloma. Leukemia. 2009;23:3-9.
4. Russell SJ, Rajkumar SV. Multiple myeloma and the road to personalised medicine. Lancet Oncol. 2011;12:617-619.
5. Kumar SK, Mikhael JR, Buadi FK, et al. Management of newly diagnosed symptomatic multiple myeloma: updated Mayo stratification of
Myeloma and Risk-Adapted Therapy (mSMART) consensus guidelines. Mayo
Clinic Proceedings. 2009;84:1095-110.
6. Rajkumar SV, Gahrton G, Bergsagel PL. Approach to the treatment of
multiple myeloma: a clash of philosophies. Blood. 2011;118:3205-3211.
7. Rajkumar SV, Jacobus S, Callander NS, et al. Lenalidomide plus
high-dose dexamethasone versus lenalidomide plus low-dose dexamethasone
as initial therapy for newly diagnosed multiple myeloma: an open-label
randomised controlled trial. Lancet Oncol. 2010;11:29-37.
8. Palumbo A, Rajkumar SV, Dimopoulos MA, et al. Prevention of
thalidomide- and lenalidomide-associated thrombosis in myeloma. Leukemia.
2008;22:414-423.
9. Harousseau JL, Attal M, Avet-Loiseau H, et al. Bortezomib pPlus
dexamethasone is superior to vincristine plus doxorubicin plus dexamethasone as induction treatment prior to autologous stem-cell transplantation in
newly diagnosed multiple myeloma: results of the IFM 2005-01 phase III
10. Richardson PG, Weller E, Lonial S, et al. Lenalidomide, bortezomib,
and dexamethasone combination therapy in patients with newly diagnosed
multiple myeloma. Blood. 2010;116:679-686.
11. Cavo M, Tacchetti P, Patriarca F, et al. Bortezomib with thalidomide
plus dexamethasone compared with thalidomide plus dexamethasone as
induction therapy before, and consolidation therapy after, double autologous
stem-cell transplantation in newly diagnosed multiple myeloma: a randomised phase 3 study. Lancet. 2010;376:2075-2085.
12. Moreau P, Facon T, Attal M, et al. Comparison of reduced-dose
bortezomib plus thalidomide plus dexamethasone (vTD) to bortezomib plus
dexamethasone (VD) as induction treatment prior to ASCT in de novo
multiple myeloma (MM): Results of IFM2007-02 study. J Clin Oncol. 2010;28
13. Kumar S, Flinn IW, Richardson PG, et al. Novel Three- and Four-Drug
Combination Regimens of Bortezomib, Dexamethasone, Cyclophosphamide,
and Lenalidomide, for Previously Untreated Multiple Myeloma: Results From
the Multi-Center, Randomized, Phase 2 EVOLUTION Study. Abstract presented at ASH Annual Meeting; December 2010; Orlando, FL.
14. Mateos MV, Oriol A, Martnez-Lopez J, et al. Bortezomib, melphalan,
and prednisone versus bortezomib, thalidomide, and prednisone as induction
therapy followed by maintenance treatment with bortezomib and thalidomide
versus bortezomib and prednisone in elderly treated patients with untreated
multiple myeloma: a randomized trial. Lancet Oncol. 2010;11:934-941.
15. Palumbo A, Bringhen S, Rossi D, et al. Bortezomib-melphalanprednisone-thalidomide followed by maintenance with bortezomibthalidomide compared with bortezomib-melphalan-prednisone for initial
treatment of multiple myeloma: a randomized controlled trial. J Clin Oncol.
2010;28:5101-5109.
16. Moreau P, Pylypenko H, Grosicki S, et al. Subcutaneous versus intra-
514
venous administration of bortezomib in patients with relapsed multiple

myeloma: a randomised, phase 3, non-inferiority study. Lancet Oncol. 2011;
12:431-440.
17. Barlogie B, Anaissie E, van Rhee F, et al. Incorporating bortezomib
into upfront treatment for multiple myeloma: early results of total therapy 3.
British Journal of Haematology. 2007;138:176-185.
18. Facon T, Mary JY, Hulin C, et al. Melphalan and prednisone plus
thalidomide versus melphalan and prednisone alone or reduced-intensity
autologous stem cell transplantation in elderly patients with multiple myeloma (IFM 99-06): a randomised trial. Lancet. 2007;370:1209-1218.
19. Hulin C, Facon T, Rodon P, et al. Efficacy of melphalan and prednisone
plus thalidomide in patients older than 75 years with newly diagnosed
multiple myeloma: IFM 01/01 Trial. J Clin Oncol. 2009;27:3664-3670.
20. Palumbo A, Bringhen S, Caravita T, et al. Oral melphalan and
prednisone chemotherapy plus thalidomide compared with melphalan and
prednisone alone in elderly patients with multiple myeloma: randomised
controlled trial. Lancet. 2006;367:825-831.
21. Waage A, Gimsing P, Fayers P, et al. Melphalan and prednisone plus
thalidomide or placebo in elderly patients with multiple myeloma. Blood.
2010;116:1405-1412.
22. Wijermans P, Schaafsma M, Termorshuizen F, et al. Phase III study
of the value of thalidomide added to melphalan plus prednisone in elderly
patients with newly diagnosed multiple myeloma: the HOVON 49 Study.
J Clin Oncol. 2010;28:3160-3166.
23. Beksac M, Haznedar R, Firatli-Tuglular T, et al. Addition of thalidomide to oral melphalan/prednisone in patients with multiple myeloma not
eligible for transplantation: results of a randomized trial from the Turkish
Myeloma Study Group. Eur J Haematol. 2011;86:16-22.
24. Kapoor P, Rajkumar SV, Dispenzieri A, et al. Melphalan and prednisone versus melphalan, prednisone and thalidomide for elderly and/or transplant ineligible patients with multiple myeloma: A meta-analysis. Leukemia.
2011;25:689-696.
25. Fayers PM, Palumbo A, Hulin C, et al. Thalidomide for previously
untreated elderly patients with multiple myeloma: meta-analysis of 1685
individual patient data from 6 randomized clinical trials. Blood. 2011;118:
1239-1247.
26. San Miguel JF, Schlag R, Khuageva NK, et al. Bortezomib plus
melphalan and prednisone for initial treatment of multiple myeloma. N Engl
J Med. 2008;359:906-917.
27. Nair B, van Rhee F, Shaughnessy JD Jr., et al. Superior results of Total
Therapy 3 (2003-33) in gene expression profiling-defined low-risk multiple
myeloma confirmed in subsequent trial 2006-66 with VRD maintenance.
Blood. 2010;115:4168-4173.
28. van Rhee F, Szymonifka J, Anaissie E, et al. Total therapy 3 for
multiple myeloma: Prognostic implications of cumulative dosing and premature discontinuation of VTD maintenance components, bortezomib, thalidomide and dexamethasone, relevant to all phases of therapy. Blood. 2010;116:
1220-1227.
29. Palumbo A, Bringhen S, Liberati AM, et al. Oral melphalan, prednisone, and thalidomide in elderly patients with multiple myeloma: updated
results of a randomized controlled trial. Blood. 2008;112:3107-3114.
30. Mateos MV, Richardson PG, Schlag R, et al. Bortezomib plus melphalan and prednisone compared with melphalan and prednisone in previously untreated multiple myeloma: updated follow-up and impact of
subsequent therapy in the phase III VISTA trial. J Clin Oncol. 2010;28:22592266.
Maintenance Therapy for Myeloma: How

Much, How Long, and at What Cost?
By Michel Attal, MD, and Murielle Roussel, MD
Overview: Maintenance therapy in multiple myeloma has

been under investigation for more than 3 decades and has
been without evidence of clear advantage in terms of
progression-free survival (PFS) until the mid-2000s. Neither
conventional chemotherapy, prednisone, nor interferon-based
maintenance regimens offered any benefit after conventional
or high-dose therapy. Thalidomide was the first drug, mainly
given as maintenance after high dose therapy, to demonstrate
clinical benefits in terms of PFS and, in some studies, of
overall survival (OS). The role of other novel agents such as
lenalidomide and bortezomib as maintenance therapy is
emerging. Lenalidomide has been shown to reduce the risk of
relapse with longer follow-up needed to see if this will
HERAPEUTIC ADVANCES in the treatment of multiple myeloma have improved remission duration and
OS.1 While agents such as thalidomide, bortezomib, and
lenalidomide have had a major effect on response rates, most
patients with multiple myeloma will ultimately relapse.
Further strategies are needed to both improve response
depth and prolong response duration in order to induce a
clonal extinction and, at last, cure the disease.
Maintenance therapy can be defined as any treatment
given after completion of induction treatment, for a prolonged period of time, with the goal of extending the duration of response, prolonging PFS and OS, while maintaining
a good quality of life.2,3 In addition, an optimal maintenance
treatment should have an acceptable toxicity profile and not
compromise treatment at time of relapse. Unfortunately,
this strategy remains elusive in multiple myeloma.
Early attempts at maintenance therapy with conventional
chemotherapy has always failed to prolong PFS and OS4 and
the role of long-term steroids has also been controversial.5
The effect of alpha-interferon was shown to be modest in
terms of prolonging PFS and OS.6,7 Most investigators
concluded that the benefit was small and needed to be
balanced against the cost and potential toxicity of prolonged
treatment with alpha-interferon.
The advent of novel agents (thalidomide, bortezomib,
and lenalidomide) renewed the concept of maintenance or
long-term treatment, even for elderly patients or those not
eligible for high dose therapy (HDT). Several phase II and
III trials on maintenance therapy are ongoing.
Maintenance Therapy with Immunomodulatory
Drugs (IMiDs)
The IMiDs thalidomide and lenalidomide are the most

frequently studied maintenance drugs, given their ease of
oral administration and established antimyeloma efficacy
(Table 1).
translate into a survival benefit. At present, a number of key

questions remain unanswered. What are the optimal dose and
duration of those treatments? Is the risk of toxicity and
second primary malignancies acceptable? Will the disease be
more aggressive at time of relapse? Is the clinical benefit
predicted by initial prognostic factors and response to previous therapy? Does maintenance therapy work by further
eradication of minimal residual disease or by immunological
control of the malignant clone? Ongoing randomized trials are
evaluating lenalidomide and bortezomib, both in the transplant and nontransplant settings, to better define the role of
these drugs as maintenance in multiple myeloma.
spite differences, such as duration of treatment and/or

concomitant use of prednisone/prednisolone, thalidomide
maintenance was consistently associated with a longer PFS.
The benefit with respect to OS, however, was variable.
Along with the Arkansas group, the Intergroupe Francophone du Myelome (IFM) was one of the first to show that
thalidomide as maintenance after tandem autologous stem
cell transplantation (ASCT) was superior to no maintenance
or pamidronate alone (IFM 9902 trial9). Continuous thalidomide increased the complete response (CR) and very good
partial response (VGPR) rate (67% vs. 55% and 57%, respectively; p 0.03), the 3-year PFS (52% vs. 36% and 37%,
respectively; p 0.009), and the 4-year OS (87% vs. 77% and
74%, respectively; p 0.04). The Australian group10 obtained similar results when comparing thalidomide (for 12
months) plus prednisolone (until progression) with prednisolone alone. The thalidomide-containing arm achieved
higher rates of VGPR with increased PFS and OS. Within
the total therapy 2 program,8 thalidomide was given as
maintenance until disease progression or intolerance. In the
initial report, CR rate and 5-year PFS were notably better in
the thalidomide arm (62% vs. 43% and 56% vs. 44%, respectively) but there was no improvement in OS. After relapse,
survival was shorter in patients pre-exposed to thalidomide.
However, in an updated analysis (with a median follow-up of
72 months), prolonged OS was eventually confirmed even in
a subgroup of patients with poor-risk cytogenetics.14
Lastly, a recently presented meta-analysis of five transplant studies revealed a significant improvement both in
PFS (p 0.001) and OS (p 0.002) with thalidomide
maintenance therapy.15
Lenalidomide. Given the more favorable efficacy profile,
lenalidomide was considered an attractive agent for maintenance. This has prompted several ongoing trials designed
to compare continuous treatment until relapse with no
maintenance after ASCT. Two large randomized phase III
Thalidomide and Lenalidomide Prolong Duration of Response

after HDT
Thalidomide. There are five randomized studies with

thalidomide maintenance after HDT that have been published to date; one additional trial was reported during the
American Society of Hematology meeting in 2010.8-13 De-
From the Hematology Department, CHU Purpan, Toulouse, France.

Address reprint requests to M. Attal, MD, Hematologie Clinique, CHU Purpan, Place du
Dr Baylac, TSA 40031, 31059 Toulouse Cedex 9, France; email: attal.m@chu-toulouse.fr.
1092-9118/10/1-10
515
ATTAL AND ROUSSEL

16
trials, one conducted by the IFM and the second by the

Cancer and Leukemia Group B (CALGB),17 were updated
during the last International Myeloma Workshop in Paris,
France.
The IFM 2005 02 phase III trial16 included 614 patients
under age 65 who were randomly selected after HD melphalan 200 mg/m2 followed by ASCT and consolidation with
lenalidomide (25 mg/day, 21 days/month for 2 months) to
receive either maintenance therapy with lenalidomide (10 to
15 mg/day, 28 days/month) or placebo until relapse. The first
interim analysis showed a statistically significant (p 107)
benefit favoring maintenance therapy with lenalidomide
and the study was unblinded in July 2010. There was no
planned cross-over and all patients stayed in their allocated
arm until January 2011. After a median follow-up of 36
months from randomization, median PFS was 41 months for
patients in the lenalidomide arm versus 24 months in the
placebo arm (p 0.0001). Thus, 3 years after randomization,
61% of patients in the lenalidomide maintenance group did
not relapse versus 34% in the placebo group. The benefit of
lenalidomide maintenance in terms of PFS, the primary
objective of this study, was notable, but there was no OS
benefit recorded (4-year OS estimates in the lenalidomide
arm were 79% vs. 73% in the placebo arm). The benefit of
lenalidomide maintenance was confirmed by the CALGB
100104 study.17 The protocol was similar to the IFM study
with the exception of consolidation component. This trial
was also closed prematurely because of a significant advantage in the lenalidomide arm in terms of time to progression
(TTP) calculated from the day 0 of intensification. The study
was unblinded in December 2009 and 86 patients in the
placebo arm went on to receive lenalidomide. Based on an
intent-to-treat analysis, after a median follow-up of 28
KEY POINTS
516
Maintenance therapy with thalidomide can improve

the depth of response and reduce the risk of progression or death in both intensive and nonintensive
programs.
On average, patients tolerate thalidomide for no more
than 1 year; the optimal dose of thalidomide should
be the minimal effective dose that is associated with
superior tolerance and less toxicity (often around 50
mg/day).
Maintenance therapy with lenalidomide is clearly
effective in prolonging the duration of response and
diminishes the risk of relapse by 35% after frontline
therapy, irrespective of the patients age; however,
patients treated with lenalidomide have the ongoing
risk of new cancer and no clear OS benefit has yet
been shown.
Bortezomib maintenance therapy can also increase
response rate and prolong PFS after initial therapy
but questions regarding scheduling, duration of therapy, and combination with other drugs, remain.
At present, none of the drugs evaluated are approved
for maintenance therapy.
months, the TTP was 48 months in the lenalidomide arm

versus 31 months in the placebo arm (p 0.0001).
One can thus conclude that lenalidomide maintenance can
reduce the risk of relapse in the transplant setting.
What Group of Patients Will Benefit from IMiD Maintenance?
Thalidomide. A number of concerns exist regarding the

use of thalidomide maintenance after ASCT, which include
the presence of adverse cytogenetic abnormalities or the
quality of response after HDT.
In the IFM 99 02 trial,9 only patients who failed to
achieve at least VGPR and who lacked the del 13q cytogenetic abnormality had a longer PFS after receiving the
thalidomide. Within the intensive pathway of the British
MRC IX trial,13 thalidomide maintenance had no effect on
PFS (9 months vs. 12 months, p 0.48) in patients with
adverse iFISH [t (4; 14), t (14; 16), del17p, gain (1q21)] and
was associated with worse OS (p 0.009). Conversely in the
Australian trial,10 the PFS advantage with thalidomide
maintenance was irrespective of response to HDT or cytogenetic abnormalities.
On balance however, the evidence would support avoiding
thalidomide maintenance in patients with adverse cytogenetics.
Lenalidomide. Interestingly, with lenalidomide maintenance, subgroup analysis has shown that the benefit of
maintenance therapy was seen irrespective of response to
HDT and/or consolidation and initial prognostic factors. In
the IFM 2005 02 trial,16 3-year PFS estimates were higher
in all stratified subgroups of patients who received lenalidomide maintenance compared with those who received placebo. This included patients who had achieved VGPR at time
of randomization (64% vs. 49%; p 0.004) or not (51% vs.
18%; p 0.0001), and patients with 13q deletion (53% vs.
24%; p 0.0001) or not (67% vs. 44%; p 0.0001).
Can We Ultimately Prolong the OS with
IMiD Maintenance?
The shorter postrelapse survival observed in several studies may be caused by a number of factors, such as the
duration of maintenance treatment, selection of more resistant clones (especially in high-risk patients or patients who
received IMiDs as part of the induction), the age of patients
at time of relapse, toxicities from previous treatments, and
the availability of salvage treatments. The Australian trial10
showed that maintenance thalidomide for only 1 year did not
adversely affect the outcome after relapse. However, three
additional studies suggested that the long-term use of thalidomide may induce more resistant disease at relapse.8,11,13
Interestingly, within the meta-analysis described previously, outcomes did not differ between trials that used
thalidomide during the maintenance phase only and those
that used thalidomide both for induction and maintenance
treatment.15 For OS, a major effect variability between
trials was noted (test for heterogeneity, p 0.03) and the
positive result for overall effect must be interpreted with
caution.
Conversely, based on our own experience, it seems that
continuous therapy with lenalidomide does not have a negative influence on survival after relapse. It will be important
to determine whether patients given lenalidomide maintenance will have same median OS at first progression and
MAINTENANCE THERAPY FOR MULTIPLE MYELOMA

Table 1. IMiDs Thalidomide and Lenalidomide Maintenance Studies after ASCT and Conventional Chemotherapy
Trial
Author and Year
n
Age
Maintenance Dose
Comparator
Duration of
Maintenance
PFS/EFS
OS
Survival after Relapse
Thalidomide Trials
Post ASCT
IFM 9502
Attal et al 2006
Total therapy 2
Barlogie et al
20062008
Thal
597
Thalidomide
Pamidronate or Until progression 52%*
200400 mg/d
Observation
(3-yr from
65 y Pamidronate
randomization)
668
Thalidomide
100 mg/d
1st year
then 50 mg
alternate days
75 y IFN
IFN
Thal
75%*
75%
(1-yr OS)
Thal
75%
44%
Not reached
(median, yr)
7 yr
27%
(5-yr OS)
23%
42%*
(3-yr PFS)
23%*
86%*
(3-yr OS)
75%*
79%
(1-yr OS)
77%
28 mo*
(median, mo)
17 mo
NR
(median, yr)
Thalidomide
Prednisolone
100200 mg/d
Spencer et al 2009 70 y Prednisolone
12 mo
(prednisolone
until progression)
NCIC CTG MY.10
332
48 mo or until
progression
Stewart et al 2010
Thal
87%*
(4-yr from
enrollment)
56%*
(5-yr)
243
Observation
Thal
37%*
Until progression
ALLG MM6
Thalidomide
200 mg/d
65 y prednisone
Thal
5y
N/A
HOVON-50
556
Thalidomide
Lokhorst et al 2010 65 y 50 mg/d
IFN
Until progression
34 mo*
(median, mo)
25 mo*
73 mo
(median, mo)
60 mo
20 mo*
31 mo*
(median, mo)
MRC myeloma IX
Morgan et al
2012
Observation
Until progression
30 mo*
(median, mo)
27 mo*
75%
(3-yr OS)
80%
20 mo*
36 mo*
(median, mo)
Thal
Thal
MPT Thalidomide MP
100 mg/d
observation
Until progression
22 mo*
(median, mo)
14.5 mo*
MP
observation
Until progression
34%*
(2-yr PFS)
14%*
MP
placebo
Until progression
15 mo
(median, mo)
Thalidomide
200 mg/d
65 y IFN
IFN
Until progression
327
Thalidomide
65 y 50100 mg/d
Observation
Until progression
493
Thalidomide
65 y 50100 mg/d
Post Chemotherapy
GIMEMA
Palumbo et al
2008
HOVON 49
Wijermans et al
2010
NMSG
Waage et al 2010
CEMSG
Ludwig et al 2010
MRC myeloma IX
Morgan et al
2012
331
Thal
Thal
Thal
Thal
48 mo
11.5 mo*
24 mo*
40 mo*
(median, mo)
31 mo*
N/A
14 mo
29 mo
(median, mo)
32 mo
N/A
28 mo*
(median, mo)
13 mo*
53 mo
(median, mo)
51 mo
8 mo
25 mo
(median, mo)
11 mo*
(median, mo)
9 mo*
38 mo
(median, mo)
39 mo
21 mo
26 mo
(median, mo)
Len
Len
Len
Len
45 mo
(median, mo)
65 y
333
MPT
65 y Thalidomide
50 mg/d
357
MPT
65 y Thalidomide
200 mg/j
128
Lenalidomide Trials
Post ASCT
Len
IFM 200502
Attal et al 2011
614
65y
Lenalidomide
1015 mg/d
placebo
Until progression
41 mo*
(median, mo)
24 mo*
79%
(5-yr OS)
73%
CALBG 100104
McCarthy et al
2011
568
65y
Lenalidomide
1015 mg/d
placebo
Until progression
43 mo*
31 mo*
(median, mo)
N/A
N/A
23 versus
39 deaths*
Len
Len
Len
Len
Len
Until progression
31 mo*
(median, mo)
14/13 mo*
73%
(3-yr OS)
65%
N/A
Post chemotherapy
MM 015
Palumbo et al
2011
499
MPR
MPR or MP
65 y Lenalidomide
observation
10md/d 2128d
Len
12 mo
12 mo
(median, mo)
Len
Abbreviations: PFS, progression-free survival; EFS, event-free survival; OS, overall survival; y, years; mo, months; N/A, not available.
* Statistically significant.
517
ATTAL AND ROUSSEL

Table 2. Bortezomib Maintenance Studies after ASCT and Conventional Chemotherapy
Trial
Author and Year
n
age
Maintenance Dose
Duration of
Maintenance
Comparator
PFS/EFS
OS
Bortezomib Trials
Post ASCT
Bor
Bor
Bor
Bor
HOVON 65/HD4
Sonneveld et al 2010
800
65y
Bortezomib
1.3 mg/m2/2 wk
Thal 50
2 yr
48%*
(3-yr PFS)
42%*
78%*
(3-yr OS)
71%*
PETHEMA/GEM
Rosinol et al 2011
266
65y
Bortezomib
1.3 mg/m2/3 mo
Thalidomide 100 mg/d
Thalidomide 100 mg/d

Or
IFN
3 yr
78%*
(2-yr PFS)
63/49%*
N/A
No difference
GIMEMA
Palumbo et al 2010
511
65 y
VMPT
VT
Bortezomib 1.3 mg/m2,
d1,15/4 wk
Thalidomide 50 mg/d
VMP
observation
NR*
27 mo*
89%
(3-yr OS)
GEM2005MAS65
Mateos et al 2011
260
65 y
VMP or VTP
VT
Bortezomib 1.3 mg/m2
d 1, 4, 8, 11/3 mo
Thalidomide 50 mg/d
VMP or VTP
VP
Bortezomib 1.3 mg/m2
VT 39 mo*
VP 32 mo*
VT NR
VP 60 mo
(median, mo)
(median, mo)
Post Chemotherapy
d 1, 4, 8, 11/3 mo
Prednisone 50 mg
alternate day
Until
progression
3 yr
87%
Abbreviations: PFS, progression-free survival; EFS, event-free survival; OS, overall survival; y, years; mo, months; N/A, not available.
* Statistically significant.
will respond to increased doses of lenalidomide (25 mg) and

the addition of dexamethasone at the time of myeloma
progression. Longer follow-up is currently needed to assess
that issue.
Can Maintenance Therapy with IMiDs Also Prolong
Duration of Response in Elderly Patients or Those
Ineligible for HDT?
Fewer data are available concerning maintenance therapy

with IMiDS in elderly patients and in the nontransplant
setting.
Thalidomide. Three melphalan, prednisone, and thalidomide (MPT) trials were reported that included thalidomide
maintenance (MPTT).18-20 The MPTT schedule was compared to melphalan and prednisone (MP) with placebo and
thus it may be difficult to determine the value of maintenance therapy per se as the studies were designed to
evaluate the entire treatment program. Median PFS was
improved in MPTT arm in 2 of the 3 trials but there was
no OS advantage with long-term thalidomide use. The MRC
Myeloma IX maintenance study, was properly designed to
evaluate the influence of thalidomide maintenance in the
nonintensive pathway.13 Results indicated that thalidomide
had only a marginal effect in improving PFS in nonintensively treated patients (median PFS 11 months vs. 9
months, p 0.014). Furthermore, it suggests that it may be
related to thalidomide exposure at some point during the
treatment course rather than directly attributable to its use
in maintenance.
Lenalidomide. Prolonged treatment with lenalidomide
has shown a clear benefit in elderly patients, making it a
good choice for long-term maintenance. Several trials are
examining this role with the most mature data emerging
from the International study MM015.21
In this large, phase III trial, a total of 459 patients older
518
than or equal to age 65 with newly diagnosed multiple

myeloma (MM) were enrolled. Induction consisted of nine
28-day cycles of melphalan 0.18 mg/kg (d1 4), prednisone
2 mg/kg (d1 4), and lenalidomide 10 mg (d121) (MPR) or
MP. After induction, patients receiving MPR-R received
lenalidomide 10 mg (d121) maintenance until progression;
patients receiving MPR and MP received placebo. The MPR
regimen resulted in significantly higher response rates
(MPR-R: 77%, MPR: 68%, MP: 50%, p 0.001). Sixty percent of responses were achieved within 3 months following
induction treatment initiation. Further improvements in the
quality of response occurred with continued treatment, particularly during the first year, with a few patients achieving
further tumor reduction thereafter. A landmark analysis
showed that the addition of lenalidomide maintenance to
MPR decreased the risk of progression by 68% (p 0.001)
and significantly prolonged the median PFS (MPR-R 31
months vs. MPR 15 months (p 0.001) and MP 12 months
(p 0.001), respectively). In addition, a subgroup landmark
analysis showed that the benefit of MPR-R over MPR was
maintained regardless of Internatinal Staging System (ISS)
stage (ISS I and II vs. III), response ( VGPR vs. PR) and
age (6575 vs. 75 years old). With a median follow-up of
41 months, the estimated 4-year OS was similar between the
3 groups (58% to 59%).
Maintenance Therapy with Proteasome Inhibitors
Data concerning maintenance with bortezomib are less

mature. Several randomized studies by European and
American study groups are ongoing. The IV formulation
makes bortezomib a less attractive option for long-term
treatment. However, SC administration and the future
availability of oral forms of proteasome inhibitors will certainly boost potential maintenance trials with these agents
(Table 2).
Only two phase III trials of bortezomib maintenance

post-ASCT presenting PFS and OS data have been reported
to date.22,23 In the HOVON 65 MM/GMMG-HD4 trial,22
bortezomib 1.3 mg/m2 was given every 2 weeks for 2 years
after bortezomib-adriamycin-dexamethasone (PAD) induction and one or two transplantations. All outcomes were
significantly better in the bortezomib-containing arm (3year PFS: 48% vs. 42%, p 0.047 and OS: 78% vs. 71%, p
0.048) compared with the control arm using vincristineadriamycin-dexamethasone (VAD) induction, ASCT, and
low-dose thalidomide maintenance (50 mg daily). Patients
with high-risk MM derived a particular benefit from bortezomib: the 3-year PFS for patients with t(4;14) was 32%
compared with 22% in the controls. Unfortunately, the
design of the study does not allow a clear dissection of the
role of bortezomib maintenance therapy.
The Spanish group PETHEMA/GEM23 recently presented
updated results of their large phase III randomized trial
comparing induction with thalidomide/dexamethasone (TD)
versus bortezomib/thalidomide/dexamethasone (VTD) versus alternate polychemotherapy regimen with vincristine/
melphalan,
cyclophosphamide/prednisone,
vincristine/
carmustine/doxorubicin/dexamethasone, and bortezomib
(VBMCP/VBAD/B) in patients less than age 65 with newly
diagnosed symptomatic MM. The maintenance program
after HDT consisted of bortezomib and thalidomide (VT; 1
cycle of bortezomib-1.3 mg/m2 on days 1, 4, 8, and 11 every
3 months plus thalidomide 100 mg daily) compared with
thalidomide (T; 100 mg daily) compared with alfa2b-IFN
(subcutaneous, 3 MU 3 per week). The planned maintenance duration was 3 years or until disease progression or
toxicity.
The CR rate with maintenance was improved by 23% with
VT, 11% with T and 19% with alfa2-IFN (p not significant). After a median follow-up of 24 months, the PFS was
significantly longer with VT compared with T and alfa2-IFN
(PFS at 2 years: 78% vs. 63% vs. 49%, respectively, p 0.01).
However, OS was not significantly different among the 3
arms.
For elderly or nontransplant eligible patients, three large
phase III trials studied the role of bortezomib maintenance
alone or in combination with either T or prednisone.24
UPFRONT24,25 is a randomized, open-label, multicenter,
phase IIIb study conducted in the U.S. community practice
setting, designed to evaluate the use of single-agent bortezomib as a maintenance therapy following bortezomib/
dexamethasone (VD), VTD, or bortezomib/melphalan/
prednisone (VMP) induction in newly diagnosed MM
patients older than age 65 or ineligible for HDT. Bortezomib
1.6 mg/m2 was given once a week for five 35-day cycles. After
a median follow-up of 26 months, the 1-year PFS estimates
for VD, VTD, and VMP were 57.4%, 63.8%, and 67.3%,
respectively. Additonally, the 2-year OS estimates for VD,
VTD, and VMP were 73.7%, 73.6%, and 77.6%, respectively.
In the GEM2005MAS65, the Spanish myeloma group
investigated, a 3-year maintenance program with bortezomib/thalidomide (VT) or bortezomib/prednisone (VP)
(bortezomib: 1,3 mg/m2/d 1,4,8,11/3 months; thalidomide:
50 mg/d; prednisone: 50 mg alternating days). Updated data
were recently presented.25 This maintenance regimen increased the CR from 24% to 42%. After a median follow-up
of 46 months, there was no difference between VT and VP
maintenance with respect to PFS (39 months vs. 32 months)
or OS (not reached vs. 60 months). Neither maintenance

regimens overcame the poor prognosis of high-risk cytogenetics.
Maintenance with VT was also tested in an Italian study
comparing VMPT/VT with VMP. Median PFS was not
reached in the VT continuous therapy versus 27 months in
the no maintenance arm.26
Does Maintenance Therapy Enhance the Depth of
Response and/or Control a Subclinical Relapse?
All published trials with thalidomide and bortezomib

maintenance have shown that response rates can be increased but it is unclear whether this is the result of a late
consolidation effect within the first months of maintenance
(as proposed in the thalidomide trials) and/or the result of a
gradual clearance of residual disease over time. Further
trials examining sequential minimal residual disease evaluation are needed to address this issue.
With lenalidomide maintenance, less data concerning response rates are available. However, it appears that lenalidomide may also further enhance the depth of response
and not only control the residual tumor cells through active
immune surveillance. Within the IFM 2005 02 trial, the
best response during maintenance therapy was slightly, but
not significantly, higher (CR: 25% vs. 22%, p 0.4; VGPR:
77% vs. 70%, p 0.08).16 Nevertheless, the IFM recently
reported the results of a phase II study suggesting that
lenalidomide maintenance can increase response rate in
nearly 30% of patients, even up to stringent CR based on
flow assessment.27
Can Those Agents Be Administered Safely for a Long
Period of Time? The Benefit/Risk Ratio
There is a urgent need for future studies aimed at establishing the appropriate dose and optimal duration of maintenance therapy in both intensive and nonintensive
programs (Table 3).
Despite the longer PFS, the major caveat that precludes
widespread use of thalidomide maintenance is the toxicity
related to long-term administration. Peripheral neuropathy,
sedation, and constipation are common and often lead to a
reduction in quality-of-life parameters and premature discontinuation of therapy even when low doses are used. In
the French trial, patients received 200 mg/day of thalidomide, for a median of 15 months (range: 0.150 months). In
the other trials, the median duration of treatment varied
between 9 months and 2 years. The optimal dose of thalidomide should be the minimal effective dose that is associated
with superior tolerance and least toxicity; 50 mg/day for less
than 1 year appears to be the appropriate dose and duration.
Subsequently, lenalidomide became a logical and attractive alternative to thalidomide because of its lack of neurologic toxicity and a favorable prerequisite for long-term use.
Within the phase III reported trials, toxicity was acceptable
and only 5% of patients in the MM 015 trial,21 12% in the
CALBG trial,17 and 21% of patients in the IFM trial16
discontinued lenalidomide maintenance before myeloma
progression. Neutropenia was the most common side effect
noted, but febrile neutropenia was rare (2% 6%). Thromboembolic disease has been reported and long-term use of
prophylactic antiplatelets agent or low molecular weight
heparin should be considered.
519
ATTAL AND ROUSSEL

Table 3. Toxical Patterns of Maintenance Trials
Trial
Author and Year
Median Duration of
Treatment and Median
Received Dose
Discontinuation of
Treatment due to Drug-Related
Adverse Events (%)
Adverse Events* (%)

*All Grade Unless Specified
Thalidomide Maintenance
Post ASCT
IFM 9502
Attal et al 2006
Neuropathy (68%), fatigue (34%), constipation (20%), neutropenia

(7%)
NB: grade 3/4 neuropathy (7%), infection (6%)
Total therapy 2
grade2 neuropathy (15%), thrombosis (6%)
Barlogie et al 20062008
39%
(mainly related to neuropathy)
N/A
15 mo (0.150)
200 mg/d
80% stopped thalidomide
within 2 yr
ALLG MM6
Spencer et al 2009
Neuropathy (52%), infection 23%, fatigue (14%), constipation (18%)

NB: grade 3/4 neuropathy (10%)
30%
58% still on therapy at 12 mo

100 mg/d
HOVON-50
Lokhorst et al 2010
Neuropathy (64%)
33%
2 yr
NCIC CTG MY.10

Stewart et al 2010
Grade 3/4 neuropathy (10%), thrombosis (7%), fatigue (7%)
Post ASCT or Chemotherapy

MRC myeloma IX
N/A
Morgan et al 2012
Post Chemotherapy
GIMEMA
Palumbo et al 2008
16 mo
52%
7 mo (050)
50 mg/d
55% AE grade 3/4
N/A
8 mo (0.0339.40)
within the entire program
HOVON 49
Wijermans et al 2010
grade 2 neuropathy (54%)
N/A
8.4 mo (1.435.9)
NMSG
Waage et al 2010
Grade 34 non-hematological AE (40%, neuropathy (6%), thrombosis

(8%)
CEMSG
Ludwig et al 2010
Leucopenia (59%), fatigue (78%), neurological (69%), infection (28%),

constipation (44%), skin (33%)
Lenalidomide Maintenance
Post ASCT
IFM 200502
Attal et al 2011
CALBG 100104
McCarthy et al 2011
Post Chemotherapy
MM 015
Palumbo et al 2011
Post ASCT
HOVON 65/HD4
Sonneveld et al 2010
PETHEMA/GEM
Rosinol et al 2011
Post Chemotherapy
GIMEMA
Palumbo et al 2010
GEM2005MAS65
Mateos et al 2011
Neutropenia (68%), febrile neutropenia (2%), thrombocytopenia

(24%), diarrhea (40%), fatigue (47%), upper respiratory infection
(70%), neuropathy (23%), skin rash (20%), cramps (39%)
SPMs n 26
56% at 1 yr (mainly related to

neuropathy)
23%
7,5 mo
13 mo
75 mg/d
25% at time of unblinding
N/A
Grade 3/4 Neutropenia (43%), febrile neutropenia (6%),

thrombocytopenia (13%), diarrhea (4%), fatigue (5%), infection
(33%), skin rash (4%)
SPMs n 15
12%
N/A
Grade 4 hematologic AEs: thrombocytopenia (5%), neutropenia (4%),

anemia (3%); Grade 3/4 nonhematologic AEs: bone pain (5%),
diarrhea (5%) SPMs 8%
Bortezomib Maintenance
5%
N/A
Grade 2/4 neuropathy (23%), GI symptoms (23%), infections (53%)
9%
49% during 2 yr
Grade 1/3 neuropathy (12.2%)
15.6%
N/A
N/A
N/A
Grade 3/4 neutropenia (38%), thrombocytopenia (22%),

Infections (13%), neuropathy (8%), fatigue (6%), DVT (5%, Rash (4%)
VT: grade 3/4 neuropathy (9%), GI symptoms (4%), neutropenia (4%)
VT 13%
20 mo (136)
VP 9% (mainly related to neuropathy)
Abbreviation: N/A, not available.
An unexpected finding from the three lenalidomide maintenance trials was an increase in the incidence of secondary
cancers (SPMs), including hematological malignancies and
solid tumors; nearly 8% in both post-ASCT studies compared
with 2% in the placebo groups. The difference compared with
520
placebo was statistically significant (p 0.001), mainly after

24 months of treatment.
Intense investigation regarding the risk of SPMs in recipients of lenalidomide has ensued. In addition to a suspected
intrinsic risk of second cancers in this disease, antimyeloma
drugs other than lenalidomide may predispose to SPMs. In

the IFM 2005 02 trial, the cumulative incidence of SPMs
was 3.1 per 100 patient-years. Risk factors identified in
multivariate analysis were treatment with lenalidomide,
being over age 55, male gender, ISS III, and induction DCEP
(IFM 2005 01 protocol). In the MM 015 study, the cumulative incidence was 2.5 per 100 patient-years.
Palumbo A and colleagues28 recently presented the pooled
data of 1,798 patients from nine European protocols. The
risk/benefit of lenalidomide maintenance therapy was
largely in favor of the IMiDs, either regarding the risk of
progression or the risk of death as a result of myeloma.
Therefore, the strong efficacy of continuous therapy with
lenalidomide outweighs the potential risk of SPMs. Longer
follow-up is needed to definitively assess the risk of SPMs
in patients receiving lenalidomide, especially in context of
alkylating agents.
With respect to bortezomib maintenance, only sporadic
SPMs were reported with an incidence rate of 0.3 per 100
person-years.29 This advantage proved partially abrogated
by its specific toxicity profile, in particular the neurotoxicity.
After ASCT, in the HOVON joint trial,22 57% of patients
started maintenance, of which 27% required dose reductions
and 9% discontinued bortezomib because of toxicity, mainly
peripheral neuropathy. Less than half of the patients could
receive the planned 2 years of maintenance.
In elderly patients, similar limitations were observed. In
the UPFRONT study,24 50% of patients effectively received
the maintenance with 13% to 25% requiring dose reductions.
Surprisingly, maintenance with single-agent bortezomib was
well tolerated, with limited additional toxicity compared
with induction. In the Spanish trial GEM2005MAS65,25 up
to 10% of patients experienced grade 3/4 peripheral neuropathy and nearly 60% of patients discontinued bortezomib
because of adverse events. Of note, 5% of deaths were

related to drug toxicity. The median duration of bortezomib
maintenance was 20 months (range: 136 months).
Overall, the neurotoxicity related to bortezomib will likely
limits its use as a maintenance agent and dose adjustments
as well as a limited administration are therefore suggested.
Conclusion
Currently, maintenance therapy with novel agents is not

approved in most countries. Thalidomide and bortezomib
effectively increase response rates and prolong PFS, but that
benefit is hampered by their neurologic toxicity likely making them more useful as consolidation agents rather than
long-term maintenance.
Maintenance therapy with lenalidomide is clearly effective in prolonging the duration of response and diminishing
the risk of relapse in frontline multiple myeloma patients
irrespective of their age. However, patients treated with
lenalidomide have the ongoing risk of new cancer and a clear
OS benefit has yet to be shown. In the case of elderly
patients, where one hopes to maximize the response to
upfront treatment, the current robust data supports the use
of lenalidomide maintenance in these patients, this balancing the potential risk of SPMs.
In younger patients, whether lenalidomide maintenance
therapy should be routinely offered to patients is controversial. The increased incidence of SPMs is an important risk
and we await for more mature survival data before making
firm recommendations in this regard.
Acknowledgment
We are indebted to Christopher P. Venner, MD, for his critical
reading of the manuscript.
Author
Michel Attal
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Celgene;
Janssen-Cilag
Research
Funding
Expert
Testimony
Other
Remuneration
Celgene;
Janssen-Cilag
Murielle Roussel
Celgene;
Janssen
REFERENCES
1. Kumar SK, Rajkumar SV, Dispenzieri A, et al. Improved survival in
multiple myeloma and the impact of novel therapies. Blood. 2008;111:25162520.
2. Cavo M, Rajkumar SV, Palumbo A, et al. International Myeloma
Working Group consensus approach to the treatment of multiple myeloma
patients who are candidates for autologous stem cell transplantation. Blood.
2011;117:6063-6073.
3. Mihelic R, Kaufman JL, Lonial S. Maintenance therapy in multiple
myeloma. Leukemia. 2007;21:1150-1157.
4. Belch A, Shelley W, Bergsagel D, et al. A randomized trial of maintenance versus no maintenance melphalan and prednisone in responding
multiple myeloma patients. Br J Cancer. 1988;57:94-99.
5. Berenson JR, Crowley JJ, Grogan TM, et al. Maintenance therapy with
alternate-day prednisone improves survival in multiple myeloma patients.
Blood. 2002;99:3163-168.
6. Fritz E, Ludwig H. Interferon-alpha treatment in multiple myeloma:
meta-analysis of 30 randomised trials among 3948 patients. Ann Oncol.
2000;11:1427-1436.
7. Group MTC. Interferon as therapy for multiple myeloma: an individual
patient data overview of 24 randomized trials and 4012 patients. Br J
Haematol. 2001;113:1020-1034.
8. Barlogie B, Tricot G, Anaissie E, et al. Thalidomide and hematopoieticcell transplantation for multiple myeloma. N Engl J Med. 2006;354:10211030.
9. Attal M, Harousseau JL, Leyvraz S, et al. Maintenance therapy with
thalidomide improves survival in patients with multiple myeloma. Blood.
2006;108:3289-3294.
10. Spencer A, Prince HM, Roberts AW, et al. Consolidation therapy with
low-dose thalidomide and prednisolone prolongs the survival of multiple
myeloma patients undergoing a single autologous stem-cell transplantation
procedure. J Clin Oncol. 2009;27:1788-1793.
11. Lokhorst HM, van der Holt B, Zweegman S, et al. A randomized phase
3 study on the effect of thalidomide combined with adriamycin, dexamethasone, and high-dose melphalan, followed by thalidomide maintenance in
patients with multiple myeloma. Blood. 2010;115:1113-1120.
12. Stewart AK, Trudel S, Bahlis NJ, et al. A randomized phase III trial of
thalidomide and prednisone as maintenance therapy following autologous
stem cell transplantation (ASCT) in patients with multiple myeloma (MM):
The NCIC CTG MY. 10 Trial. ASH Annual Meeting Abstracts. 2010;116:39.
13. Morgan GJ, Gregory WM, Davies FE, et al. The role of maintenance
thalidomide therapy in multiple myeloma: MRC Myeloma IX results and
meta-analysis. Blood. 2012;119:7-15.
521
ATTAL AND ROUSSEL

14. Zangari M, van Rhee F, Anaissie E, et al. Eight-year median survival in
multiple myeloma after total therapy 2: roles of thalidomide and consolidation
chemotherapy in the context of total therapy 1. Br J Haematol. 2008;141:433444.
15. Hahn-Ast C, Lilienfeld-Toal M, Heteren P, et al. Improved progressionfree and overall survival with thalidomide maintenance therapy after autologous stem cell transplantation in multiple myeloma: a metaanalysis of five
randomized trials. Haematologica. 2011;96:1-678, 884.
16. Attal M, Olivier P, Cances Lauwers V, et al. Maintenance treatment
with lenalidomide after transplantation for myeloma: analysis of secondary
malignancies within the IFM 2005-02 trial and updated data. Haematologica.
2011;96:S23.
17. McCarthy PL, Owzar K, Anderson KC, et al. Phase III intergroup study
of lenalidomide versus placebo maintenance therapy following single autologous stem cell transplant (ASCT) for multiple myeloma (MM): CALGB ECOG
BMT-CTN 100104. Haematologica. 2011;96:S23.
18. Palumbo A, Bringhen S, Liberati AM, et al. Oral melphalan, prednisone, and thalidomide in elderly patients with multiple myeloma: updated
results of a randomized controlled trial. Blood. 2008;112:3107-114.
19. Wijermans P, Schaafsma M, Termorshuizen F, et al. Phase III study
of the value of thalidomide added to melphalan plus prednisone in elderly
patients with newly diagnosed multiple myeloma: the HOVON 49 Study.
J Clin Oncol. 2010;28:3160-3166.
20. Waage A, Gimsing P, Fayers P, et al. Melphalan and prednisone plus
thalidomide or placebo in elderly patients with multiple myeloma. Blood.
2010;116:1405-1412.
21. Palumbo A, Adam Z, Kropff M, et al. A phase 3 study evaluating the
efficacy and safety of lenalidomide (Len) combined with melphalan and
prednisone followed by continuous lenalidomide maintenance (MPR-R) in
patients (Pts) 65 years (yrs) with newly diagnosed multiple myeloma
(NDMM): updated results for pts aged 65-75 yrs enrolled in MM-015. ASH
Annual Meeting Abstracts. 2011;118:475.
22. Sonneveld P, Schmidt-Wolf I, van der Holt B, et al. HOVON-65/GMMGHD4 randomized phase III trial comparing bortezomib, doxorubicin, dexa-
522
methasone (PAD) Vs VAD followed by high-dose melphalan (HDM) and

maintenance with bortezomib or thalidomide in patients with newly diagnosed multiple myeloma (MM). ASH Annual Meeting Abstracts. 2010;116:40.
23. Rosinol L, Cibeira MT, Mateos MV, et al. A phase III PETHEMA/GEM
randomized trial of postransplant (ASCT) maintenance in multiple myeloma:
superiority of bortezomib/thalidomide compared with thalidomide and alfa-2b
interferon. ASH Annual Meeting Abstracts. 2011;118:3962.
24. Niesvizky R, Flinn IW, Rifkin R, et al. Efficacy and safety of three
bortezomib-based combinations in elderly, newly diagnosed multiple myeloma patients: results from all randomized patients in the community-based,
phase 3b UPFRONT study. ASH Annual Meeting Abstracts. 2011;118:478.
25. Mateos M-V, Oriol A, Teruel A-I, et al: Maintenance Therapy with
Bortezomib Plus Thalidomide (VT) or Bortezomib Plus Prednisone (VP) In
Elderly Myeloma Patients Included In the GEM2005MAS65 Spanish Randomized Trial. ASH Annual Meeting Abstracts. 2011;118:477.
26. Palumbo A, Bringhen S, Rossi D, et al. Bortezomib-melphalanprednisone-thalidomide followed by maintenance with bortezomibthalidomide compared with bortezomib-melphalan-prednisone for initial
treatment of multiple myeloma: a randomized controlled trial. J Clin Oncol.
2010;28:5101-5109.
27. Roussel M, Robillard N, Moreau P, et al: Bortezomib, lenalidomide, and
dexamethasone (VRD) consolidation and lenalidomide maintenance in frontline multiple myeloma patients: updated results of the IFM 2008 phase II
VRD intensive program. ASH Annual Meeting Abstracts. 2011;118:1872.
28. Palumbo A, Larocca A, Zweegman S, et al. Second primary malignancies in newly diagnosed multiple myeloma patients treated with lenalidomide:
analysis of pooled data in 2459 patients. ASH Annual Meeting Abstracts.
2011;118:996.
29. San Miguel JF, Richardson PG, Orlowski RZ, et al. Risk of second
primary malignancies (SPMs) following bortezomib (Btz)-based therapy:
analysis of four phase 3 randomized controlled trials in previously untreated
or relapsed multiple myeloma (MM). ASH Annual Meeting Abstracts. 2011;
118:2933.
NEW OPTIONS, NEW QUESTIONS: HOW TO

SELECT AND SEQUENCE THERAPIES FOR
METASTATIC MELANOMA
CHAIR
Michael B. Atkins, MD
Beth Israel Deaconess Medical Center
Boston, MA
SPEAKERS
Keith T. Flaherty, MD
Boston, MA
Jeff A. Sosman, MD
Vanderbilt Medical Center
Nashville, TN
New Options and New Questions: How to

Select and Sequence Therapies for Patients
with Metastatic Melanoma
By Keith T. Flaherty, MD, Jeff A. Sosman, MD, and Michael B. Atkins, MD
Overview: Recent advances in the understanding of the melanoma biology and tumor immunology have yielded new treatment strategies for patients with advanced melanoma. Within
the past year, the selective BRAF inhibitor vemurafenib and
immune checkpoint inhibitor ipilimumab have been added to
the treatment armamentarium. In addition, other molecularly
targeted agents and immunotherapies are showing considerable promise. The availability of multiple, effective treatment
ELANOMA IS currently the fifth and seventh most

common cancer in American men and women, respectively.1 Although early-stage patients can be treated
successfully with surgical resection, the prognosis for metastatic melanoma is dismal, with an overall 5-year mortality
rate of 90%. An estimated 8000 Americans will die of
melanoma in 2012. Many of these patients are young (median around 56 years) and otherwise healthy; therefore,
their loss represents a societal burden disproportionate to
other cancers. Historically, common treatment approaches
have included cytotoxic chemotherapy, interleukin-2-based
immunotherapy, and combinations of chemotherapy and
immunotherapyso called biochemotherapy. Although some
treatment approaches, most notably, high-dose interleukin
(IL)-2, have produced extremely durable tumor responses in
a small percentage of patients; unfortunately, no treatment
approach was able to reproducibly show a survival advantage in phase III trials.2 Hence, there has been an urgent
need for new treatment approaches for patients with advanced melanoma.
Molecular Biology of Melanoma
Elucidation of the somatic genetic alterations responsible

for the formation of melanoma has provided a matrix for
developing molecularly targeted therapies. Therapies aimed
at the mutated gene products themselves have been developed. In cases where oncogenes of interest cannot be feasibly
targeted directly or when pathways are activated by loss of
tumor suppressor gene function, critical nodes of signaling
are being sought for development of indirect targeting strategies.
Three well-defined driver oncogenes have been identified
in melanoma: BRAF, NRAS, and CKIT. Each is capable of
activating the MAP kinase pathway, whereas NRAS and
CKIT activate the PI3 kinase pathways among others.
BRAF is the most commonly activated oncogene in melanoma, with approximately 50% of advanced melanomas
harboring such mutations.3 Eighty to 90% of the BRAF
mutations found in melanoma result in a substitution of
glutamate for valine at the 600 position of the kinase domain
within the BRAF aminoacid sequence. The second most
common mutation results in a lysine substitution in the
same position. The two substitutions account for 95% of all
BRAF mutations in melanoma, and both create a kinase
that is constitutively active, phosphorylating MEKindependent of upstream activation by receptor tyrosine
kinases or RAS. Being a serine-threonine kinase that con-
524
options for patients with melanoma, although long sought, has

complicated treatment decisions. This article will review the
advances in our understanding of melanoma biology and
tumor immunology, the current status of immunotherapy, the
advances in molecularly targeted therapy for patients with
BRAF mutant melanomas, the possible approaches to patients
with BRAF wild-type (WT) tumors, and the current considerations for treatment selection of individual patients.
sumes ATP as an energy source, BRAF is amenable to

targeting with ATP competitive, small molecular inhibitors.4 The two most clinically advanced inhibitors, vemurafenib and dabrafenib, will be discussed in detail below.
NRAS, which activates RAF kinases, as well several other
downstream pathways, is activated by mutations that
disable the GTPase activity within the molecule.5 Such
mutations are found in 20% of advanced melanomas.6 Pharmacologic agents that can restore the function of the GTPase
have been technically difficult to generate. Thus, direct
targeting strategies for NRAS remain elusive. Indirect therapeutic strategies will be discussed in detail later in this
manuscript. Lastly, CKIT mutations, many of which overlap
with the ones found in gastrointestional tumors, are found
in approximately 1% of all melanomas. They are found
nearly exclusively in melanomas that arise on the nail beds,
palms, soles, and mucosal surfaces. In the advanced melanoma population, melanomas arising from these sites account for only 5% to 10% of all cases.7 Within these subsets,
CKIT mutations are found in approximately 10%. Mutations
in exons 11 and 13 are common, and many have previously
been described as responses to imatinib and other small
molecule KIT inhibitors. With roughly 70% of all melanomas
harboring BRAF, NRAS, or CKIT mutations, the remaining
tumors have yet to be associated with a described driver
oncogene, and thus lack a primary target for therapy.
Important secondary or complementary mutations have
been identified in two additional pathways, PI3 kinase and
p16/Rb. Both of these pathways have been implicated in
melanoma formation, and preliminary evidence suggests
that they serve as important modifiers of resistance to
single-agent targeted therapies aimed at driver oncogenes,
although direct targeting of these pathways in preclinical
models is not associated with significant antitumor effects.
Activating mutations of PI3 kinase itself have not been
described in melanoma, and activating AKT mutations have
been observed only rarely.8 Amplification of one AKT isoform, AKT3, has been identified in a subset of melanoma cell
From the Massachusetts General Hospital Cancer Center, Boston, MA; Vanderbilt
University Medical Center, Nashville, TN; Beth Israel Deaconess Medical Center, Boston,
MA.
Address reprint requests to Keith T. Flaherty, MD, Massachusetts General Hospital
Cancer Center, 55 Fruit St., Yawkey 9E, Boston MA 02114; email: kflaherty@partners.org.
1092-9118/10/1-10
THERAPIES FOR PATIENTS WITH METASTATIC MELANOMA
lines, but has not yet been validated in a large population of

human tumor samples.9 PTEN deletion and inactivating
mutation are the common causes of PI3 kinase pathway
activation, occurring in approximately 20% of all melanomas.6 These aberrations commonly coexist with BRAF mutation, and on occasion with NRAS mutation, and are
critical cofactors in melanoma formation in experimental
models.10,11 Mutations of p16 are well described in familial
melanoma, and mutations or deletion of p16 are commonly
found in sporadic melanoma as well (approximately 40% of
advanced tumors).6 Loss of p16 function co-occurs with
BRAF mutation as well as BRAF WT tumors in many cases.
In cooperation with BRAF mutation, p16 loss can facilitate
formation of invasive melanomas in mouse models.12 CDK4,
a key regulator of the cell cycle, is negatively regulated by
p16. Thus, loss of p16 function permits CDK4 to drive
progression through the cell cycle, independent of its usual
upstream regulation. CDK4 itself is amplified or can harbor
activating mutations in some melanomas (approximately
5%) and cyclin D, the binding partner of CDK4, is amplified
in a distinct subset (20%), resulting in the same cell cycle
dysregulation as p16 loss.6 To intercept the consequences of
p16 loss, cyclin D amplification or CDK4 mutation, or
selective CDK4 inhibitor are thought be relevant.
Based on currently available evidence, targeting BRAF,
CKIT, and the pathways downstream of NRAS is a promising starting point for melanoma-targeted therapies. At least
for BRAF, cotargeting of constituents of the PI3 kinase
pathway or CDK4 appears to be a plausible strategy for
overcoming resistance to BRAF inhibitors and antagonizing
multiple oncogenic pathways to achieve greater degrees of
tumor control.
molecules, such as CTLA4 on the surface of activated T cells.

Mechanisms identified for tumor-induced immune suppression have included stimulation of CD4, CD25 Tregulatory cell (Treg) production, which serves to limit immune
activation, inhibition of T-cell receptor signaling, and melanoma cell expression of PDL1, blocking the cytolytic function
of tumor-infiltrating T lymphocytes (Lisee; Greenwald, Freeman). These biologic discoveries have been made clinically
relevant through the development of therapeutic agents that
directly or indirectly target these immunomodulatory pathways, potentially enabling the restoration of effective,
tumor-specific immune destruction. Ipilimumab, an antibody directed against CTLA4, has shown a significant survival benefit in two phase III studies of patients with
advanced melanoma, leading to its U.S. Food and Drug
Administration (FDA) approval in 2011. In addition, antibodies that block the effects of tumor PDL1 expression
(anti-PD1 antibodies) have shown encouraging response
rates and toxicity profiles in phase I trials, prompting
extensive additional investigation. Other efforts have attempted to selectively deplete Treg cells through lymphodepletion with nonmyeloablative chemotherapy coupled
with tumor specific adoptive T-cell immunotherapy. Another
strategy has focused on identifying the patients most likely
to respond to immunotherapy, based on melanoma genomic
profile or gene-expression pattern; immune-cell infiltration;
or, checkpoint inhibitor or plasma cytokine expression.
Treatment is restricted to those most likely to benefit or
focuses on identifying ways of manipulating the immune
system to convert tumors to a more immune-responsive
phenotype.
Immunotherapy for Melanoma
Melanoma Tumor Immunology
Studies of immune regulation and the mechanisms of

tumor-induced immune suppression have identified specific
obstacles to effective immunotherapy of melanoma. These
include the physiologic down-modulation of the immune
response through the upregulation of the expression of
KEY POINTS
Delineation of genetic alterations in melanoma has

provided the opportunity to develop targeted therapy
in the treatment of refractory disease.
For the approximately 50% of melanoma patients
whose tumors harbor mutated BRAF, BRAF inhibitors produce unprecedented response rates and a
survival advantage compared with chemotherapy.
Advances in the understanding of immune-system
function and tolerance have led to the development of
a new generation of targeted immunotherapies.
Ipilimumab represents the first immunotherapy to
improve survival in phase III trials, with durable
responses observed in a subset of patients.
Selection of optimal therapy for individual patients
requires consideration of patients and disease characteristics to devise best first-line and sequential
therapy approaches.
IL 2Based Therapy
High-dose bolus IL2 (HD IL-2) received FDA approval in

1998 for the treatment of patients with metastatic melanoma, largely based on its ability to produce durable complete responses in 5% to 10% of patients. In a retrospective
review of 270 patients treated on multiple phase II studies,
the objective response rate was 16%, with a median duration
of 9 months (range from 4 months to over 106 months).
Despite the low objective response rate, 59% of complete
responders remained progression-free at 7 years, and no
patient responding for longer than 30 months had progressed, suggesting that some patients were cured.2 Treatment, however, was associated with significant toxicity
limiting its application to a select group of patients treated
in specialized centers.
Efforts to improve on the efficacy of IL-2 in patients with
melanoma have included combinations with chemotherapy,
i.e., biochemotherapy, vaccines, and adoptive-cell therapy.
Although several phase II trials, a small phase III trial, and
two meta-analyses suggested that combinations of IL-2
and cisplatin-based biochemotherapy offered benefit relative
to either chemotherapy or IL-2 alone, several multiinstitutional phase III trials have failed to confirm this
benefit.
Another approach to improving the activity of HD IL-2
involved the addition of a gp100 peptide vaccine. A recently
reported phase III trial randomly assigned 185 patients with
metastatic melanoma to HD IL-2 given alone every three
weeks or in combination with a gp100 peptide vaccine.13
525
FLAHERTY, SOSMAN, AND ATKINS
Because of the specificity of the peptide vaccine for certain

histocompatibility receptor subtypes, enrollment was limited to HLA-A*0201 patients. The study reported an objective response rate of 16% for the combination compared with
6% for HD IL-2 alone. There were eight complete responses
(9%) in the combination arm, but only one (1%) among those
treated with IL-2 alone. There was a trend toward increased
overall survival (median 17.8 months vs. 11.1 months; p
0.06), although the trial was not adequately powered to
assess this endpoint. The clinical significance of this finding
is uncertain, considering the relatively poor response rate
in patients treated with HD IL-2 alone, the current lack of
availability of the specific formulation of vaccine adjuvant
used in this trial, and the observations that this same
vaccine did not improve the efficacy of ipilimumab in a phase
III trial (see below).
Others have explored the efficacy of HD IL-2 in combination with adoptive transfer of tumor derivedtumor reactive
T cells. These approaches have included preparative regimens involving myeloablative chemotherapy with or without total-body irradiation in order to delete host immune
cells and promote engraftment of adoptively transferred
tumor-reactive T cells.14 Autologous hematopoietic progenitorcell support was used in patients who received TBI. The
Surgery Branch of the National Cancer Institute recently
reported the combined results from three separate trials.
There were 52 objective responses in 93 patients (56%
response rate), including 20 (22%) complete responses. Complete responses were ongoing at 37 months to 82 months in
19 of the 20 responders, and the 3- and 5-year actuarial
survival rates for patients achieving a complete response
were 100% and 93%, respectively. Efforts to confirm these
results at other centers, including in a proposed multicenter
trial, as well as to develop a more practical treatment
regimen are currently underway.
Ipilimumab
The CTLA-4 receptor on T lymphocytes is a negative

regulator of T-cell activation that blocks positive stimulatory
effects to these cells, mediated through their costimulatory
and antigen specificT-cell receptors. The monoclonal antibodies ipilimumab and tremelimumab bind to CTLA-4 and
thus prevent this feedback inhibition. Both have been studied in patients with melanoma, with the most extensive data
and promising results being observed with ipilimumab.
Ipilimumab was studied in a placebo-controlled phase III
trial in which 676 patients with previously treated advanced
melanoma were randomly assigned in a 3:1:1 ratio to ipilimumab plus gp100 peptide vaccine, ipilimumab alone or
gp100 vaccine alone.15 Ipilimumab (3 mg/kg) and/or vaccine
were given every three weeks for four doses. Patients with
confirmed partial or complete response or stable disease for
three months or more after completion of the 12-week
induction period were allowed to receive reinduction with
their original treatment if they subsequently had disease
progression.
In this study, overall survival was significantly increased
in the two groups who received ipilimumab (median 10.0
months and 10.1 months vs. 6.4 months, in the ipilimumab
plus gp100, ipilimumab alone, and gp100 groups, hazard
ratios for death 0.68 and 0.66 compared with gp100 alone,
respectively). Treatment benefits appeared to be independent of sex, age ( age 65 or age 65), stage at presentation
526
(M0, M1a, and M1b compared with M1c), baseline lactate

dehydrogenase (LDH), or prior use of IL-2. Tumor response
rate was also significantly improved in both groups of
patients treated with ipilimumab compared with gp100
alone (5.7% and 10.9% vs. 1.5%, respectively). Further,
objective partial or complete responses were maintained for
at least 2 years in four of 23 (17%) patients treated with
ipilimumab plus gp100 and nine of 15 (60%) patients with
ipilimumab alone. Among 31 patients who initially received
ipilimumab either alone or with gp100 and then underwent
reinduction therapy with ipilimumab, six (21%) had an
objective response to retreatment, and 15 (48%) had stable
disease. Although this phase III trial limited enrollment to
patients who were HLA-A*0201 positive, a retrospective
analysis of four phase II trials involving ipilimumab alone
showed similar activity regardless of HLA type. Although
patients in this trial did not have tumor profiling for BRAF
mutations, limited unpublished data from Bristol-Myers
Squibb (BMS) suggests that the activity of ipilimumab is
independent of BRAF mutational status. As a consequence
of this study, ipilimumab was approved for the treatment of
all patients with advanced melanoma.
The presumed mechanism of action of ipilimumab is to
break down tolerance to tumor-associated antigens in the
melanoma. At the same time, this break down of tolerance
may result in autoimmune reactions against self-antigens. A
wide range of immune-mediated adverse events have been
observed. The most common, serious manifestations include
enterocolitis, hepatitis, dermatitis, and endocrinopathies. In
this trial using a 3 mg/kg dose of ipilimumab, immunerelated adverse events occurred in approximately 60% of
patients treated with ipilimumab. Grade 3 or 4 toxicity was
seen in 10% to 15% of ipilimumab-treated patients, compared to 3% of those receiving only gp100. These side effects
were typically not seen until 6 or more weeks into therapy.
A somewhat higher incidence of side effects was observed
with a dose of 10 mg/kg every 3 weeks in a randomized phase
II trial that assessed the effects of dose on activity and
toxicity.16 Several investigators have suggested that the
development of immune-related toxicities correlated with
benefit from therapy.
Although patients with untreated brain metastases were
excluded from the phase III trial, other studies have observed antitumor activity with ipilimumab for patients with
brain metastases.17 Finally, data from phase II trials suggested that a number of patients (up to 10% of those treated)
exhibited apparent disease progression after 12 weeks of
ipilimumab (with either larger lesions or new lesions),
followed by subsequent disease regression. The overall survival outcome of these patients was similar to those exhibiting a tumor response. This led to the establishment of
Immune-related Response Criteria that endeavored to capture these patients in the subset of patients achieving
treatment benefit.18
A second phase III trial involved previously untreated
patients who were randomly assigned to dacarbazine plus
either ipilimumab or placebo.19 In this study, overall survival was significantly increased in patients assigned to the
dacarbazine plus ipilimumab arm (median 11.2 months vs.
9.1 months). The overall incidence of grade 3 or 4 toxicity
was significantly higher with dacarbazine plus ipilimumab
(56% vs. 28%). In particular, hepatic toxicity was significantly more common with the combination than with dacar-
bazine alone or than that observed with ipilimumab alone.

The increase in hepatic toxicity may be due to its combination with dacarbazine, which is also known to be hepatotoxic. On other hand, the incidence of other immune-related
toxicities (colitis, rash, hypophysitis) was less than that seen
in prior studies with ipilimumab alone, perhaps suggesting
that dacarbazine may have blunted the immune-toxicity
profile,and/or the higher incidence of hepatotoxicity may
have pre-empted or altered the immune toxicity profile.
Whether this blunting of immune toxicity by dacarbazine
might have also blunted the antitumor effect of ipilimumab
is a matter of speculation. However, the overall pattern of
toxicity and efficacy of this trial do not support the addition
of dacarbazine to ipilimumab. The relative value of the use
of ipilimumab at the 10 mg/kg dose used in this study and in
multiple phase II studies versus the already approved 3
mg/kg dose awaits the completion of an ongoing phase III
trial directly comparing the two doses.
Other Immune Regulatory Checkpoints
Monoclonal antibodies targeted against a number of other

regulatory checkpoints are being evaluated in patients with
advanced melanoma based on our current understanding of
the development of cellular immunity. The PD-1 receptor
acts as an inhibitory receptor of T cells, in a manner
analogous to CTLA-4S.20 In a preliminary report of a phase
I study, a monoclonal antibody directed against the PD-1
receptor (MDX-1106) caused significant regression of metastatic melanoma lesions in three cases and appears to be
associated with less autoimmunity than reported with ipilimumab.21 Ongoing studies are exploring various doses of
this antibody as well as a variety of different antibodies
targeting either PD1 or PDL1 patients with melanoma. A
member of the tumor necrosis factor (TNF) family, 4 1BB
(CD137), acts as a costimulatory molecule that causes T-cell
proliferation. A humanized Monoclonal Ab (MAb), BMS663513, targeted at CD137 acts as an agonist and can cause
costimulation of CD8 and CD4 cells. In a preliminary
report of the initial phase I study with this agent, three
partial responses were observed among 54 patients with
melanoma.22 OX-40 is another TNF receptor, which also
acts as a costimulatory factor for T cells. A MAb that targets
this receptor has begun a phase I study.23
Treatment Selection Options
Considerable effort has focused on identifying patients

who respond to immunotherapy in the hope of restricting
such treatment to those most likely to benefit. IL-2 responsiveness has been shown to be more likely in patients with
normal serum LDH, or low plasma vascular endothelial
growth (VEGF) and fibronectin levels.24 In addition, response appears to be more frequent in patients whose
tumors contain mutations in BRAF or NRAS, or possess an
inflammatory gene-expression signature.25 More recent
studies have suggested that response to IL-2 is associated
with enhancement of a pre-existing gene expression pattern
within the tumor associated with immune-mediated tissuespecific destruction under the control of IFNgamma.26
Benefit from vaccination has also been linked to tumors
expressing an IFN driven chemokine signature.27 Preliminary results suggest that both PD1 antibody responsiveness
and IL-2 responsive in patients with renal cell carcinoma
(RCC) may be correlated with tumor cellsurface expression
of PDL1. Furthermore, research suggests that tumor PDL1

expression is not constitutive, but is related to the secretion
of IFNg by of tumor reactive CD8 T cells in the microenvironment. Thus, effective immunotherapy may require
pre-existence of tumor-specific immunity within the microenvironment and the use of agents that can either drive
T-cell function (HD IL-2 or vaccines) or block inherent
immunoregulatory signals (ipilimumab, or anti-PD1). Several current studies are underway to validate these predictive biomarkers for specific immunotherapies as well as
to determine if combinations of immunotherapy with either
other immunotherapies or molecularly targeted agents
could convert nonimmune responsive tumors into those
capable of responding.
BRAF Inhibitor Therapy for BRAFV600
Mutant Melanoma
As previously described, approximately 50% of patients

with cutaneous melanoma, have tumors that express the
mutated BRAF oncogene. Those patients with melanoma
carrying a V600E/K mutation are responsive to selective
BRAF inhibitors of which vemurafenib is the first to obtain
FDA-approval.28,29 Other BRAF inhibitors are in the late
stages of development and look equally promising.30
Through a series of clinical trials performed in rapid succession from phase I to phase II and III, the BRAF inhibitor,
vemurafenib, was shown to effectively inhibit the MAPK
pathway constitutively activated by the BRAF mutation;
affect clinical tumor regression in most patients with objective clinical responses in over 50% of patients; and, finally,
improve overall survival compared to dacarbazine chemotherapy.28,29 The phase I trial established that vemurafenib
inhibited its target and demonstrated clinical efficacy in
most of the 32 patients with a BRAFV600E/K mutation.28 A
much larger phase II trial enrolled patients with 132
BRAFV600E/K mutated-melanoma who had failed either
standard immunotherapy or chemotherapy and documented
an objective response rate of 53%, of which 6% were complete responses.31 The median progression-free survival for
the entire population was 6.8 months. The median overall
survival was a remarkable 15.9 months (95% CI 11.8 18.3).
Finally, a randomized phase III trial allocated 675 patients
to open label vemurafenib at 960 mg twice daily orally or
dacarbazine at 1,000 mg/m2 every 3 weeks with no crossover allowed. At the first interim analysis, the HR for death
was 0.37 (95% CI 0.26 0.55; p 0.0001). But the median
follow-up was short, only 3.75 months. Vemurafenib was
approved in August of 2011.
Although vemurafenib is generally well tolerated, it can
cause frequent skin toxicities dominated by hyperproliferative skin lesions, including keratoacanthomas, a low-grade
cutaneous squamous cell carcinoma.32 Approximately 25%
of patients develop such lesions in a median time of only 8
weeks, but they can be easily treated by excision without
interrupting treatment. However, there may be a more
significant issue in the adjuvant setting that could be
representative of the agents potential to enhance the
growth of other subclinical malignancies.
With a median progression-free survival of 6 to 8 months,
many responses to BRAF inhibitor therapy are short-lived.
A minority are maintained for more than 12 months, and
these patients are typically those with baseline normal LDH
and nonbulky disease. Resistance to these BRAF inhibitors
527
has been intensively studied. These studies have proposed

and validated a number of resistance mechanisms in small
numbers of patient tumor samples including: (1) Mutations
of NRAS, the protein upstream of BRAF in the MAPK
pathway.33 (2) Activation of downstream signaling proteins
including MEK1 through mutation or activation by overexpression of MAP kinases34,35 (3) The expression of alternate
shortened forms of the BRAFV600 mutated molecule that
are capable of binding together and reactivating the pathway even in the presence of vemurafenib.36 All three of these
mechanisms appear to reactivate the MAPK pathway, providing a direction to pursue in order to effectively delay or
overcome these forms of resistance. (4) Alternatively, activation or overexpression of alternate pathways through
activation of receptor tyrosine kinases, such as IGFR1 or
PDGFRb. Both of these RTK appear to signal at least in part
through the PI3K/Akt/mTOR pathway.33,37
Ongoing trials are combining BRAF and MEK inhibitors
with the goal of delaying or overcoming resistance as a
result of reactivation of the MAPK pathway. On the other
hand, combinations of BRAF-inhibitor therapy with either
PI3 kinase, PI3 kinase/mTOR, Akt, mTORC1- and mTORC2dual inhibitors could be effective inhibiting mechanisms of
resistance involving alternate pathways and different receptor tyrosine kinases.
Although BRAF inhibitors induce rapid disease regression
and symptom benefit in patients with bulky disease, markedly delay progression, and improve overall survival for the
overall metastatic melanoma population, there is clearly
need for further improvement of these outcomes.
For patients with BRAFV600E/K mutant melanoma,
there is a choice of therapy between new and older immunotherapy approaches, including high dose IL-2, ipilimumab, or a clinical trial investigating anti-PD1 compared
with a BRAF inhibitor. It is apparent that patients with
symptomatic, bulky, rapidly growing, and high-serum LDH
associated melanoma are much more likely to have rapid
clinical improvement when treated with vemurafenib. The
improvement may be relatively short term but can last up to
12 months in some patients. These patients are unlikely to
benefit from immunotherapy, in part, because immunebased treatments can take a period of time and some
experience progression initially before later improvement.
Furthermore, the percent of patients having an objective
clinical response is much lower with IL-2 (15% to 20%) or
ipilimumab (5% to 10%) than with vemurafenib ( 50%).
The greater dilemma surrounds the patient with BRAFmutant melanoma who has nonbulky, asymptomatic disease, and with a normal LDH. These patients have a choice.
If young enough, with excellent organ function and at a
treatment center with extensive experience, IL-2 may be the
first choice. Or, in others who do not fulfill IL-2 criteria,
ipilimumab. However, because of the delay in beneficial
effect, it may require waiting up to 4 to 6 months before one
can be certain that the melanoma is not responsive. Also, in
this subset of patients, the response rate to vemurafenib is
above 60% in the phase II trial, and overall survival is quite
long with a median of 15.9 months.31 This compares to 10
months on the previously treated trial and 11.2 months on
the treatment-naive trial with ipilimumab. The definite
answer to the question of which drug to use first will only be
answered through trials including combinations of the two
agents.
528
Treatment Options for BRAF WT Melanoma
BRAF WT melanoma is a heterogeneous disease that

includes patients with NRAS mutated melanoma (about
one-third of these BRAF WT patients) as well as very small
subset of patients with CKIT mutant melanoma from mucosal, acral, and chronic sun-damaged sites. Characterizing
these mutations can be justified currently for the purposes
of directing patients to relevant clinical trials. There is
supporting literature showing that a fraction of patients
with L597 or K642 mutations in CKIT are responsive to
imatinib, some of which the results are quite durable ( 12
months).38,39 NRAS approaches will be discussed below.
Immunotherapy remains a strong consideration in BRAF
WT patients. Those who are eligible for IL-2 should consider
this treatment at an experienced center. Since it has the
longest follow-up, we are confident of both its response rate
and the frequency of extremely durable disease responses.
Those patients are frequently assessed at 7 to 8 weeks and
again at 11 to 12 weeks and in the case of progression or
even stable disease, this treatment is rarely continued.
Toxicities are acute and rarely, if ever, delayed. For these
reasons, we believe that consideration of ipilimumab should
come after these patients have received IL-2 if the treatment
is appropriate. As stated previously, new trials with antiPD1 antibodies are ongoing and demonstrate exciting, early
results with response rates over 20% of which most are
durable lasting over 12 months.
It is critical to continue efforts to define other driver
kinases within the BRAF WT population that could be
targeted with drug therapy. For now, efforts are underway
to investigate approaches to NRAS mutant melanoma,
These include MEK inhibitors (GSK 1120212, TAK733, or
others) alone or in combination with inhibitors of the PI3K/
AKT, mTOR pathway. These trials are still in dose-finding
phases, but soon phase II efforts will be undertaken in
NRAS mutant melanoma.
Finally, chemotherapy still represents an option for patients without either good standard options, as listed above,
or clinical phase I and II trial options. Because of its limited
long-term benefit, chemotherapy is best considered for patients who are symptomatic or with rapidly growing disease
with a goal of improving quality of life. Agents, including
dacarbazine, temozolomide, and taxane-based regimens,
have all demonstrated similar antitumor activity with below
20% response rates. Rarely, long-lasting responses are observed.
Conclusion
Given the current availability of multiple treatment options, patients and physicians now often have choices regarding initial treatment and the sequence of various
treatments. For patients with BFAFV600E mutant melanoma, options could include HD IL-2, ipilimumab or vemurafenib. At the moment, there appears to be only limited
information to guide this choice. Although vemurafenib
appears to be equally active in patients whose disease has
progressed following IL 2-based immunotherapy, there is no
data on its activity following resistance to ipilimumab.
Similarly, there is no data on the activity of ipilimumab (or
even the feasibility of stopping vemurafenib and administering ipilimumab) following disease progression on vemurafenib. It is conceivable that patients treated initially with
ipilimumab may get the benefit of a durable response (up to

25%) without compromising the benefit of subsequent vemurafenib for those patients who exhibit disease progression,
whereas those started on vemurafenib may not be able to
benefit from subsequent ipilimumab at time of progression.
It is also possible that the high response rate and prolonged
progression free survival associated with vemurafenib therapy relative to ipilimumab may overwhelm any benefit that
might be achieved with initial ipilimumab therapy. Further,
it is possible that subsets of patients, defined by tumor
and/or clinical characteristics (for example, PTEN loss, immune infiltration, high serum LDH), might do better with
one initial therapy or sequence than with the other. A
randomized phase III trial of ipilimumab (followed by vemurafenib) versus vemurafenib (followed by ipilimumab) is
being planned within the Cooperative Group mechanism in
an effort to obtain real data to address these important
therapeutic questions and guide future treatment choices for
this patient population. In addition, this population repre-
sents an ideal group for combination treatment approaches.

Studies involving selective BRAF inhibitors with immunotherapy (ipilimumab, HD IL-2, PD1 pathway blockers),
VEGF pathway inhibitors, or other molecularly targeted
agents (MEK or PI3Kinase inhibitors or apoptosis inducers)
are currently underway and represent opportunities to further advance treatment results.
Options are more limited for patients with BRAF WT
melanoma. For the few patients with mutations in C-Kit,
imatinib or participation in a C-kit-inhibitor clinical trial
seems to represent the best choice. For those with NRAS
mutations, HD IL-2 might be a reasonable option, although
data supporting this selection requires validation. Additional research is needed to determine the activity of ipilimumab or other immunotherapies in this population.
Efforts are also underway to identify potential therapeutic
targets downstream of NRAS or for BRAF, NRAS, or CKIT
triple WT tumors that may allow for development of personalized treatment options for this patient population.
Author
Keith T. Flaherty
Jeff A. Sosman
Michael B. Atkins
Employment or
Leadership
Positions
Consultant or
Advisory Role
Genentech;
GlaxoSmithKline;
Metamark
Genetics
GlaxoSmithKline;
Roche
Bristol-Myers
Squibb; Celgene;
Curetech;
Genentech;
Merck; Novartis;
Prometheus
Stock
Ownership
Honoraria
Genentech;
GlaxoSmithKline
Research
Funding
Expert
Testimony
Other
Remuneration
Roche
REFERENCES
1. Siegel R, Ward E, Brawley O, et al. Cancer Statistics, 2011: the impact
deaths. CA Cancer J Clin. 2011;61:212-236.
2. Atkins MB, Lotze MT, Dutcher JP, et al. High-dose recombinant interleukin 2 therapy for patients with metastatic melanoma: analysis of 270
patients treated between 1985 and 1993. J Clin Oncol. 1999;17:2105-2116.
3. Davies H, Bignell GR, Cox C, et al. Mutations of the BRAF gene in
human cancer. Nature. 2002;417:949-954.
4. Tsai J, Lee JT, Wang W, et al. Discovery of a selective inhibitor of
oncogenic B-Raf kinase with potent antimelanoma activity. Proc Natl Acad
Sci U S A. 2008;105:3041-3046.
5. Neal SE, Eccleston JF, Webb MR. Hydrolysis of GTP by p21NRAS, the
NRAS protooncogene product, is accompanied by a conformational change in
the wild-type protein: use of a single fluorescent probe at the catalytic site.
Proc Natl Acad Sci U S A. 1990;87:3562-3565.
6. Curtin JA, Fridlyand J, Kageshita T, et al. Distinct sets of genetic
alterations in melanoma. N Engl J Med. 2005;353:2135-2147.
7. Curtin JA, Busam K, Pinkel D, et al. Somatic activation of KIT in
distinct subtypes of melanoma. J Clin Oncol. 2006;24:4340-4346.
8. Davies MA, Stemke-Hale K, Tellez C, et al. A novel AKT3 mutation in
melanoma tumours and cell lines. Br J Cancer. 2008;99:1265-1268.
9. Stahl JM, Sharma A, Cheung M, et al. Deregulated Akt3 activity
promotes development of malignant melanoma. Cancer Res. 2004;64:70027010.
10. Tsao H, Goel V, Wu H, et al. Genetic interaction between NRAS and
BRAF mutations and PTEN/MMAC1 inactivation in melanoma. J Invest
Dermatol. 2004;122:337-341.
11. Dankort D, Curley DP, Cartlidge RA, et al. Braf(V600E) cooperates
with Pten loss to induce metastatic melanoma. Nat Genet. 2009;41:544-552.
12. Dhomen N, Reis-Filho JS, da Rocha Dias S, et al. Oncogenic Braf
induces melanocyte senescence and melanoma in mice. Cancer Cell. 2009;15:

294-303.
13. Schwartzentruber DJ, Lawson DH, Richards JM, et al. Gp100 peptide
vaccine and interleukin-2 in patients with advanced melanoma. N Engl
J Med. 2011;364:2119-2127.
14. Rosenberg SA, Yang JC, Sherry RM, et al. Durable complete responses
in heavily pretreated patients with metastatic melanoma using T-cell transfer immunotherapy. Clin Cancer Res. 2011;17:4550-4557.
ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010;363:
711-723.
16. Wolchok JD, Neyns B, Linette G, et al. Ipilimumab monotherapy in
patients with pretreated advanced melanoma: A randomised, double-blind,
multicentre, phase 2, dose-ranging study. Lancet Oncol. 2010;11:155-64.
17. Lawrence DPH, McDermott DF. Phase II trial of ipilimumab monotherapy in melanoma patients with brain metastasis. J Clin Oncol. 2010;28
18. Wolchok JD, Hoos A, ODay S, et al. Guidelines for the evaluation of
immune therapy activity in solid tumors: immune-related response criteria.
Clin Cancer Res. 2009;15:7412-7420.
for previously untreated metastatic melanoma. N Engl J Med. 364:2517-2526.
20. Sharpe AHW, EJ, Ahmed R, Freeman GJ. The function of programmed
cell death 1 and its ligands in regulating autoimmunity and infection. Nat
Immun. 2007;8:239-245.
21. Brahmer JRD, Wollner I, Powderly JD, et al. Phase I study of singleagent anti-programmed death-1 (MDX-1106) in refractory solid tumors:
safety, clinical activity, pharmacodynamics, and immunologic correlates.
J Clin Oncol. 2010;28:3167-3175.
22. Sznol MH, Margolin K, McDermott D, et al. Phase I study of BMS-
529

663513, a fully human anti-CD137 agonist monoclonal antibody in patients
with advanced cancer. J Clin Oncol. 2008;26;15s (suppl; abstr 3007).
23. Curti BW, Morris N. A phase I trial of monoclonal antibody to OX40 in
patients with advanced cancer (abstract). International Society for Biological
Therapy of Cancer Annual Meeting. 2007.
24. Sabatino MK, Panelli MC. Serum vascular endothelial growth factor
and fibronectin predict clinical response to high-dose interleukin-2 therapy.
J Clin Oncol. 2009;27:2645-2652.
25. Joseph RW, Sullivan RJ, Harrell R, et al. Correlation of NRAS mutations with clinical response to high-dose IL-2 in patients with advanced
melanoma. J Immunother. 2012;35:66-72.
26. Sullivan RJ, Hoshida Y, Brunet J, et al. A single center experience with
high-dose IL-2 treatment for patients with advanced melanoma and pilot
investigation of a novel gene expression signature as a predictor of response.
27. Gajewski TF, Louahed J, Brichard VG. Gene signature in melanoma
associated with clinical activity: A potential clue to unlock cancer immunotherapy. Cancer J. 2010;16:399-403.
BRAF in metastatic melanoma. N Engl J Med. 2010;363:809-819.
2011;364:2507-2516.
30. Kefford RA, Brown MP, Millward M, et al. Phase I/II study of
GSK2118436, a selective inhibitor of oncogenic mutant BRAF kinase, in
patients with metastatic melanoma and other solid tumors. J Clin Oncol.
2010;28;15s (suppl; abstr 8503).
530
31. Ribas AK, Schuchter L, Gonzalez R, et al. BRIM 2: An open-label,

multicenter phase II study of vemurafenib (PLX4032, RG7204) in previously
treated patients with BRAFV600E mutation-positive metastatic melanoma.
J Clin Oncol. 2011;29:15s (abstr; suppl 8509).
32. Su F, Viros A, Milagre C, et al. RAS mutations in cutaneous squamouscell carcinomas in patients treated with BRAF inhibitors. N Engl J Med.
2012;366:207-215.
33. Nazarian R, Shi H, Wang Q, et al. Melanomas acquire resistance to
B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature. 2010;468:
973-977.
34. Wagle N, Emery C, Berger MF, et al. Dissecting therapeutic resistance
to RAF inhibition in melanoma by tumor genomic profiling. J Clin Oncol.
2011;29:3085-3096.
35. Johannessen CM, Boehm JS, Kim SY, et al. COT drives resistance to
RAF inhibition through MAP kinase pathway reactivation. Nature. 2010;468:
968-972.
36. Poulikakos PI, Persaud Y, Janakiraman M, et al. RAF inhibitor
resistance is mediated by dimerization of aberrantly spliced BRAF(V600E).
Nature. 2011;480:387-390.
37. Villanueva J, Vultur A, Lee JT, et al. Acquired resistance to BRAF
inhibitors mediated by a RAF kinase switch in melanoma can be overcome by
cotargeting MEK and IGF-1R/PI3K. Cancer Cell. 2010;18:683-695.
38. Guo J, Si L, Kong Y, et al. Phase II, open-label, single-arm trial of
imatinib mesylate in patients with metastatic melanoma harboring c-Kit
mutation or amplification. J Clin Oncol. 2011;29:2904-2909.
39. Carvajal RD, Antonescu CR, Wolchok JD, et al. KIT as a therapeutic
target in metastatic melanoma. JAMA. 2011;305:2327-2334.
ANTIEMETICS: CURRENT STANDARDS, EMERGING

APPROACHES, AND PERSISTENT GAPS
CHAIR
Ethan Basch, MD, MSc
New York, NY
SPEAKERS
Steven M. Grunberg, MD
Fletcher Allen Health Care
Burlington, VT
University of Kansas Cancer Center
Kansas City, KS
Matti S. Aapro, MD
Clinique de Genolier
Genolier, Switzerland
Antiemetic Use in Oncology: Updated

Guideline Recommendations from ASCO
By Ethan Basch, MD, Ann Alexis Prestrud, MPH, Paul J. Hesketh, MD, Mark G. Kris, MD,
Mark R. Somerfield, PhD, and Gary H. Lyman, MD
Overview: In 2011, ASCO updated its guideline for the use of

antiemetics in oncology, informed by a systematic review of
the medical literature. This is an abbreviated version of that
guideline, which is available in full at www.asco.org/guide
lines/antiemetics. Key changes from the prior update in 2006
include the following: Combined anthracycline and cyclophosphamide regimens were reclassified as highly emetic. Patients
who receive this combination or any highly emetic agents
should receive a 5-HT 3 receptor antagonist, dexamethasone,
and an NK 1 receptor antagonist. A large trial validated the
equivalency of fosaprepitant, a single-day intravenous formulation, with aprepitant; either therapy is appropriate. Preferential use of palonosetron is recommended for moderate
IKE ALL ASCO guidelines, these recommendations are

based on a systematic review of the literature, with
recommendations developed by an expert panel, followed by
review by ASCOs Clinical Practice Guideline Committee
(CPGC), blinded external review, ASCO Board of Directors
review, and blinded peer review by the Journal of Clinical
Oncology.
Literature Review and Analysis Methods
The initial search for this guidelines systematic review

was based on an Agency for Healthcare Research and
Quality (AHRQ)-funded Evidence-Based Practice Center report completed at Oregon Health and Science University
(OHSU).3 The dates of the OHSU literature search of Medline were 1966 through October 2008. That evidence review
was limited to trials including the newer antiemetics:
aprepitant (the NK1 receptor antagonist) and the 5-HT3
receptor antagonists. Initially, two literature searches were
completed by ASCO staff in Medline. The first included all
relevant search terms, overlapping minimally with the
OHSU search, from September 2008 through December
2009. A second search, excluding the OHSU intervention
search terms, overlapped briefly with the search for the 2006
ASCO update, ranging from February 2004 to February
2010. This second search was designed to identify new
adjunctive therapy. The Cochrane Collaboration Library
electronic database was also searched through 2011, using
the terms emesis, vomiting, and nausea. Data presented at
the ASCO and the Multinational Association of Supportive
Care in Cancer (MASCC) annual meetings was also searched
systematically using the terms vomiting, emesis, and
nausea, but only presentations or posters were included.
Data presented only in abstract form was excluded.
Eligible reports were identified in multiple rounds of
review by ASCO staff. Full-text copies were obtained for
assessment of inclusion/exclusion criteria. Articles that provisionally met inclusion criteria underwent data extraction
by ASCO staff for patient characteristics, study design and
quality, interventions, outcomes, and adverse events. Evidence summary tables (Data Supplement, available online
at www.asco.org/guidelines/antiemetics) were reviewed for
accuracy and completeness by an ASCO staff member who
was not involved in their original preparation.
532
emetic risk regimens, combined with dexamethasone. For

low-risk agents, patients can be offered dexamethasone before the first dose of chemotherapy. Patients undergoing high
emetic risk radiation therapy should receive a 5-HT 3 receptor
antagonist before each fraction and for 24 hours following
treatment and may receive a 5-day course of dexamethasone
during fractions 1 to 5. Continued symptom monitoring
throughout therapy is recommended. Clinicians often underestimate the incidence of nausea, which is not as well controlled
as vomiting. Detailed information about the development of the
guideline as well as practice tools are available at www.asco.
org/guidelines/antiemetics.
Literature Review Results
The literature search yielded a total of 271 unique citations from Medline and 48 from the MASCC and ASCO
meetings. Additional materials evaluated were from the
personal libraries of Update Committee members. Of those,
36 reports met inclusion criteria (previously described) and
were selected for full-text review. Eleven (30.6%) of those
included were either posters or presentations from meetings.
Nine studies evaluated antiemetic regimens according
to emetic risk, six of which applied to highly emetic
chemotherapy4-8 and the remainder to moderately
emetic.9-11 Five trials evaluated the comparative efficacy of
5-HT3 receptor antagonists including a systematic review
from the Cochrane Collaboration12-16; five described findings
from dosing studies specifically for palonosetron.17-21 Two
trials described new delivery methods of two previously
approved therapies.22,23 A number of studies assessed special populations. Three trials detailed results in patients
undergoing myeloablative therapy before transplant,24-26
two described efforts in patients receiving multiday chemotherapy regimens,27,28 and three trials evaluated antiemetic
therapies in pediatric patients undergoing cancer
therapy.29-31
Three studies examined complementary therapies in patients receiving cancer treatment.32-34 Two studies that
specifically considered therapy for delayed nausea and vomiting were identified.35,36 Among the studies reviewed, only
one trial evaluating therapy for patients undergoing radiation was identified.37
Update Panel
A multidisciplinary expert panel was assembled in accordance with ASCOs Conflict of Interest Management Proce-
From the Memorial-Sloan Kettering Cancer Center, New York, NY; American Society of
Clinical Oncology, Alexandria, VA; Lahey Clinic Medical Center; Burlington, MA; Duke
University, Durham, NC.
Address reprint requests to Ethan Basch, MD, MSc, Genitourinary Oncology Service,
Department of Medicine, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New
York, NY 10065; email: basche@mskcc.org.
1092-9118/10/1-10
ANTIEMETICS: ASCO GUIDELINE UPDATE
dures for Clinical Practice Guidelines (Procedures,

summarized at www.asco.org/guidelinescoi). The panel developed the recommendations based on the systematic review, in keeping with ASCO standard practice (www.
asco.org/guidelines).
Guideline Recommendations
Highly and moderately emetogenic antineoplastic agents

have the potential to induce both acute ( 24 hours) and
delayed ( 24 hours) nausea and vomiting following chemotherapy. The guideline recommendations include prophylaxis for both types of nausea and vomiting where
appropriate.
This guideline update includes the most recent recommendations developed by the Update Committee (Table 1).
Detailed discussions of the evidence are available in the full
guideline, available at www.asco.org/guidelines/antiemetics
or in the published guideline in the Journal of Clinical
Oncology.
A table with intravenous agents organized by emetic
risk (Table 2) is included. The intravenous risk stratification schema was originally published in 199738 and was
updated at the MASCC/European Society for Medical Oncology (ESMO) 2009 consensus conference.39 The modified
stratification from MASCC was adopted by ASCO for this
guideline update.40 Dosing schedules are also detailed
herein (Table 3).
Chemotherapy-Induced Nausea and Vomiting (CINV)
Clinical Question 1. What is the optimal treatment to

prevent nausea and vomiting from highly emetogenic antineoplastic agents?
KEY POINTS
A list of drugs and radiation therapies associated

with high, moderate, and low emetic risk, as well as a
list of recommended antiemetic regimens based on
emetic risk, is provided in the ASCO Antiemetic
Guideline (www.asco.org/guidelines/antiemetics) and
in this abbreviated version of the guideline.
Continued symptom monitoring throughout therapy
is recommended. Clinicians often underestimate the
incidence of nausea.
Patients receiving highly emetic agents should receive a 5-HT3 receptor antagonist, dexamethasone,
and an NK1 receptor antagonist. Combined anthracycline and cyclophosphamide regimens are now classified as highly emetic.
Preferential use of palonosetron is now recommended
for moderate emetic risk regimens, combined with
dexamethasone. For low-risk agents, patients can be
offered dexamethasone before the first dose of chemotherapy.
Patients undergoing high emetic risk radiation therapy should receive a 5-HT3 receptor antagonist before
each fraction and for 24 hours following treatment
and may receive a 5-day course of dexamethasone
during fractions 1 to 5.
Recommendation 1. The three-drug combination of an

NK1 receptor antagonist (day 1 through 3 for aprepitant; day
1 only for fosaprepitant), a 5-HT3 receptor antagonist (day 1
only), and dexamethasone (day 1 through 3 or 1 4) is
recommended for patients receiving highly emetogenic chemotherapy. This recommendation is unchanged since the
2006 update, but reworded for clarification. The Update
Committee also recommended reclassification of the combined anthracycline and cyclophosphamide (AC) regimen as
highly emetogenic.
prevent nausea and vomiting from moderately emetogenic
antineoplastic agents?
Recommendation 2. The two-drug combination of palonosetron (day 1 only) and dexamethasone (day 1 through 3) is
recommended for patients receiving moderately emetogenic
chemotherapy. If palonosetron is not available, clinicians
may substitute a first-generation 5-HT3 receptor antagonist,
preferably granisetron or ondansetron. Limited evidence
also supports adding aprepitant to the combination. Should
clinicians opt to add aprepitant in patients receiving
moderate-risk chemotherapy, any one of the 5-HT3 receptor
antagonists is appropriate.
prevent nausea and vomiting from low emetogenic antineoplastic agents?
Recommendation 3. A single 8 mg dose of dexamethasone
before chemotherapy is suggested. No change since 2006.
prevent nausea and vomiting from minimally emetogenic
antineoplastic agents?
Recommendation 4. No antiemetic should be administered routinely before or after chemotherapy. No change
from the original guideline.
prevent nausea and vomiting from combination chemotherapy?
Recommendation 5. Patients should be administered
antiemetics appropriate for the component chemotherapeutic (antineoplastic) agent of greatest emetic risk. No
change from the original guideline. Anthracycline-cyclophosphamide combinations are now classified as highly
emetogenic.
Clinical Question 6. What is the role of adjunctive drugs
for nausea and vomiting induced by cancer treatments?
Recommendation 6. Lorazepam or diphenhydramine are
useful adjuncts to antiemetic drugs, but are not recommended as single-agent antiemetics. No change since 2006.
Clinical Question 7. What is the role of complementary
and alternative medicine therapies to prevent or control
nausea and vomiting induced by chemotherapy?
Recommendation 7. No published randomized controlled
trial data which met inclusion criteria are currently available to support a recommendation about such therapies.
Special Populations

prevent nausea and vomiting associated with cancer therapy for pediatric patients?
Recommendation 8. The combination of a 5-HT3 receptor
antagonist plus a corticosteroid is suggested before chemotherapy in children receiving chemotherapy of high or mod-
533
BASCH ET AL
Table 1. Summary of Recommendations
2006
Chemotherapy-Induced Nausea and Vomiting

Highly Emetogenic Agents
The three-drug combination of a 5-HT3 receptor antagonist,
dexamethasone, and aprepitant is recommended before
chemotherapy. In all patients receiving cisplatin and all other
agents of high emetic risk, the two-drug combination of
dexamethasone and aprepitant is recommended. The Update
Committee no longer recommends the combination of a 5HT3 serotonin receptor antagonist and dexamethasone on
days 2 and 3.
Moderately Emetogenic Agents
Low Emetogenic Agents

Minimally Emetogenic Agents
Combination Chemotherapy
The three-drug combination of a 5-HT3 receptor antagonist,

dexamethasone, and aprepitant is recommended for patients
receiving AC.
For patients receiving chemotherapy of moderate emetic risk
other than AC, we recommend the two-drug combination of a
5-HT3 receptor antagonist and dexamethasone. In patients
receiving AC, aprepitant as a single agent is recommended
on days 2 and 3. For all other chemotherapies of moderate
emetic risk, single-agent dexamethasone or a 5-HT3 receptor
antagonist is suggested for the prevention of emesis on days 2
and 3.
Dexamethasone 8 mg is suggested. No routine preventive use of

antiemetics for delayed emesis is suggested.
No change from the original guideline. No antiemetic should be
administered routinely before or after chemotherapy.
No change from the original guideline. Patients should be
administered antiemetics appropriate for the chemotherapeutic
agent of greatest emetic risk.
Adjunctive Drugs
Lorazepam and diphenhydramine are useful adjuncts to

antiemetic drugs, but are not recommended as single agents.
Complementary Therapy
New question for 2011 update.
Pediatric Patients
The combination of a 5-HT3 antagonist plus a corticosteroid is

suggested before chemotherapy in children receiving
chemotherapy of high or moderate emetic risk. Due to
variation of pharmacokinetic parameters in children, higher
weight-based doses of 5-HT3 antagonists than those used in
adults may be required for antiemetic protection.
High-Dose Chemotherapy with

Stem Cell or Bone Marrow
Transplant
No change from original guideline. A 5-HT3 receptor antagonist

antiemetic combined with dexamethasone is suggested.
Aprepitant should be considered although evidence to support
its use specifically in these patients is lacking.
No change from the original guideline. It is suggested that
antiemetics appropriate for the risk class of the chemotherapy,
as outlined above, be administered for each day of the
chemotherapy and for 2 d after, if appropriate.
Multiday Chemotherapy
534
2011 Update
The three-drug combination of an NK1 receptor antagonist (days

13 for aprepitant; day 1 only for fosaprepitant), a 5-HT3
receptor antagonist (day 1 only), and dexamethasone (days
13 or 14) is recommended for patients receiving highly
emetogenic chemotherapy. This recommendation is
unchanged since the 2006 update, but reworded for
clarification.
The Update Committee also recommended reclassification of the
combined anthracycline and cyclophosphamide regimen as
highly emetogenic.
The two-drug combination of palonosetron (day 1 only) and
dexamethasone (days 13) is recommended for patients
receiving moderately emetogenic chemotherapy. If
palonosetron is not available, clinicians may substitute a first
generation 5-HT3 receptor antagonist, preferably granisetron
or ondansetron.
Limited evidence also supports adding aprepitant to the

combination. Should clinicians opt to add aprepitant in
patients receiving moderate-risk chemotherapy, any one of the
5-HT3 receptor antagonists is appropriate.
A single 8 mg dose of dexamethasone before chemotherapy is
suggested. No change since 2006.
No antiemetic should be administered routinely before or after
chemotherapy. No change from the original guideline.
Patients should be administered antiemetics appropriate for the
component chemotherapeutic (antineoplastic) agent of greatest
emetic risk. No change from the original guideline.
Anthracycline-cyclophosphamide combinations are now
classified as highly emetogenic.
Lorazepam or diphenhydramine are useful adjuncts to
antiemetic drugs, but are not recommended as single agent
antiemetics. No change since 2006.
No published randomized controlled trial data which met
inclusion criteria are currently available to support a
recommendation about such therapies.
The combination of a 5-HT3 receptor antagonist plus a
corticosteroid is suggested before chemotherapy in children
receiving chemotherapy of high or moderate emetic risk. Due
to variation of pharmacokinetic parameters in children, higher
weight-based doses of 5-HT3 receptor antagonists than those
used in adults may be required for antiemetic protection. No
change since 2006.
A 5-HT3 receptor antagonist combined with dexamethasone is
suggested. Aprepitant should be considered, although
evidence to support its use is limited.
It is suggested that antiemetics appropriate for the emetogenic
risk class of the chemotherapy be administered for each day
of the chemotherapy and for two days after, if appropriate.
No change from the original guideline.
The Update Committee suggests, based on limited data, that
patients receiving five-day cisplatin regimens be treated with a
5-HT3 receptor antagonist in combination with
dexamethasone and aprepitant.

Table 1. Summary of Recommendations (Contd)
2006
2011
Chemotherapy-Induced Nausea and Vomiting (contd)

Emesis or Nausea Despite
No change from original guideline. The Update Committee
Optimal Prophylaxis
suggests that clinicians (1) conduct a careful re-evaluation of
emetic risk, disease status, concurrent illnesses, and
medications; (2) ascertain that the best regimen is being
administered for the emetic risk; (3) consider adding an
lorazepam or alprazolam to the regimen; and (4) consider
substituting a high-dose intravenous metoclopramide for the
5-HT3 antagonist or adding a dopamine antagonist to the
regimen.
Anticipatory Nausea and

Vomiting
No change since the original guideline. Use of the most active

antiemetic regimens appropriate for the chemotherapy being
administered to prevent acute or delayed emesis is suggested.
Such regimens may be used with the initial chemotherapy,
rather than assessing the patients emetic response with less
effective treatment. If anticipatory emesis occurs, behavioral
therapy with systematic desensitization is effective and
suggested.
Radiation Induced Nausea and Vomiting
High Risk
suggests administration a 5-HT3 antagonist with or without a
corticosteroid before each fraction and for at least 24 h after.
There is no change from the original guideline.
Moderate Risk
The Update Committee recommends a 5-HT3 antagonist before

each fraction.
Low Risk

recommends a 5-HT3 antagonist before each fraction.
Minimal Risk

suggests that treatment be administered on an as-needed
basis only. Dopamine or serotonin receptor antagonists are
advised. Antiemetics should be continued prophylactically for
each remaining radiation treatment day.
No change. Patients should receive rescue therapy with a
dopamine- receptor antagonists or a 5-HT3 receptor
antagonist. Antiemetics should be continued prophylactically
for each remaining radiation treatment day.
Combined Chemotherapy and

Radiation Therapy
erate emetic risk. Because of variation of pharmacokinetic

parameters in children, higher weight-based doses of 5-HT3
receptor antagonists than those used in adults may be
required for antiemetic protection. No change since 2006.
prevent nausea and vomiting in patients who are undergoing high-dose chemotherapy with stem cell or bone marrow
transplant?
Recommendation 9. A 5-HT3 receptor antagonist combined with dexamethasone is recommended. Aprepitant
should be considered, although evidence to support its use is
limited.
Question 10. What is the optimal treatment to prevent
nausea and vomiting for patients receiving multiday chemotherapy?
Recommendation 10. It is suggested that antiemetics appropriate for the emetogenic risk class of the chemotherapy
be administered for each day of the chemotherapy and for
2 days after, if appropriate. No change from the original
Clinicians should:
(1) Re-evaluate emetic risk, disease status, concurrent
illnesses, and medications;
(2) Ascertain that the best regimen is being administered for
the emetic risk;
(3) Consider adding lorazepam or alprazolam to the
regimen; and
(4) Consider adding olanzapine to the regimen or substituting
high-dose intravenous metoclopramide for the 5-HT3
receptor antagonist or adding a dopamine antagonist to
the regimen.
Use of the most active antiemetic regimens appropriate for the
chemotherapy being administered to prevent acute or delayed
emesis is suggested. Such regimens should be used with initial
chemotherapy, rather than assessing the patients emetic
response with less effective treatment. If anticipatory emesis
occurs, behavioral therapy with systematic desensitization is
effective and suggested. No change since the original
guideline.
Based on extrapolation of evidence, the Update Committee
recommends that all patients should receive a 5-HT3 receptor
antagonist before each fraction and for at least 24 h after
completion of radiotherapy. Patients should also receive a
five-day course of dexamethasone before fractions 15.
The Update Committee recommends that patients receive a 5HT3 receptor antagonist before each fraction for the entire
course of radiotherapy. Patients may be offered a short
course (fractions 15) of dexamethasone before treatment.
The Update Committee recommends a 5-HT3 receptor antagonist
alone as either prophylaxis or rescue. For patients who
experience RINV while receiving rescue therapy only,
prophylactic treatment should continue until radiotherapy is
complete.
Patients should receive rescue therapy with either a dopamine
receptor antagonist or a 5-HT3 receptor antagonist.
Prophylactic antiemetics should continue throughout radiation
treatment if a patient experiences RINV while receiving rescue
therapy.
Patients should receive antiemetic prophylaxis according to the
emetogenicity of chemotherapy, unless the emetic risk with the
planned radiotherapy is higher. No change from the original
guideline.
guideline. The Update Committee suggests, based on limited

data, that patients receiving 5-day cisplatin regimens be
treated with a 5-HT3 receptor antagonist in combination
with dexamethasone and aprepitant.
Clinical Question 11. What is the optimal antiemetic
regimen for patients who experience nausea and vomiting
secondary to cancer therapy despite optimal prophylaxis?
Recommendation 11. Language from the 2006 guideline
was reformatted for clarity. Clinicians should:
(1) Re-evaluate emetic risk, disease status, concurrent
illnesses, and medications
(2) Ascertain that the best regimen Is being administered
for the emetic risk
(3) Consider adding lorazepam or alprazolam to the regimen
(4) Consider adding olanzapine to the regimen or substituting high-dose intravenous metoclopramide for the 5-HT3
receptor antagonist or adding a dopamine antagonist to the
regimen.
535
BASCH ET AL
Table 2. Emetic Risk of Intravenous Antineoplastic Agents*
High 90%
Moderate
Low
Minimal
Carmustine
Cisplatin
Cyclophosphamide 1500 mg/m2
Dacarbazine
Azacitidine
Alemtuzumab
Bendamustine
Carboplatin
Clofarabine
Cyclophosphamide 1500 mg/m2
Cytarabine 1000 mg/m2
5-Fluorouracil
Bortezomib
Carbezetaxel
Catumaxumab
Cytarabine 1000 mg/m2
Docetaxel
Doxorubicin HCL liposome injection
Etoposide
Gemcitabine
Ixabepilone
2-Chlorodeoxyadenosine
Bevacizumab
Bleomycin
Busulfan
Cetuximab
Fludarabine
Dactinomycin
Mechlorethamine
Streptozotocin
Daunorubicin**
Doxorubicin**
Epirubicin**
Idarubicin**
Ifosfamide
Irinotecan
Oxaliplatin
Methotrexate
Mitomycin
Mitoxantrone
Paclitaxel
Panitumumab
Pemetrexed
Temsirolimus
Topotecan
Trastuzumab
Pralatrexate
Rituximab
Vinblastine
Vincristine
Vinorelbine
* List is not exhaustive.

** These anthracyclines, when combined with cyclophosphamide, are now
designated as high emetic risk.
Clinical Question 12. What treatment options are available for patients who experience anticipatory nausea and
vomiting?
Recommendation 12. Use of the most active antiemetic
regimens appropriate for the chemotherapy being administered to prevent acute or delayed emesis is suggested. Such
regimens should be used with initial chemotherapy, rather
than assessing the patients emetic response with less effective treatment. If anticipatory emesis occurs, behavioral
therapy with systematic desensitization is effective and
suggested. No change since the original guideline.
Radiation-Induced Nausea and Vomiting (RINV)
This guideline update includes an updated risk stratification table according to site of radiation treatment. MASCC
updated the radiation therapy emetic risk table at the
MASCC/ESMO 2009 consensus conference (Table 4) and it
was adopted by ASCO for this guideline update.40 Dosing
schedules, according to risk level, are detailed in Table 5.
Clinical Question 13. What is the optimal prophylaxis for
nausea and vomiting caused by high emetic risk radiation
therapy?
Recommendation 13. Based on extrapolation of indirect
evidence, the Update Committee recommends that all patients should receive a 5-HT3 receptor antagonist before
each fraction and for at least 24 hours after completion of
radiotherapy. Patients should also receive a 5-day course of
dexamethasone during fractions 1 to 5.
Clinical Question 14. What is the optimal prophylaxis for
nausea and vomiting caused by moderate emetic risk radiation therapy?
Recommendation 14. The Update Committee recommends that patients receive a 5-HT3 receptor antagonist
536
Table 3. Antiemetic Dosing by Chemotherapy Risk Category*

Day of Chemotherapy
High Emetic Risk (HEC)

NK1 Antagonist
Aprepitant
Fosaprepitant
Subsequent Days
125 mg oral
80 mg oral; days
2 and 3
150 mg IV
5-HT3 Receptor Antagonist

Granisetron
2 mg oral OR
1 mg or 0.01 mg/kg IV
Ondansetron
8 mg oral twice daily OR
8 mg or 0.15 mg/Kg IV
Palonosetron
0.50 mg oral OR
0.25 mg IV
Dolasetron
100 mg oral only
Tropisetron
5 mg oral OR 5 mg IV
Ramosetron
0.3 mg IV
Corticosteroidb
Dexamethasone
12 mg oral or IV
Moderate Emetic Risk (MEC)c

Palonosetron
0.25 mg IV OR
0.50 mg oral
Corticosteroid
8 mg oral or IV
Dexamethasone
Low Emetic Risk
Corticosteroid
Dexamethasone
8 mg oral or IV; days

23 or days 24
8 mg; days 2 and 3
8 mg oral or IV
* For patients receiving multiday chemotherapy, clinicians must first determine

the emetic risk of the agent(s) included in the regimen. Patients should receive
the agent of the highest therapeutic index daily during chemotherapy and for two
days thereafter. Patients can also be offered the granisetron transdermal patch
(Sancuso) that delivers therapy over multiple days rather than taking a serotonin
antagonist daily.
a
Includes AC (combination of anthracycline and cyclophosphamide).
b
The dexamethasone dose is for patients who are receiving the recommended
three-drug regimen for highly emetic chemotherapy. If patients do not receive
aprepitant, the dexamethasone dose should be adjusted to 20 mg on day 1 and
16 mg on days 2 4.
c
Clinicians who choose to utilize an NK 1 antagonist should follow HEC dosing. Importantly, corticosteroid is only given on day one; dexamethasone dose is
12 mg.
before each fraction for the entire course of radiotherapy.

Patients may be offered a short course of dexamethasone
during fractions 1 to 5.
manage nausea and vomiting associated with low emetic
risk radiation therapy?
Recommendation 15. The Update Committee recommends a 5-HT3 receptor antagonist alone as either prophylaxis or rescue. For patients who experience RINV while
Table 4. Emetic Risk by Site of Radiation Therapy 40
High
Moderate
Low
Minimal
Total Body Irradiation

Total Nodal Irradiation
Upper Abdomen
Upper Body Irradiation
Half Body Irradiation
Cranium
Craniospinal
Head and Neck
Lower Thorax Region
Pelvis
Extremities
Breast

Table 5. Antiemetic Dosing by Radiation Risk Category
Dose
Schedule
High Emetic Risk

Granisetrona
Ondansetrona
Palonosetronb
Dolasetron
Tropisetron
Corticosteroid
Dexamethasone
2 mg oral
5-HT3 antagonist before each fraction throughout XRT. Continue for at

least 24 h following completion of XRT.
8 mg oral twice daily

0.50 mg oral
0.25 mg IV
100 mg oral only
5 mg oral or IV
4 mg oral or IV
Before fractions 15
Moderate Emetic Risk

Any of the above listed agents are acceptable, note preferred options
Corticosteroid
Dexamethasone
4 mg IV or oral
5-HT3 antagonist before each fraction throughout XRT

Before fractions 15
Low Emetic Risk

5-HT3 either as rescue or prophylaxis. If rescue is utilized, then

prophylactic therapy should be given until the end of XRT.
Minimal Emetic Risk

Dopamine Receptor Antagonist
Metoclopramide
Prochlorperazine
Patients should be offered either class as rescue therapy. If rescue is

utilized, then prophylactic therapy should be given until the end of XRT.
20 mg oral
10 oral or IV
Abbreviations: IV, intravenous; XRT, radiation therapy; bid, twice daily; qid, four times daily; q, every; h, hours;
a
Preferred agents.
b
No data are currently available on the appropriate dosing frequency with palonosetron in this setting. The Update Committee suggests dosing every second or third
day may be appropriate for this agent.
receiving rescue therapy only, prophylactic treatment

should continue until radiotherapy is complete.
manage nausea and vomiting associated with minimal
emetic risk radiation therapy?
Recommendation 16. Patients should receive rescue therapy with either a dopamine receptor antagonist or a 5-HT3
receptor antagonist. Prophylactic antiemetics should continue throughout radiation treatment if a patient experiences RINV while receiving rescue therapy.
manage nausea and vomiting during concurrent radiation
and chemotherapy?
Recommendation 17. Patients should receive antiemetic
prophylaxis according to the emetogenicity of chemotherapy,
unless the emetic risk with the planned radiotherapy is
higher. No change from the original guideline.
Drug Formulations, Agent Dosing
A study published in 2007 compared an orally disintegrating tablet (ODT) of ondansetron with a standard tablet
(Data Supplement).22 No differences were reported in emesis or nausea control between the two agents. Notably, it is
not clear whether this study was designed as a nonequivalence trial a priori. The ODT formulation is an acceptable
alternative to the standard ondansetron tablet.
Two antiemetic agents received regulatory approval in
alternative formulations since the 2006 update. Granisetron
is also available as a transdermal patch that delivers therapy over 7 days. As described earlier, this is an option for
patients receiving multiday, high-risk chemotherapy regimens.47 The panel also suggests that the granisetron patch
may be useful for patients undergoing high- or moderaterisk radiation.
Oral palonosetron was approved by the U.S. Food and
Drug Administration (FDA) in 2008.48 Data detailing antiemetic similarity of the oral and intravenous formulations
and agent safety was presented at the 2007 European
Conference of Clinical Oncology meeting.18 This trial also
supported the 0.50 mg dose.
Three studies assessed dosing of intravenous palonosetron. A meta-analysis of eight studies suggested similar
outcomes (Data Supplement) with respect to complete response among patients treated with either 0.25 mg or
0.75 mg doses.20 The other two trials reported findings that
a dose of 0.075 mg is clearly inferior to both 0.25 mg and
0.75 mg.19,21
Patient and Clinician Communication
Clinicians are encouraged to provide patients with a

prescription for a rescue antiemetic therapy before the
patient leaves the treatment facility on the first day of
treatment. Data suggest that physicians frequently underestimate rates of nausea and vomiting secondary to radiation therapy and chemotherapy.49
In order to ensure optimal symptom management, clini-
537
BASCH ET AL
cians should assess symptoms throughout the course of

therapy. Clinicians and clinical researchers should consider
collecting direct reports of symptom presence and severity
by patients with a checklist. Patient response to treatment
may change over time, thus requiring ongoing assessments
and modification to antiemetic strategies. For example, the
National Cancer Institute has developed a Patient-Reported
Outcomes version of its Common Terminology Criteria for
Adverse Events (PRO-CTCAE), which includes two items to
assess nausea50:
1) In the last 7 days, how OFTEN did you have NAUSEA
(Never/Rarely/Occasionally/Frequently/Almost constantly)
2) In the last 7 days, what was the SEVERITY of your
NAUSEA at its WORST
(None/Mild/Moderate/Severe/Very severe).
Clinicians and patients are also encouraged to discuss
costs of treatment, particularly to assess if cost is prohibitive, a hardship to patients, or if it may impact treatment
compliance.
Additional Resources
The data supplement, including evidence tables, and clinical tools and resources can be found at www.asco.org/
guidelines/antiemetics. Patient information is available
there as well and also at www.cancer.net.
Acknowledgments
Thanks to additional Guideline Panel members Petra C.
Feyer, Maurice Chesney, Rebecca Anne Clark-Snow, Anne Ma-
rie Flaherty, Barbara Freundlich, Gary Morrow, Kamakshi V.

Rao, and Rowena N Schwartz. The Panel also wishes to thank
Dr. Steven Grunberg, Dr. Kristopher Dennis, Dr. Edward Chow,
Dr. Carlo DeAngelis, Dr. Amy Abernethy, Dr. Michael Danso,
Dr. James L. Abbruzzese, Dr. Robert Langdon, the ASCO
Clinical Practice Guidelines Committee, and the ASCO Board of
Directors for their thoughtful reviews of guideline drafts. Special thanks to Shauniece Morris, Pamela B. Mangu, and Sarah
Temin for their assistance data checking and data extraction.
APPENDIX: Updated Committee Members, 2011

Ethan Basch, MD, Co-Chair
Gary H. Lyman, MD, Co-Chair
Paul J. Hesketh, MD,
Steering Committee
Mark G. Kris, MD,
Steering Committee
Maurice Chesney
Petra C. Feyer MD
Anne Marie Flaherty, RN
Barbara Freundlich, BA
Gary Morrow, PhD
Kamakshi V. Rao, PharmD, BCOP, CPP
Rowena N Schwartz, PharmD, BCOP
Memorial-Sloan Kettering Cancer Center

Duke University
Lahey Clinic Medical Center
Patient Representative
Lawrence Memorial Hospital
Oncology Center
Vivantes Clinic of Radiooncology and
Nuclear Medicine
Patient Representative
University of Rochester Cancer Center
University of North Carolina Hospital
The Johns Hopkins Hospital
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Ethan Basch*
Ann Alexis Prestrud*
Paul J. Hesketh
Helsinn; Merck;
Tesaro
Mark G. Kris*
Mark R. Somerfield*
Gary H. Lyman
Amgen
REFERENCES
1. Gralla RJ, Osoba D, Kris MG, et al. Recommendations for the use of
antiemetics: evidence-based, clinical practice guidelines. American Society of
Clinical Oncology. J Clin Oncol. 1999;17:2971-2994.
2. Kris MG, Hesketh PJ, Somerfield MR, et al. American Society of Clinical
Oncology guideline for antiemetics in oncology: Update 2006. J Clin Oncol.
2006;24:2932-2947.
3. Peterson K, McDonagh MS, Carson S, et al. Drug class review: Newer
antiemetics. Update 1. http://www.ohsu.edu/drugeffectiveness/reports/final.
cfm. Accessed September 1, 2011.
4. Grunberg SM, Chua DT, Maru A, et al. Phase III randomized doubleblind study of single-dose fosaprepitant for prevention of cisplatin-induced
nausea and vomiting. J Clin Oncol. 2010;28:641s (suppl; abstr 9021).
5. Herrington JD, Jaskiewicz AD, Song J. Randomized, placebo-controlled,
pilot study evaluating aprepitant single dose plus palonosetron and dexamethasone for the prevention of acute and delayed chemotherapy-induced
nausea and vomiting. Cancer. 2008;112:2080-2087.
6. Hoshi E, Takahashi T, Takagi M, et al. Aprepitant prevents
chemotherapy-induced nausea and vomiting in Japanese cancer patients
receiving high-does cisplatin: a multicenter randomized, double-blind, placebocontrolled study (abstract P-20). Presented at: 20th Anniversary International MASCC/ISOO Symposium, St Gallen, Switzerland; June 2007;30.
538
7. Navari R, Gray S, Kerr A. Olanzapine versus aprepitant for the prevention of chemotherapy-induced nausea and vomiting (Cinv): a randomized trial
(abstract 02-010). Presented at: 2010 MASCC/ISOO Symposium, Vancouver,
Canada; June 2010;26.
8. Yeo W, Mo FK, Suen JJ, et al. A randomized study of aprepitant,
ondansetron and dexamethasone for chemotherapy-induced nausea and vomiting in Chinese breast cancer patients receiving moderately emetogenic
chemotherapy. Breast Cancer Res Treat. 2009;113:529-535.
9. Aapro M, Fabi A, Nole F, et al. Double-blind, randomised, controlled
study of the efficacy and tolerability of palonosetron plus dexamethasone for
1 day with or without dexamethasone on days 2 and 3 in the prevention of
nausea and vomiting induced by moderately emetogenic chemotherapy. Ann
Oncol. 2010;21:1083-1088.
10. Celio L, Bajetta E, Denaro A, et al. Single-day regimen of palonosetron
(PALO) and dexamethasone (DEX) for the prevention of emesis associated
with moderately emetogenic chemotherapy (MEC): subgroup analysis from a
randomized phase III trial. J Clin Oncol. 2009;27 (suppl; abstr 9620).
11. Rapoport BL, Jordan K, Boice JA, et al. Aprepitant for the prevention
of chemotherapy-induced nausea and vomiting associated with a broad range
of moderately emetogenic chemotherapies and tumor types: a randomized,
double-blind study. Support Care Cancer. 2010;18:423-431.

12. Aapro MS, Grunberg SM, Manikhas GM, et al. A phase III, doubleblind, randomized trial of palonosetron compared with ondansetron in preventing chemotherapy-induced nausea and vomiting following highly
emetogenic chemotherapy. Ann Oncol. 2006;17:1441-1449.
13. Saito M, Aogi K, Sekine I, et al. Palonosetron plus dexamethasone
versus granisetron plus dexamethasone for prevention of nausea and vomiting during chemotherapy: a double-blind, double-dummy, randomised, comparative phase III trial. Lancet Oncol. 2009;10:115-124.
14. Yu Z, Liu W, Wang L, et al. The efficacy and safety of palonosetron
compared with granisetron in preventing highly emetogenic chemotherapyinduced vomiting in the Chinese cancer patients: a phase II, multicenter,
randomized, double-blind, parallel, comparative clinical trial. Support Care
Cancer. 2009;17:99-102.
15. Ezzo J, Richardson MA, Vickers A, et al. Acupuncture-point stimulation for chemotherapy-induced nausea or vomiting. Cochrane Database of
Systematic Reviews, Issue 1, 2010.
16. Ho CL, Su WC, Hsieh RK, et al. A randomized, double-blind, parallel,
comparative study to evaluate the efficacy and safety of ramosetron plus
dexamethasone injection for the prevention of acute chemotherapy-induced
nausea and vomiting. Jpn J Clin Oncol. 2010;40:294-301.
17. Boccia RV, Gonzalez EF, Pluzanska G, et al. Palonosetron (PALO),
administered orally or intravenously (IV), plus dexamethasone for prevention
of chemotherapy-induced nausea and vomiting (CINV). J Clin Oncol. 2008;26
18. Grunberg S, Voisin D, Zufferli M, et al. Oral palonosetron is as effective
as intravenous palonesetron: A Phase III dose ranging trial in patients
receiving moderately emetogenic chemotherapy. Presented at: 14th European
Conference of Clinical Oncology (ECCO); Barcelona, Spain, September 2007.
19. Maemondo M, Masuda N, Sekine I, et al. A phase II study of palonosetron combined with dexamethasone to prevent nausea and vomiting induced by highly emetogenic chemotherapy. Ann Oncol. 2009;20:1860-1866.
20. Raftopoulos H, Bria E, Gralla RJ, et al. Is there a preferred dose of
palonosetron? Results of an abstracted data (AD) meta-analysis (MA) of all
eight randomized double-blind (RDB) studies and the impact on guideline
considerations. J Clin Oncol. 2009;27 (suppl; abstr 9636).
21. Segawa Y, Aogi K, Inoue K, et al. A phase II dose-ranging study of
palonosetron in Japanese patients receiving moderately emetogenic chemotherapy, including anthracycline and cyclophosphamide-based chemotherapy. Ann Oncol. 2009;20:1874-1880.
22. Pectasides D, Dafni U, Aravantinos G, et al. A randomized trial to
compare the efficacy and safety of antiemetic treatment with ondansetron and
ondansetron zydis in patients with breast cancer treated with high-dose
epirubicin. Anticancer Res. 2007;27:4411-4417.
23. Boccia RV, Gordan LN, Clark G, et al. Efficacy and tolerability of
transdermal granisetron for the control of chemotherapy-induced nausea and
vomiting associated with moderately and highly emetogenic multi-day chemotherapy: a randomized, double-blind, phase III study. Support Care Cancer, 2010;19:1609-1617.
24. Giralt S, Mangan K, Maziarz R, et al. Palonosetron (PALO) for
prevention of chemotherapy-induced nausea and vomiting (CINV) in patients
receiving high-dose melphalan prior to stem cell transplant (SCT). J Clin
Oncol. 2008;26 (suppl; abstract 9617).
25. Stiff P, Fox-Geiman M, Kiley K, et al. Aprepitant Vs. Placebo Plus Oral
Ondansetron and Dexamethasone for the Prevention of Nausea and Vomiting
Associated with Highly Emetogenic Preparative Regimens Prior to Hematopoietic Stem Cell Transplantation: A Prospective, Randomized Double-Blind
Phase III Trial. Blood. 2009;114 (suppl; abstr 2267).
26. Walsh T, Morris AK, Holle LM, et al. Granisetron vs ondansetron for
prevention of nausea and vomiting in hematopoietic stem cell transplant
patients: results of a prospective, double-blind, randomized trial. Bone Marrow Transplant. 2004;34:963-968.
27. Brames MJ, Johnson EL, Nichols CR, et al. A Double-Blind, PlaceboControlled, Crossover Study Evaluating the Oral Neurokinin-1 Receptor
Antagonist Aprepitant in Combination with a 5HT3 Antagonist and Dexamethasone in Patients with Germ Cell Tumors Undergoing 5-day CisplatinBased Chemotherapy Regimens. A Hoosier Oncology Group Study (H.O.G.).
Support Care Cancer. 2010 (suppl 3; abstr 02-004):S67-S220.
28. Herrstedt J, Sigsgaard TC, Nielsen HA, et al. Randomized, doubleblind trial comparing the antiemetic effect of tropisetron plus metopimazine
with tropisetron plus placebo in patients receiving multiple cycles of multipleday cisplatin-based chemotherapy. Support Care Cancer. 2007;15:417-426.
29. Gore L, Chawla S, Petrilli A, et al. Aprepitant in adolescent patients for
prevention of chemotherapy-induced nausea and vomiting: A randomized,
double-blind, placebo-controlled study of efficacy and tolerability. Pediatr
Blood Cancer. 2009;52:242-247.
30. Sepulveda-Vildosola AC, Betanzos-Cabrera Y, Lastiri GG, et al. Palono-
setron hydrochloride is an effective and safe option to prevent chemotherapyinduced nausea and vomiting in children. Arch Med Res. 2008;39:601-606.
31. Pillai AK, Sharma KK, Gupta YK, et al. Anti-emetic effect of ginger
powder versus placebo as an add-on therapy in children and young adults
receiving high emetogenic chemotherapy. Pediatr Blood Cancer. 2011;56:234238.
32. Ryan JL, Heckler C, Dakhil SR, et al. Ginger for chemotherapy-related
nausea in cancer patients: A URCC CCOP randomized, double-blind, placebocontrolled clinical trial of 644 cancer patients. J Clin Oncol. 2009;27 (suppl;
abstr 9511).
33. Tan L, Liu J, Liu X, et al. Clinical research of olanzapine for prevention
of chemotherapy-induced nausea and vomiting. J Exp Clin Cancer Res.
2009;28:131.
34. Zick SM, Ruffin MT, Lee J, et al. Phase II trial of encapsulated ginger
as a treatment for chemotherapy-induced nausea and vomiting. Support Care
Cancer. 2009;17:563-572.
35. Fabi A, Ciccarese M, Metro G, et al. Oral ondansetron is highly active
as rescue antiemetic treatment for moderately emetogenic chemotherapy:
results of a randomized phase II study. Support Care Cancer. 2008;16:13751380.
36. Lajolo PP, de Camargo B, del Giglio A. Omission of day 2 of antiemetic
medications is a cost saving strategy for improving chemotherapy-induced
nausea and vomiting control: Results of a randomized phase III trial. Am J
Clin Oncol. 2009;32:23-26.
37. Wong RK, Paul N, Ding K, et al. 5-hydroxytryptamine-3 receptor
antagonist with or without short-course dexamethasone in the prophylaxis of
radiation induced emesis: A placebo-controlled randomized trial of the National Cancer Institute of Canada Clinical Trials Group (SC19). J Clin Oncol.
2006;24:3458-3464.
38. Hesketh PJ, Kris MG, Grunberg SM, et al. Proposal for classifying the
acute emetogenicity of cancer chemotherapy. J Clin Oncol. 1997;15:103-109.
39. Roila F, Herrstedt J, Aapro M, et al. Guideline update for MASCC and
ESMO in the prevention of chemotherapy- and radiotherapy-induced nausea
and vomiting: results of the Perugia consensus conference. Ann Oncol 21
Suppl. 2010;5:v232-243.
40. Gralla RJ, Roila F, Tonato M, et al. MASCC/ESMO Antiemetic Guideline 2010. http://data.memberclicks.com/site/mascc/MASCC_Guidelines_
English_2010.pdf. Accessed September 1, 2011.
41. Grunberg S, Chua D, Maru A, et al. Single-dose fosaprepitant for the
prevention of chemotherapy-induced nausea and vomiting associated with
cisplatin therapy: randomized, double-blind study protocol-EASE. J Clin
Oncol. 2011;29:1495-1501.
42. Billio A, Morello E, Clarke MJ. Serotonin receptor antagonists for
highly emetogenic chemotherapy in adults. Art. No. CD006272. DOI: 10.1002/
14651858.CD006272.pub2. Cochrane Database of Systematic Reviews, Issue
1, 2010.
43. Jordan K, Hinke A, Grothey A, et al. A meta-analysis comparing the
efficacy of four 5-HT3-receptor antagonists for acute chemotherapy-induced
emesis. Support Care Cancer. 2007;15:1023-1033.
44. Prescribing information for EMEND (fosaprepitant dimeglumine)
for
injection.
http://www.accessdata.fda.gov/drugsatfda_docs/label/2010/
022023s002s003s005lbl.pdf. Accessed September 1, 2011.
45. Celio L, Frustaci S, Denaro A, et al. Palonosetron in combination with
1-day versus 3-day dexamethasone for prevention of nausea and vomiting
following moderately emetogenic chemotherapy: a randomized, multicenter,
phase III trial. Support Care Cancer, 2010;19:1217-1225.
46. Phillips RS, Gopaul S. Antiemetic medication for prevention and
treatment of chemotherapy induced nausea and vomiting in childhood. Art.
No. CD007786. DOI: 10.1002/14651858.CD007786.pub2. Cochrane Database
of Systematic Reviews, Issue 9, 2010.
47. Granisetron Drug Information. http://www.accessdata.fda.gov/drug
satfda_docs/appletter/2008/022198s000ltr.pdf, 2010. Accessed September 1,
2011.
48. Food and Drug Administration Drug Information on Oral Palonosetron
http://www.accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm?fuseaction
Search.DrugDetails. Accessed September 1, 2011.
49. Basch E. The missing voice of patients in drug-safety reporting. N Engl
J Med. 2010;362:865-869.
50. Patient-Reported Outcomes version of the CTCAE (PRO-CTCAE).
https://wiki.nci.nih.gov/display/PROCTCAE/Patient-ReportedOutcomes
versionoftheCTCAE%28PRO-CTCAE%29. Accessed September 1,
2011.
51. U.S. Cancer Statistics Working Group. United States Cancer Statistics:
1999-2002 Incidence and Mortality Web-based Report http://apps.nccd.cdc.
gov/uscs/ U.S. Department of Health and Human Services, Centers for Disease
Control and Prevention and National Cancer Institute. Atlanta, 2005.
539
BASCH ET AL
52. American Cancer Society. Cancer facts and figures for African Americans 2005-2006. Atlanta, 2005.
53. Mead H, Cartwright-Smith L, Jones K, et al. Racial and ethnic
disparities in U.S. health care: A chartbook. New York: The Commonwealth
Fund, 2008.
54. Ries LAG, Eisner MP, Kossary CL, et al. SEER cancer statistics review
1973-1997. National Cancer Institute. Bethesda, MD, 2000.
55. Hickok JT, Roscoe JA, Morrow GR, et al. Nausea and emesis remain
540
significant problems of chemotherapy despite prophylaxis with 5-hydroxytryptamine-3 antiemetics: A University of Rochester James P. Wilmot Cancer
Center Community Clinical Oncology Program Study of 360 cancer patients
treated in the community. Cancer. 2003;97:2880-2886.
56. Guidance for Industry. Patient-Reported Outcome Measures: Use in
Medical Product Development to Support Labeling Claims. http://www.fda.
gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/
UCM193282.pdf. Accessed September 1, 2011.
Chemotherapy-Induced Nausea and Vomiting

Incidence and Prevalence
By Steven M. Grunberg, MD
Overview: Although chemotherapy-induced nausea and vomiting is recognized as having been an important problem
during the initial introduction of chemotherapy into the antineoplastic armamentarium, the assumption that this problem
has already been solved can restrict optimal management and
further advances. Underestimation of nausea and vomiting
may have many causes. If these toxicities are assumed to be
necessary properties of chemotherapy, then their incidence
may be taken for granted. If nausea and vomiting appear after
discharge from the clinic several days after chemotherapy,
these toxicities may not be reported because of poor recall or
HE JEWISH Passover service1 contains the story of the

four sons one wise, one evil, one simple, and one who
does not know how to ask a question. In the explanation of
the story, we learn that the most dangerous son is not the
evil one, but rather the one who does not know how to ask a
question. If we do not know where a lack of knowledge lies,
then we cannot begin to address that ignorance or to answer
that question. This is also true in medical science and has
become pertinent in recent years in several areas of oncology. Approximately 80% of patients with lung cancer have a
significant smoking exposure, and we have long assumed
that the remainder had a smoking exposure that had not yet
been identified. Now that we realize that there is a different
unanswered and unaddressed question concerning the role
of epidermal growth factor receptor mutations2 and EML4ALK translocations,3 we can begin to address this problem
with appropriate targeted agents. We have also always
known that smoking and drinking are important risk factors
for head and neck cancer, and we assumed that patients
without those risk factors simply had bad luck. Now that we
appreciate the role of human papillomavirus infection in
oropharyngeal cancers,4 we can begin to address the unique
natural history of this illness and the possibility of identifying specific remedies.
Although significant progress has been made in the prevention and control of chemotherapy-induced nausea and
vomiting, underappreciation of the incidence and prevalence
of this problem still limits our effectiveness in addressing its
true magnitude. This misunderstanding begins with the
systems used to grade intrinsic emetogenicity, which generally divide agents into those that are highly emetogenic,
moderately emetogenic, low emetogenic, or minimally emetogenic.5 The wide range of frequencies of emesis within a
given category, particularly moderately emetogenic chemotherapy (30% to 90% vomiting), can lead to significant
variability in interpreting the emetogenic risk of any chemotherapeutic agent. It must also be remembered that these
grading systems were specifically designed to describe the
acute emetogenicity of chemotherapy during the 24 hours
after administration of a single intravenous bolus dose
without concurrent administration of other chemotherapies
and without the administration of antiemetic agents. Numerous modifying factors are therefore not considered, including the presence of delayed nausea and vomiting (24 to
120 hours after chemotherapy) or anticipatory nausea and
because of efforts by patients to avoid unnecessary complaints. Physician education may be compromised if physicians see nausea and vomiting as population problems but not
problems for their own patients. Failure to recognize nausea
and vomiting as two distinct entities that may appear independently of each other can also limit understanding of the
prevalence of these problems and efforts at effective management. Continued attention to the impact of nausea and vomiting on the patient experience will be necessary to insure
optimal maintenance of quality of life.
vomiting (before chemotherapy), the possibility of extended

infusion or divided doses of chemotherapy, or the route of
administration. A single dose of an oral chemotherapy may
be less emetogenic than a single dose of an intravenous
formulation, but oral chemotherapy regimens are often
multiple daymultiple dose regimens and emetogenicity
may gradually increase as the latter days of administration
are reached.5 The emetogenicity of combination regimens is
also only rarely considered in such tables.
Of greater importance is the source of emetogenicity data
itself. Phase I studies of new agents are designed to identify
dose-limiting toxicities (grade 3 and grade 4 toxicities), and
lesser toxicities (such as grade 1 nausea/vomiting) may go
unreported, thereby underestimating the true incidence of
the problem. Prophylactic use of antiemetics may not be
restricted in a phase I study of a new cytotoxic chemotherapy, even if it is not known whether the agent is emetogenic.
This can lead to overestimation of the emetogenicity of a new
chemotherapy, where administration of a prophylactic antiemetic never was required, or to underestimation of the
emetogenicity of a new chemotherapy, where prophylactic
administration of antiemetic really was partially effective.
Even when the existence of the problem is recognized, the
magnitude of the problem may not be appreciated, particularly if pertinent events occur away from direct observation
by medical personnel. Advances in clinical practice have
moved administration of chemotherapy for most solid tumor
patients from the hospital to the clinic setting. Although
patients are directly observed on the day of treatment, the
following days (delayed period) are often spent at home. In
an observational study,6 we asked 24 physicians and nurses
from 14 oncology practices in the United States and Europe
to estimate the percentage of patients in their own practices
who were having chemotherapy-induced nausea or vomiting
after highly or moderately emetogenic chemotherapy. We
then surveyed 298 patients from these practices using 5-day
antiemetic diaries to determine the actual frequency of
From the Division of Hematology/Oncology, University of Vermont, Burlington, VT.

Address reprint requests to Steven M. Grunberg, MD, Division of Hematology/Oncology,
University of Vermont, 89 Beaumont Avenue - Given Building E214, Burlington, VT 05405;
email: steven.grunberg@uvm.edu
1092-9118/10/1-10
541
STEVEN M. GRUNBERG
acute and delayed nausea or vomiting after administration

of chemotherapy. Comparison of physician/nurse perception
to patient reality demonstrated that estimations of the
incidence of acute nausea and vomiting were quite accurate
but that the incidence of delayed nausea or vomiting was
markedly underestimated. In fact, the incidence of delayed
nausea or vomiting could be twice that estimated by experienced health professionals. Underestimation of nausea
was more extreme than underestimation of vomiting, and
the problem was particularly apparent in patients receiving
moderately emetogenic chemotherapy (often patients with
breast cancer receiving a combination of cyclophosphamide
and an anthracycline). Studies using similar methodology
have now been repeated in various parts of the world,
including Mexico,7 Venezuela,8 and Taiwan.9 Although absolute values may change (in Taiwan9 there was an overestimation of the incidence of acute vomiting while in
Venezuela8 delayed vomiting was a greater problem than
delayed nausea), the overall pattern of underestimation of
nausea and vomiting, particularly in the delayed setting
where patients are not being directly observed, remains. We
therefore urge oncology practices to adopt as a standard
operating procedure for administration of the first cycle of a
new chemotherapy either a single telephone call to the
patient or a single administration of a reporting tool such as
the Multinational Association of Supportive Care in Cancer
(MASCC) Antiemetic Tool (MAT) (which has been validated
for 3-day recall) several days after administration of chemotherapy10 to more accurately describe problems that should
be addressed during future chemotherapeutic cycles.
Even if the existence of greater amounts of nausea and
vomiting than expected is accepted intellectually as a problem for the overall population of patients receiving chemotherapy, failure to accept the problem in terms of ones own
patients may limit adoption of appropriate remedies. In a
very interesting study, Mertens and colleagues11 evaluated
different strategies to encourage greater compliance with
generally accepted antiemetic guidelines in a single hospital. Distribution of written educational materials concerning
the guidelines led to a transient increase in compliance to
these standards, but prescription habits soon returned to
KEY POINTS
542
The current incidence of chemotherapy-induced nausea and vomiting, particularly delayed nausea and
vomiting, is underestimated by physicians and
nurses.
The incidence of clinically significant toxicities such
as nausea and vomiting for new agents should be
better defined during the drug development process.
Real-time recording of patient-reported outcomes
through telephone contact or questionnaire is more
accurate than reports of toxicities recalled several
weeks after the event.
Physicians should be aware that patient efforts to
appear strong and cooperative may lead to underreporting of toxicities.
Nausea and vomiting are separate phenomena that
may require unique remedies.
baseline. A single Grand Rounds presentation concerning

antiemetics by a nationally recognized expert speaker had
no effect on antiemetic guideline compliance within the
hospital. The investigators then used diaries to survey
patients receiving chemotherapy concerning their emetic
experience and shared these results with the patients own
physicians. When the physicians were able to see persistent
nausea and vomiting not as abstract problems but as events
documented to be happening to their own patients, then a
durable increase in compliance with antiemetic guidelines
as reflected in antiemetic prescribing habits was achieved.
This is consistent with the observations by Stuebe12 concerning the value of Level I as compared to Level IV evidence
in clinical practice. Physicians learn intellectually from the
randomized double-blind clinical trials that are the basis of
Level I evidence. However, we remember and react to our
experiences with our own patients, and it is the Level IV
evidence of adverse anecdotes that is most likely to lead to
a durable improvement in practice. However, even if we are
willing to change practice based on the experience of our
patients, we must be able to understand what those experiences are, and questions of communication therefore become
important. Salsman and colleagues13 interviewed physicians and patients concerning barriers to effective communication regarding treatment of chemotherapy-induced
nausea and vomiting and found interesting similarities as
well as interesting differences. Both physicians and patients
indicated a desire to minimize the number of medications that were being taken to avoid drug-drug interactions
and side effects. This can be a positive goal as long as it
does not lead to underutilization of effective and necessary
remedies. A significant percentage of physicians and patients suggested that the presence of nausea and vomiting
is an indication that chemotherapy is working. This is
disturbing since no relationship between the intensity of
chemotherapy-induced nausea and vomiting and the effectiveness of chemotherapy has ever been demonstrated. Failure to use available effective antiemetics due to a mistaken
belief that such a course of action was preserving therapeutic efficacy would be a disservice to everyone involved. Of
particular concern was the finding that patients wanted to
be strong and not complain to their physicians, while
physicians believed that patients would report serious adverse events if they occurred. This disconnect in communication in and of itself could lead to undertreatment of
chemotherapy-induced nausea and vomiting even if potentially effective agents were available.
An additional concept that may lead to the development of
more effective antiemetic agents is the realization that
nausea and vomiting are not manifestations of the same
phenomenon but rather two separate phenomena.14 It has
long been assumed that nausea is simply the prodrome of
vomiting and that agents that relieve vomiting will therefore certainly relieve nausea as well. This belief is the
justification for the use of antiemetic clinical trial endpoints
such as complete response (no vomiting and no use of rescue
medication), which do not even mention the term nausea
as sufficient for regulatory approval of agents not just to
prevent vomiting but rather to prevent nausea and vomiting. In reality, vomiting is an objective endpoint (expulsion
of stomach contents), while nausea is a subjective endpoint
that is more difficult to localize or define. Because nausea is
INCIDENCE AND PREVALENCE OF CHEMOTHERAPY-INDUCED NAUSEA AND VOMITING
subjective, it is also much more difficult to identify animal

models that reliably display nausea and that can be quantified to a sufficient degree to screen new antinausea agents
before clinical testing. It is certainly true that nausea and
vomiting are correlated and that agents that are effective
against one of these phenomena are likely to have some
effect against the other. However, it must also be recognized
that neurotransmitter pathways that control vomiting and
nausea are probably not identical and that these differences
may provide guidance to develop new agents specifically
effective against chemotherapy-induced nausea. It is of
interest that several agents suggested to have significant
activity against chemotherapy-induced nausea (such as
corticosteroids,15 megestrol acetate,16 cannabinoids,17 and
olanzapine18) are not agents that have been highly effective
against chemotherapy-induced vomiting but rather agents

suggested to have activity against cancer anorexia and
cachexia. Perhaps the common ground between nausea and
anorexia will be more promising in leading to increased
understanding of this phenomenon than the common ground
between nausea and vomiting.
Although advances in antiemetic care have been based on
identification of relevant neurotransmitter/neurotransmitter receptor pathways and on carefully conducted clinical
trials,19 it is still necessary to understand the extent of the
problem and the nature of the problem before the problem
can be fully addressed. Issues of communication and reporting as well as continued efforts to more accurately define
and address the specific problem of nausea will therefore be
critical to further advances in this field.
Author
Steven M. Grunberg
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
A.P. Pharma;
Archimedes;
Helsinn; Merck;
Tesaro
Merck
Merck
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Goldberg N. Passover Haggadah. Hoboken NJ: Ktav Publishing House;
1993.
2. Sequist LV, Bell DW, Lynch TJ, et al. Molecular predictors of response
to epidermal growth factor receptor antagonists in non-small-cell lung cancer.
J Clin Oncol. 2007;25:587-595.
5. Grunberg SM, Warr D, Gralla RJ, et al. Evaluation of new antiemetic
agents and definition of antineoplastic agent emetogenicitystate of the art.
Support Care Cancer. 2011;19 (suppl 1):S43-S47.
6. Grunberg SM, Deuson RR, Mavros P, et al. Incidence of chemotherapyinduced nausea and emesis after modern antiemetics. Cancer. 2004;100:22612268.
7. Erazo Valle A, Wisniewski T, Figueroa Vadillo JI, et al. Incidence of
chemotherapy-induced nausea and vomiting in Mexico: healthcare provider
predictions versus observed. Curr Med Res Opin. 2006;22:2403-2410.
8. De Jongh-Garcia C, Poli S, Ananth C, et al. Health care provider
perception of nausea and vomiting and patients reported incidence: the
Venezuela Emesis Registry. Support Care Cancer. 2005;13:414-415 (abstr).
9. Liau CT, Chu NM, Liu HE, et al. Incidence of chemotherapy-induced
nausea and vomiting in Taiwan: physicians and nurses estimation vs
patients reported outcomes. Support Care Cancer. 2005;13:277-286.
10. Molassiotis A, Coventry PA, Stricker CT, et al. Validation and psychometric assessment of a short clinical scale to measure chemotherapy-induced
nausea and vomiting: the MASCC antiemesis tool. J Pain Symptom Manage.
2007;34:148-159.
11. Mertens WC, Higby DJ, Brown D, et al. Improving the care of patients
with regard to chemotherapy-induced nausea and emesis: the effect of
feedback to clinicians on adherence to antiemetic prescribing guidelines.
J Clin Oncol. 2003;21:1373-1378.
12. Stuebe AM. Level IV evidenceadverse anecdote and clinical practice.
N Engl J Med. 2011;365:8-9.
13. Salsman JM, Grunberg SM, Beaumont JL, et al. Communicating about
chemotherapy-induced nausea and vomiting: a comparison of patient and
provider perspectives. J Natl Compr Canc Netw. 2012;10:149-157.
14. Molassiotis A, Stricker CT, Eaby B, et al. Understanding the concept of
chemotherapy-related nausea: The patient experience. Eur J Cancer Care.
2008;17:444-453.
15. Parry H, Martin K. Single-dose i.v. dexamethasonean effective antiemetic in cancer chemotherapy. Cancer Chemother Pharmacol. 1991;28:231232.
16. Loprinzi C, Jatoi A. Antiemetic properties of megestrol acetate. J
Palliat Med. 2006;9:239-240.
17. Storr MA, Sharkey KA. The endocannabinoid system and gut-brain
signalling. Curr Opin Pharmacol. 2007;7:575-582.
18. Navari RM, Brenner MC. Treatment of cancer-related anorexia with
olanzapine and megestrol acetate: a randomized trial. Support Care Cancer.
2010;18:951-956.
19. Hesketh PJ. Chemotherapy-induced nausea and vomiting. N Engl
J Med. 2008;358:2482-2494.
543
MUCOSAL INJURY IN PATIENTS WITH CANCER:

TARGETING THE BIOLOGY
CHAIR
Douglas E. Peterson, DMD, PhD
University of Connecticut Health Center
Farmington, CT
SPEAKERS
Dorothy M. Keefe, MD, MBBS
Royal Adelaide Hospital
Adelaide, Australia
Stephen T. Sonis, DMD, DMSc
Biomodels
Watertown, MA
New Frontiers in Mucositis

By Douglas E. Peterson, DMD, PhD, Dorothy M. Keefe, MD, and
Stephen T. Sonis, DMD, DMSc
Overview: Mucositis is among the most debilitating side

effects of radiotherapy, chemotherapy, and targeted anticancer therapy. Research continues to escalate regarding key
issues such as etiopathology, incidence and severity across
different mucosae, relationships between mucosal and nonmucosal toxicities, and risk factors. This approach is being
translated into enhanced management strategies. Recent
technology advances provide an important foundation for this
continuum. For example, evolution of applied genomics is
fostering development of new algorithms to rapidly screen
genomewide single-nucleotide polymorphisms (SNPs) for
patient-associated risk prediction. This modeling will permit
individual tailoring of the most effective, least toxic treatment
in the future. The evolution of novel cancer therapeutics is
changing the mucositis toxicity profile. These agents can be
associated with unique mechanisms of mucosal damage.
Additional research is needed to optimally manage toxicity
caused by agents such as mammalian target of rapamycin

(mTOR) inhibitors and tyrosine kinase inhibitors, without reducing antitumor effect. There has similarly been heightened
attention across the health professions regarding clinical
practice guidelines for mucositis management in the years
following the first published guidelines in 2004. New opportunities exist to more effectively interface this collective guideline portfolio by capitalizing upon novel technologies such as
an Internet-based Wiki platform. Substantive progress thus
continues across many domains associated with mucosal
injury in oncology patients. In addition to enhancing oncology
patient care, these advances are being integrated into highimpact educational and scientific venues including the National Cancer Institute Physician Data Query (PDQ) portfolio as
well as a new Gordon Research Conference on mucosal health
and disease scheduled for June 2013.
research over the past 15 years.10 This has in turn fostered

development of clinical practice guidelines. These have been
disseminated widely via publications and presentations at
scientific conferences. In addition, clinical management interventions and their biologic basis have been incorporated
into electronic technologies such as the National Cancer
Institute PDQ website Oral Complications of Chemotherapy and Head/Neck Radiation.11 A comprehensive update
of this website, including oral mucositis pathobiology, prevention, and treatment, was posted in March 2011. This
material was the fourth of 35 most frequently viewed Supportive Care PDQ websites in the following 3 months (April
to June 2011), with continued high activity among health
professionals and patients into 2012.
UCOSITIS is an umbrella term covering cancer

treatment regimen-related damage to mucosal
surfaces. Although mucositis has been clinically observed for
several decades, comprehensive advances in pathobiology,
clinical impact, and molecularly targeted treatment have
chiefly occurred the past 15 years. As a result mucositis
was incorporated as a medical subject headings (MeSH)
term by the National Library of Medicine in 2006.
Although targeted anticancer therapies can cause mucosal damage, the actual scope of injury is not identical to that
caused by conventional cancer therapies. While mucositis
is most often considered in relation to the oral cavity and
digestive tract, other mucosae such as respiratory and
genitourinary tract can be affected. The degree of damage is
not uniform, with oral mucositis typically most prominent
with systemic cancer treatments. Conjunctivae and vaginal
mucosa appear uniquely resistant to damage for reasons
that are not well delineated.
Incidence and severity of mucositis is related to the type of
treatment administered, the risk factors of the host, and the
tumor being treated.1,2 Chemotherapy can cause mucositis
even in a limited number of doses if the agent is sufficiently
mucotoxic, whereas radiation tends to cause local toxicity,
only leading to distant toxicities if a sufficiently high dose is
used.2 The various targeted agents cause mucositis depending on cross-reactivity of receptors in the mucosa concerned,
or presence of their known target in mucosa.3
Standard-dose chemotherapy causes oral mucositis in up
to 40% of patients, whereas high-dose chemotherapy related
to stem cell/bone marrow transplantation causes oral mucositis in up to 80% of patients.4,5 Radiotherapy for head
and neck cancer causes oral mucositis in virtually 100% of
patients.6 Oral mucositis incidence secondary to mTOR
inhibitors is approximately 50%, and incidence of diarrhea
from targeted agents is approximately 40%, albeit of low
grade.7,8 Chemotherapy causes diarrhea in up to 40% of
patients as well; it is also typically low-grade but can be
more frequent and severe in relation to irinotecan.9
Given the clinical and economic importance of the toxicity,
there has been an escalating trajectory of basic and clinical
Effect on Quality of Life (QoL)
Mucositis is but one of the toxicities typically associated

with cancer treatment. It rarely occurs alone; study of links
between toxicities provides information on shared pathobiology and opportunities for intervention in multiple toxicities.12,13 The importance of these toxicities lies in the effect
not only on patients QoL, but also on their ability to
withstand effective doses of treatment.14 Dose reductions or
delays lead to reduction in efficacy and survival. Most
clinical trials of cancer treatment do not have toxicity as a
primary endpoint, making evaluation of true incidence difficult.15 Health professionals estimates of incidence and
severity of toxicity may be lower than those reported by
patients themselves.16
From the Department of Oral Health and Diagnostic Sciences, School of Dental Medicine,
Neag Comprehensive Cancer Center, University of Connecticut Health Center, Farmington,
CT; University of Adelaide, Adelaide, Australia; Harvard School of Dental Medicine,
Brigham and Womens Hospital and the Dana-Farber Cancer Institute, and Biomodels,
LLC, Boston, MA.
Address reprint requests to Douglas E. Peterson, DMD, PhD, Department of Oral Health
and Diagnostic Sciences, School of Dental Medicine, University of Connecticut Health
Center, 263 Farmington Ave., Farmington, CT 06030-1605; email: peterson@nso.uchc.edu
1092-9118/10/110
545
PETERSON, KEEFE, AND SONIS
Painful mouth ulcers and gastrointestinal toxicity including diarrhea can significantly compromise QoL (p 0.01).17
For example, a prospective, multicenter study of the burden
of mucosal toxicity on patients receiving chemotherapy and
radiotherapy for various cancers demonstrated that for each
increasing grade of oral mucositis or diarrhea, there is a
corresponding significant reduction in QoL.18 Given the
importance of these and related toxicities there is important
opportunity in the era of the Internet and tablet computers
to enhance collection and analysis of toxicity data. This
strategy would enable clinicians to better understand toxicity being experienced by patients and to intervene more
quickly.19
New Frontier 1: Development of Molecularly Targeted

Mucositis Therapeutics
There are multiple agents at varying stages of development for the prevention or treatment of mucosal injury
(Table 1). The collective strategic feature involves targeting
of drug mechanism in relation to the molecular model for
mucositis. The drugs cover the spectrum from topical rinses
to injectable agents, and are used for radiotherapy, chemotherapy, chemoradiation, and more recently, cancertargeted therapy. The majority are in trial for radiationinduced mucosal injury, in large part due to the clinical
severity as well as feasibility in the clinical setting.
New Frontier 2: Genomics
Symptoms and Signs of Mucositis
Symptoms and signs of mucositis depend on the mucosa

affected. Oral mucositis results in pain, erythema, and
ulceration, and can lead to infection, malnutrition, and
dehydration; it can also lead to death of selected patients.20
Esophagitis has comparable features.21 Gastritis can be
accompanied by bleeding.22 Small and large bowel mucositis
is accompanied by abdominal pain, bloating, and diarrhea;
pelvic mucositis can result in bleeding as well as tenesmus.23
Typhlitis, a specific manifestation of mucositis of the
caecum, causes pain, abdominal bloating, and altered bowel
habitand can present as an acute abdomen in the setting
of neutropenia and fever.24 It can be diagnosed by abdominal imaging, but treatment is typically conservative as
patients are usually too ill to undergo surgery.24
New Frontiers
Exciting new frontiers are ahead in relation to mucositis.

Selected examples include (1) development of molecularly
targeted mucositis therapeutics, (2) genomics, (3) new types
of cancer therapies, (4) studies involving nonalimentary
mucosa, and (5) new strategies for mucositis guideline
production and utilization.
KEY POINTS
546
Mucositis is a major toxicity of cancer treatment and

affects quality of life, cancer outcome, and cost of
care.
Mucositis rarely occurs in isolation; linkage of research relative to mechanisms and treatments of
other toxicities is thus strategically important.
Targeted anticancer therapies cause mucositis by
new mechanisms and provide a fertile area for ongoing research.
Applied genomics allows genome-wide risk predictiontool development with translation to personalized cancer medicine.
Ongoing systematic evaluation of the mucositis literature, development and dissemination of guidelines,
and impact assessment of the interventions are
key to optimizing clinical and health care economic
outcomes.
Concepts that define underlying mucositis pathobiology

are rapidly evolving. A reductionist view had existed into the
mid-1990s; namely, regimen-related epithelial damage was
solely attributable to clonogenic cell death of progenitor
stem cells in the basal layer or villi. Now it is clear that
mucotoxicity represents the culmination of a series of biologically complex events that are comprehensively dependent on the tissue and its environment.6,25 Evidence exists
to support a role for at least 14 different canonical pathways
that, once activated by radiation or chemotherapy, contribute to mucosal damage (Fig. 1).26 In addition and unlike
internal organs, every mucosal surface is exposed to a
unique external environment (i.e., gut flora or saliva), which
in turn potentially modifies tissue response.
The biologic complexity of mucosal injury has provided
opportunities for application of genomics to at least three
key areas: (1) understanding the biologic basis for the
condition, (2) predicting toxicity risk among various populations, and (3) identifying responder and nonresponder populations to mucositis drug therapy.27 Of note is that this
modeling is also consonant with one of the key clinical
cancer advances in 2011 highlighted by Dr. J. Von Roenn,
namely, study of a genetic biomarker that predicts taxaneinduced neuropathy.28
Differential changes in gene expression following radiation or chemotherapy were first comprehensively described
in animal models.29 Sequential differences in levels and
actions of expressed genes provided insights into mechanistic changes that formed the biologic basis for etiopathology
of mucositis. These early studies utilized RNA extracted
from targeted tissue in animals, since sequential, frequent
biopsies from oncology patients was and is not generally
feasible. To overcome this barrier to human studies, peripheral blood was studied as an RNA source in patients treated
for solid tumors.26 This provided opportunity for study of
changes in expression over time and correlation with clinical
and surrogate (i.e., cytokine) endpoints. Furthermore,
rather than analyzing individual genes, generation of
Bayesian networks and their comparison with established
canonical pathways and mechanistic groupings yielded insight into pathogenesis and potential drug targets for intervention.
Despite these advances, questions persist. For example,
why do some patients develop mucositis and others do not?
Why is it that two individuals with comparable demographics, tumor classifications, and treatment regimens have such
disparate toxicity?
Partial insights to these challenging questions do exist.
Age, gender, body mass, and other patient characteristics
MUCOSAL INJURY CAUSED BY CANCER THERAPIES

Table 1. Selected Agents in Development for the Prevention/Treatment of Mucositis a
Stage of Trial Versus
Cancer Treatment
Chemotherapy
Radiotherapy
Targeted Cancer
Therapy
Preclinical
AMP-18/NX002
PMX-30063
TXA127
TZP-201
AMP-18/NX002
CBLB502
OralX
Transcutaneous electrical nerve stimulation
Phase I
Elsiglutide
Glucarpidase
Oral selenium
Dexlansoprazole
Glucarpidase
SGX201
Phase II
AG013
Buprenorphine
GelClair
Lactobacillus CD 2 lozenges
LED therapy
Omegaven
Oral impact
rhEGF
rhGM-CSF
Sargramostim (GM-CSF)
SCV-07
Selenomethionine
ALD518
Camomilla recutita mouthwash
Clonidine lauriad
Dexpanthenol mouthwash
GelClair
Honey mouthwash
IZN-6N4 (botanical extracts)
LED therapy
L-lysine
rhGM-CSF
Sargramostim (GM-CSF)
SCV-07
Selenomethionine
Doxycycline hyclate
Phase III
Amifostine trihydrate
Caphosol
Celecoxib
Fosaprepitant
Glutamine
Low-level laser light therapy
Palifermin
Amifostine trihydrate
Caphosol
Celecoxib
Doxepin hydrochloride rinse
Fosaprepitant
Humidification
Hyperbaric oxygen
Hyperimmune colostrum
Iseganan hydrochloride
Palifermin
Pilocarpine
rhEGF
Caphosol
Postmarketing
Glutamine popsicles
Impact enteral nutrition
Lenograstim
Morphine
MuGard
Palifermin
Impact enteral nutrition

Palifermin
Epigallocatechin gallate
Abbreviation: LED, light-emitting diode.

a
Source: ClinicalTrials.gov and industry websites.
have emerged as poor predictors of risk. In contrast, and

based on the fact that virtually every mechanistically
important pathway in human biology is genetically controlled, differences in gene expression are key risk determinants. For example, oxidative stress and proinflammatory
cytokines have been implicated in development of mucositis.30,31 Sharp increases in reactive oxygen species are noted
almost immediately after exposure to drug or radiation with
a resultant cascade of events that cause tissue destruction.32
Based on this observation, studies directed at genes controlling the metabolism of reactive oxygen species have identified deletion-specific nucleotide polymorphisms that, when
present in a patient, are associated with increased risk of
mucositis induced by both chemotherapy33 and radiation.34
In addition, both systemic and tissue-associated increases
of proinflammatory cytokines (i.e., TNF-) are associated
with regimen-related mucosal injury.30,31 Genetically controlled variability in pro-inflammatory cytokine production
and its association with disease status has been well established. For example, the presence of SNPs associated with
TNF- has a significant effect on severe toxicity risk (p

0.005), including mucositis, in an allogeneic stem cell transplant population.35
Despite the importance of these lines of research, however, the strategy of identifying the gene or SNP that
predicts toxicity is limited by the presumption that a single
genetic entity is controlling overall toxicity expression. Further, although genetically controlled deficits in metabolic
enzymes have been consistently associated with toxicity
risk, the percentage of patients manifesting these genetic
alterations is minute compared with the numbers of patients
who develop mucositis.
Perhaps a more comprehensive assessment of genetically
based risk could thus be biologically valid. An alternative
approach has thus been recently validated.36 The methodology is based on a concept that recognizes that a phenotype is
at least as likely to be the consequence of a team of genes
working cooperatively as it is due to the expression of a
single master gene. The analysis employs Bayesian networks derived from unsupervised and learned networks of
547
Fig 1. Bayesian network demonstrating overexpressed genes defined by signaling functions. Major roles for TNF- have been suggested in
pathoetiology of radiation-induced toxicities, including mucositis and fatigue. In addition to its potential primary role as a mediator of tissue
injury, the importance of TNF- as a major signaling conduit is supported by its central position in this pathway. Its prominence in more than one
pathway further reinforces its potential importance in radiation-induced toxicities.26
genes or SNPs. How does this differ from the conventional

approach? First, it is not hypothesis driven. Although this
may sound contrary to conventional wisdom, it actually
provides substantial new potential modeling as we are not
limited by historic knowledge. The unsupervised approach
assumes that we do not know what we do not know.
Consequently and second, there are no threshold expression
values required for a gene or SNP to be included in the
analysis. Third, using aggressive computer programs, algorithms test and re-evaluate the predictive value of groups of
genes or SNPs to ultimately arrive at the cluster of the
greatest predictive power. The defined cluster can then be
validated prospectively.
A totally nonclinical example of this modeling is illustrative. Lets say you are the new owner of the Chicago ASCOs,
a professional baseball team with a very limited budget. You
thus decide to take a novel approach to select the players
for your team. You have read Moneyball,37 so rather than
looking for individuals with excellent batting percentages
548
you recruit individuals with mediocre individual statistics.

These players, however, have contributed to winning teams
by adding synergies that collectively affect team performance. You eliminate any requirements for people who want
to play for your team; 1,000 individuals report for tryouts. To
determine the best team, your selection mechanism consists
of taking nine random players at a time and having them
play three innings with a team of former major leaguers.
You observe the interaction and production of the first group
of nine candidates. Two seem to work well together and are
effective at getting on base (walks are as good as hits as well
as avoiding making outs), so you keep them on the field and
put the other seven back into the pool. You next randomly
pick seven candidates and repeat the process. You repeat
this process over and over until you have not only evaluated
each candidate, but you have also evaluated every combination permutation. (You may not have much money, but you
have all the time in the world.) At the end of the process, you
produce a team that, taken together, is optimal.

Table 2. Examples of Mucositis Guidelines
Organization (Alphabetical Order)
ASCO
ESMO
MASCC/ISOO
NCCN
ONS
RTOG
Atlantic Provinces Pediatric
Hematology Oncology Network
Meta-analysis: Cochrane review
(prevention)
Meta-analysis: Cochrane review
(treatment)
URL
http://jco.ascopubs.org/content/27/1/127.full
http://annonc.oxfordjournals.org/content/22/suppl_6/vi78.full
http://www.mascc.org/mc/page.do?sitePageId88037
http://www.nccn.org/JNCCN/PDF/mucositis_2008.pdf
http://www.ons.org/Research/PEP/Mucositis
http://www.onlinecancereducationforum.com/OCEF/Oral%20mucositis%20in%20head%20and%20neck%20cancer.pdf
http://www.apphon-rohppa.com/en/guidelines/mucositis-guidelines
http://summaries.cochrane.org/CD000978/interventions-for-preventing-oral-mucositis-for-patients-with-cancer-receiving-treatment
http://summaries.cochrane.org/CD001973/interventions-for-treating-oral-mucositis-for-patients-with-cancer-receiving-treatment
Abbreviations: ASCO, American Society of Clinical Oncology; ESMO, European Society for Medical Oncology; MASCC/ISOO, Mucositis Study Group of Multinational
Association for Supportive Care in Cancer/International Society of Oral Oncology; NCCN, National Comprehensive Cancer Network; ONS, Oncology Nursing Society;
RTOG, Radiation Therapy Oncology Group.
The SNP algorithm follows a similar pattern with a couple

of caveats. First, the number of SNPs far exceeds 1,000, so
developing the team requires billions of computergenerated calculations. Second, we have the ability to compare predictive accuracy against a clinical observation
(almost like being able to see how each performed in the real
world). Third, we are not limited to only nine players being
on the team. Understanding that nongenetically controlled
factors (e.g., epigenetic pathways, concomitant disease) also
have potential to influence risk imposes an additional challenge, but the analytic algorithms have the capacity to
overlay and integrate these elements into a risk-prediction
model.
This approach has been recently applied to the practical
clinical problem of predicting oral mucositis risk for patients
with hematologic malignancies receiving standard conditioning regimens in preparation for hematopoietic stem cell
transplant (HSCT). These treatments typically result in
severe mucositis in almost 40% of treated patients.4 Records
of approximately 500 patients who had received HSCT were
reviewed, and we identified groups of patients who had
developed significant mucositis and patients who had no
stomatotoxicity. With DNA extracted from the saliva of each
patient we ran genome-wide SNP arrays. Using a Bayesian
network framework in computer algorithms that were similar to the recruitment of the Chicago ASCOs, we were able
to identify a subset of SNPs that predicted mucositis (unpublished data; ST Sonis).
Applied genomics also provides a platform for the differentiation of patients who respond to a particular drug from
those who do not. Using RNA obtained from subjects before
and after their participation in a clinical trial, a cluster of
genes was derived that was associated with drug response
as measured clinically by its ability to modify the course of
chemoradiation-associated oral mucositis. Knowing this information in a phase II study could be transformative to
subsequent study design. Inclusion criteria could be specifically adopted to reflect known responder genomics, resulting in the need for fewer subjects to power the study such
that the trial was efficient, economical, and optimized for a
successful outcome. This application of personalized medicine to determination of risk and selection of effective
therapies for mucosal toxicities is no longer hypothetical. Its
continued development represents a new frontier that will
continue to expand as new treatments become available.38
New Frontier 3: New Types of Cancer Therapy
A number of targeted anticancer therapies are now in

regular use or on the cusp of being clinically available. Still
more are in development. With them come a new battery of
mucosal toxicities, some unique to particular agents and
others synergistic with other forms of conventional therapy.27 The mucosa has not escaped as a tissue at risk and, in
fact is actually a common site for toxicities.3
Toxicity data thus continue to accumulate at a rapid rate.
In general, the mechanism underlying the pathogenesis of
mucosal injury has not been well studied, although it likely
occurs through different pathways than those described for
cytotoxic drugs or radiation. Consequently, clinical course
may vary with respect to onset, duration, cycle-associated
risk, and clinical presentation. Diarrhea is common, as is
exacerbation of mouth ulcers.
The mTOR inhibitors illustrate the unique presentation
of targeted therapies. Of patients receiving this drug class,
approximately 40% will develop severe oral mucosal lesions.39 These differ from conventional mucositis in time to
onset, duration, cycle-related dependence, and clinical
presentation.39-41 Interestingly, concomitant toxicities typically seen with oral mucositis (other gastrointestinal toxicity) are uncommon, whereas rash is more frequent.39-42
These differences are likely attributable to variations in
biologic pathways that contrast between the two types of
treatment.
As targeted therapies continue to evolve, more aggressive
investigation to understand their cellular etiologies will be
critical to develop effective interventions.
New Frontier 4: Studies Involving Nonalimentary Tract Mucosa
The mucosal surfaces share similarities of development

and structure, but have important differences relating to
their individual functions. Normal tissue has a limited
number of ways in which it can respond to insult, including
response to cancer therapy. The cancer community can learn
from other experts in mucosal health and disease such as
gastroenterologists,43 and it is also likely that members of
the oncology community can provide key insights to the
modeling of nonalimentary tract mucosal lesions. For example, and as noted previously, regimen-related mucosal injury
does not occur with equal frequency across other types of
mucosa. Why is it that patients receiving levels of chemo-
549
therapy that are mucotoxic to the gastrointestinal tract do

not develop injury of the conjunctiva or vaginal mucosa?
What characteristics of these tissues differentiate them from
other forms of mucosal stratified squamous epithelium to
reduce susceptibility to toxicity? Clearly the external environment is different, although both have a defined microflora and a secretory component as noted previously.
To further advance this research agenda, a new Gordon
Research Conference44 will take place in June 2013. It is
designed to foster new and impactful research directions
regarding mucosal health and disease. Mucosal injury in
patients with cancer will be highlighted in this context.
New Frontier 5: New Strategies for Mucositis Guideline Production
and Utilization (A Proposed Opportunity for ASCO)
The evolution of guidelines for mucositis management is

even more recent in origin than is the creation of the
contemporary pathobiologic model conceptualized in the
1998.45 Beginning in 2001, the Mucositis Study Group of
Multinational Association for Supportive Care in Cancer/
International Society of Oral Oncology (MASCC/ISOO) conducted an evidence-based review of the mucositis literature,
incorporating ASCO-level scoring criteria. The original version of the guidelines was published in 200446 and the
revised version in 2007.47 The modeling has been subsequently developed by several other health professional
groups such that there is currently a relatively robust
portfolio that is largely complementary (Table 2).
However, variability can occur in guideline production
for several reasons including scope of literature review as
well as methodology utilized to evaluate the quality of the
literature. However, perhaps the ultimate challenge is how
best to navigate the multiple challenges of incorporating
guideline-based interventions into clinical practice. As with
all guidelines, these challenges include the following:
deciding on strategies and then implementing the modeling to integrate the collective guidelines into a
user-friendly, efficient series of suggestions and recommendations for clinicians;
customizing the guidelines for country- and/or regionspecific practices, including language translations and
access to drug and device technology;
assessing health and economic outcomes in association
with guideline use.
As noted in Table 2, many academic societies produce

guidelines for the clinical management of toxicity. For example, MASCC/ISOO produces guidelines for alimentary
tract mucositis, the Cochrane collaboration produces a metaanalysis review of oral mucositis,48,49 and ASCO includes
oral mucositis in its oral health guidelines. The European
Society for Medical Oncology (ESMO) utilizes a system
based on the MASCC/ISOO guidelines, and NCCN publishes
recommendations as well. Similar guidelines for other toxicities such as nausea/vomiting or neutropenia are also
produced by these socieities, although there has been one
antiemetic guideline published jointly by MASCC/ASCO/
ESMO. Unfortunately, however, the latest round was once
again separate.
It would seem important and logical for a single combined
effort to review the literature regarding each toxicity, and
move to a more modern methodology such as use of a Wiki
platform.50 Management of clinical practice guidelines via
such technology could ultimately improve the standard and
consistency of clinical practice according to the best and
most recent scientific evidence available. Unlike written
guidelines, Wiki guidelines are continually updated as new
evidence becomes available; they are also linked to source
abstracts and other evidence-based sites that add to the
value of this technology.
A Proposal for ASCO
It thus appears to be an unprecedented time in science

and clinical practice to capitalize on the multiple important
opportunities ahead. The collective broad and deep expertise
of the ASCO organization represents an ideal venue to
further delineate, address, and overcome these clinical barriers. We thus respectfully propose the following strategy for
consideration by ASCO membership.
It is proposed that ASCOs Patient and Survivor Care
Committee, in partnership with MASCC/ISOO, assume a
leadership role in advancing new opportunities for mucositis
guideline development, dissemination, and measurement of
impact on clinical and economic outcomes in patients with
cancer.
This model could facilitate discussion across the multiple
professional organizations that have produced mucositis
guidelines over the past decade. Creating a coordinated
approach beginning in 2012 could strategically enhance care
for patients with cancer worldwide in the years to come.
Author
Douglas E. Peterson
Employment or
Leadership
Positions
Dorothy M. Keefe
Stephen T. Sonis
550
Biomodels (L)
Consultant or
Advisory Role
Alder; Amgen;
Merck
Helsinn
ActoGeniX;
Axaxia
Biologicals;
Galera
Therapeutics (U);
Inform
Genomics; Izun
Pharmaceuticals;
Merck; Novartis;
Pfizer; Polymedix
(U); ProCertus;
SciClone
Stock
Ownership
Honoraria
Research
Funding
Merck KGaA;
Pfizer
GlaxoSmithKline;
Nestec
Expert
Testimony
Other
Remuneration
REFERENCES
1. Logan RM, Stringer AM, Bowen JM, et al. Is the pathobiology of
chemotherapy-induced alimentary tract mucositis influenced by the type of
mucotoxic drug administered? Cancer Chemother Pharmacol. 2009;63:239251.
2. Raber-Durlacher JE, Elad S, Barasch A. Oral mucositis. Oral Oncol.
2010;46:452-456.
3. Keefe DM, Gibson RJ. Mucosal injury from targeted anti-cancer therapy.
Supportive Care Cancer. 2007;15:483-490.
4. Naidu MU, Ramana GV, Rani PU, et al. Chemotherapy-induced and/or
radiation therapy-induced oral mucositis-complicating the treatment of cancer. Neoplasia. 2004;6:423-431.
5. Peterson DE, Bensadoun RJ, Roila F. Management of oral and gastrointestinal mucositis: ESMO Clinical Practice Guidelines. Ann Oncol. 2011;22
(suppl 6):vi78-84.
6. Sonis ST. Oral mucositis. Anti-Cancer Drugs. 2011;22:607-612.
7. Crown JP, Burris HA, 3rd, Boyle F, et al. Pooled analysis of diarrhea
events in patients with cancer treated with lapatinib. Breast Cancer Res
Treat. 2008;112:317-325.
8. Porta C, Paglino C, Imarisio I, et al. Safety and treatment patterns of
multikinase inhibitors in patients with metastatic renal cell carcinoma at a
tertiary oncology center in Italy. BMC Cancer. 2011;11:105. Epub 2011 March
24.
9. Stein A, Voigt W, Jordan K. Chemotherapy-induced diarrhea. Pathophysiology, frequency and guideline-based management. Therap Adv Med
Oncol. 2010;2:51-63.
10. Peterson DE, Lalla RV. Oral mucositis: The new paradigms. Current
Opin Oncol. 2010;22:318-322.
11. Oral complications of chemotherapy and head/neck radiation. NCI PDQ.
2012. www.cancer.gov/cancertopics/pdq/supportivecare/oralcomplications/
HealthProfessional. Accessed February 10, 2012.
12. Aprile G, Ramoni M, Keefe D, et al. Links between regimen-related
toxicities in patients being treated for colorectal cancer. Current Opin Support
Palliat Care. 2009;3:50-54.
13. Nishimura N, Nakano K, Ueda K, et al. Prospective evaluation of
incidence and severity of oral mucositis induced by conventional chemotherapy in solid tumors and malignant lymphomas. Support Care Cancer. Epub
2011 Nov 25.
14. Lalla RV, Sonis ST, Peterson DE. Management of oral mucositis in
patients who have cancer. Dental Clin N Am. 2008;52:61-77.
15. Saad ED. Endpoints in advanced breast cancer: methodological aspects
& clinical implications. Indian J Med Res. 2011;134:413-418.
16. Bateman E, Keefe D. Patient-reported outcomes in supportive care.
Semin Oncol. 2011;38:358-361.
17. Cheng KK, Leung SF, Liang RH, et al. Severe oral mucositis associated
with cancer therapy: impact on oral functional status and quality of life.
Support Care Cancer. 2010;18:1477-1485.
18. Elting LS, Keefe DM, Sonis ST, et al. Patient-reported measurements
of oral mucositis in head and neck cancer patients treated with radiotherapy
with or without chemotherapy: demonstration of increased frequency, severity, resistance to palliation, and impact on quality of life. Cancer. 2008;113:
2704-2713.
19. Abernethy AP, Ahmad A, Zafar SY, et al. Electronic patient-reported
data capture as a foundation of rapid learning cancer care. Med Care.
2010;48:S32-8.
20. Barasch A, Epstein JB. Management of cancer therapy-induced oral
mucositis. Dermatol Ther. 2011;24:424-431.
21. Rodriguez ML, Martin MM, Padellano LC, et al. Gastrointestinal
toxicity associated to radiation therapy. Clin Transl Oncol. 2010;12:554-561.
22. Patterson WK, Sage RE, Keefe DM. Haemorrhagic gastritis in two
patients treated with all-trans-retinoic acid in acute promyelocytic leukaemia. Austral New Zealand J Med. 1994;24:314-315.
23. Gibson RJ, Keefe DM. Cancer chemotherapy-induced diarrhoea and
constipation: mechanisms of damage and prevention strategies. Support Care
Cancer. 2006;14:890-900.
24. Shvartsbeyn M, Edelman MJ. Pemetrexed-induced typhlitis in nonsmall cell lung cancer. J Thor Oncol. 2008;3:1188-1190.
25. Bowen JM, Keefe DM. New pathways for alimentary mucositis. J Oncol. 2008; 907892. Epub 2008 Sept 23.
26. Sonis S, Haddad R, Posner M, et al. Gene expression changes in
peripheral blood cells provide insight into the biological mechanisms associated with regimen-related toxicities in patients being treated for head and
neck cancers. Oral Oncol. 2007;43:289-300.
27. Keefe DM, Bateman EH. Tumor control versus adverse events with
targeted anticancer therapies. Nature Rev Clin Oncol. 2011;9:98-109.
28. Von Roenn J. Patient and survivor care. Clinical Cancer Advances
2011: ASCOs Annual Report on Progress Against Cancer. 2011;34.
29. Sonis ST, Scherer J, Phelan S, et al. The gene expression sequence of
radiated mucosa in an animal mucositis model. Cell Prolif. 2002;35 (suppl
1):93-102.
30. Logan RM, Stringer AM, Bowen JM, et al. Serum levels of NFkappaB
and pro-inflammatory cytokines following administration of mucotoxic drugs.
Cancer Biol Ther. 2008;7:1139-1145.
31. Ong ZY, Gibson RJ, Bowen JM, et al. Pro-inflammatory cytokines play
a key role in the development of radiotherapy-induced gastrointestinal
mucositis. Rad Oncol. 2010;5:22. Epub 2010 March 16.
32. Murphy CK, Fey EG, Watkins BA, et al. Efficacy of superoxide
dismutase mimetic M40403 in attenuating radiation-induced oral mucositis
in hamsters. Clin Cancer Res. 2008;14:4292-297.
33. Hahn T, Zhelnova E, Sucheston L, et al. A deletion polymorphism in
glutathione-S-transferase mu (GSTM1) and/or theta (GSTT1) is associated
with an increased risk of toxicity after autologous blood and marrow transplantation. Biol Blood Marrow Transpl. 2010;16:801-808.
34. Pratesi N, Mangoni M, Mancini I, et al. Association between single
nucleotide polymorphisms in the XRCC1 and RAD51 genes and clinical
radiosensitivity in head and neck cancer. Radiother Oncol. 2011;99:356-361.
35. Bogunia-Kubik K, Mazur G, Urbanowicz I, et al. Lack of association
between the TNF-alpha promoter gene polymorphism and susceptibility to
B-cell chronic lymphocytic leukaemia. Int J Immunogenet. 2006;33:21-24.
36. Sebastiani P, Ramoni MF, Nolan V, et al. Genetic dissection and
prognostic modeling of overt stroke in sickle cell anemia. Nat Genet. 2005;37:
435-440.
37. Lewis M. Moneyball: The Art of Winning an Unfair Game. New York:
WW Norton and Co; 2003.
38. Sharma R, Tobin P, Clarke SJ. Management of chemotherapy-induced
nausea, vomiting, oral mucositis, and diarrhoea. Lancet Oncol. 2005;6:93-102.
39. Sonis S, Treister N, Chawla S, et al. Preliminary characterization of
oral lesions associated with inhibitors of mammalian target of rapamycin in
cancer patients. Cancer. 2010;116:210-215.
40. Campistol JM, de Fijter JW, Flechner SM, et al. mTOR inhibitorassociated dermatologic and mucosal problems. Clin Transpl. 2010;24:149-456.
41. De Masson A, Fouchard N, Mery-Bossard L, et al. Cutaneous and
mucosal aphthosis during temsirolimus therapy for advanced renal cell
carcinoma: review of cutaneous and mucosal side effects of mTOR inhibitors.
Dermatol. 2011;223:4-8.
42. Gomez-Fernandez C, Garden BC, Wu S, et al. The risk of skin rash and
stomatitis with the mammalian target of rapamycin inhibitor temsirolimus:
a systematic review of the literature and meta-analysis. Eur J Cancer.
2012;48:340-346.
43. Kozarek RA. The Society for Gastrointestinal Intervention. Are we, as
an organization of disparate disciplines, cooperative or competitive? Gut
Liver. 2010;4 (suppl 1):S1-8.
44. Gordon Research Conference. Mucosal Health & Disease 9-14 June
2013. www.grc.org/programs.aspx?year2013&programmucosal. Accessed
February 10, 2012.
45. Sonis ST. Mucositis as a biological process: a new hypothesis for the
development of chemotherapy-induced stomatotoxicity. Oral Oncol. 1998;34:
39-43.
46. Rubenstein EB, Peterson DE, Schubert M, et al. Clinical practice
guidelines for the prevention and treatment of cancer therapy-induced oral
and gastrointestinal mucositis. Cancer. 2004;100:2026-2046.
47. Keefe DM, Schubert MM, Elting LS, et al. Updated clinical practice
guidelines for the prevention and treatment of mucositis. Cancer. 2007;109:
820-831.
48. Worthington HV, Clarkson JE, Bryan G, et al. Interventions for
preventing oral mucositis for patients with cancer receiving treatment.
Cochrane Database Syst Rev. 2011;13 Apr (4):CD000978.
49. Clarkson JE, Worthington HV, Furness S, et al. Interventions for
treating oral mucositis for patients with cancer receiving treatment. Cochrane Database Syst Rev. 2010;4 Aug (8):CD001973.
50. Cancer Council Australia. Clinical practice guidelines for the treatment
and management of endometrial cancer. wiki.cancer.org.au/australia/Guide
lines:Endometrial_cancer/Treatment/Early_stage/Development_process. Accessed February 10, 2012.
551
SHORT- AND LONG-TERM CARDIOVASCULAR

COMPLICATIONS OF CANCER TREATMENT:
OVERVIEW FOR THE PRACTICING ONCOLOGIST
CHAIR
Gretchen Kimmick, MD, MS
Durham, NC
SPEAKERS
Dawn L. Hershman, MD, MS
Columbia University Medical Center
New York, NY
Marianne Ryberg, MD
Herlev University Hospital
Herlev, Denmark
Doug Sawyer, MD, PhD
Nashville, TN
Short- and Long-term Cardiovascular

Complications of Cancer Treatment:
Overview for the Practicing Oncologist
By Chetan Shenoy, MBBS, and Gretchen Kimmick, MD, MS
Overview: As new therapies improve survival from cancer,

attention to comorbid illness and complications of therapy
both short- and long-term become much more important to
improving not only quality of life but also overall survival.
Recognized for its importance as the leading cause of death in
the United States, heart disease often coexists with cancer,
and cancer treatment may increase risk and/or severity. In
N ESTIMATED 1.6 million new cases of invasive

cancer are diagnosed in the United States each year.1
Modern cancer therapies have improved survival, such that
some cancers that were deemed uncontrollable a few years
ago are now considered chronic diseases and some are
cured. In other words, there are more people living with and
living after cancer. In these patients, we have recognized the
importance of not only the short-term, but also the long-term
side effects of our therapies. For instance, many cancer
survivors may have a higher risk of cardiovascular disease
than of cancer recurrence; in a study of 63,566 women with
breast cancer age 66 and older from the Surveillance,
Epidemiology and End Results (SEER)-Medicare linked
database with 12 years follow-up, survivors were equally
likely to die as a result of cardiovascular disease as they
were from breast cancer.2
In oncology, the complex business of preventing cardiovascular disease, treating predisposing factors for cardiovascular disease, ongoing management of the disease itself, and
minimizing cardiovascular complications of cancer therapy
needs multidisciplinary attentionand requires us to be
acutely aware of these issues.
Both cardiovascular disease and cancer are more prevalent with increasing age. An estimated 82.6 million American adults (more than one in three) have one or more types
of cardiovascular disease hypertension, coronary artery
disease, heart failure, or cerebrovascular disease.3 It is,
therefore, not surprising that patients diagnosed with cancer might also have cardiovascular disease or may be more
vulnerable to cardiovascular complications of therapy. In
one report, 38% of patients with cancer had hypertension.4
An analysis of the SEER-Medicare database found that
among breast cancer survivors, 1.7% had myocardial infarction, 6.7% had congestive heart failure, 2.6% had peripheral
vascular disease, and 4.3% had cerebrovascular disease at
the time of diagnosis.5 Furthermore, pre-existing cardiovascular disease, such as hypertension, coronary artery disease,
or cardiomyopathy, is a well-recognized risk factor for cardiovascular complications from chemotherapy, such as anthracyclines6,7 and trastuzumab.8,9 Some of our newer
cancer therapies also have specific adverse effects on the
cardiovascular system, regardless of age. Cardiotoxicity related to anticancer treatment, therefore, may have an important effect on the overall prognosis and survival of
patients with cancer.10
Recognized cardiovascular complications of cancer treatment are many. The more commonly recognized include
addition, there are well-recognized cardiovascular toxicities

of cancer treatment, including not only cardiomyopathy, but
also hypertension, hypercholesterolemia, and others. Oncologists and cardiologists are working closely to learn more
about the complex interaction and to improve management
and outcome for patients.
KEY POINTS
The high prevalence of cardiovascular diseases and

cancer requires attention to their interactions and
prevention.
Many cancer survivors have a higher risk of cardiovascular disease than of cancer recurrence.
Cardiovascular complications of cancer treatment
have long been recognized and we are now beginning
to understand the mechanisms and design logical
preventive strategies.
Preexisting cardiovascular diseases, including hypertension, coronary artery disease, cardiomyopathy,
and others, are well recognized risk factors for cardiovascular complications of chemotherapy.
heart failure or left ventricular dysfunction, myocardial

ischemia/infarction, hypertension, and thromboembolism.
Less commonly seen are arrhythmias, myocarditis, and
pericarditis. Also important and recently recognized are
changes in cholesterol profiles, which may increase risk of
adverse cardiovascular outcomes. These range in severity
and chronicity according to the patient and the agent (type
of chemotherapy or radiation) and likely require the attention of our cardiology colleagues.
Management of cardiovascular complications of cancer
treatment is likely to remain a significant challenge for both
cardiologists and oncologists in the future, as a result of the
growing size of the aging population of patients with cancer
and the introduction of new cancer therapies. Identifying
and understanding these effects is therefore crucial to the
successful treatment of patients with cancer. Unfortunately,
strong scientific evidence for the treatment of patients with
cardiovascular complications from cancer treatment is lack-
From the Divisions of Cardiology and Medical Oncology, Duke University Medical
Center, Durham, NC.
Address reprint requests to Gretchen G. Kimmick, MD, MS, Division of Medical Oncology,
Duke University Medical Center, Box 3204, Suite 3800 Duke South, Durham, NC 27710;
email: gretchen.kimmick@duke.edu.
1092-9118/10/1-10
553
SHENOY AND KIMMICK
ing since studies of cardiovascular disease have generally

excluded patients with cancer, and studies of patients with
cancer have generally excluded patients with concomitant

heart disease.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Chetan Shenoy*
Gretchen Kimmick
Wyeth
REFERENCES
CA Cancer J Clin. 2011;61:212-236.
2. Patnaik JL, Byers T, DiGuiseppi C, et al. Cardiovascular disease
competes with breast cancer as the leading cause of death for older females
diagnosed with breast cancer: a retrospective cohort study. Breast Cancer Res.
2011;13:R64.
3. Roger VL, Go AS, Lloyd-Jones DM, et al. Heart disease and stroke
statistics2011 update: a report from the American Heart Association.
Circulation. 2011;123:e18-e209.
4. Piccirillo JF, Tierney RM, Costas I, et al. Prognostic importance of
comorbidity in a hospital-based cancer registry. JAMA. 2004;291:2441-2447.
5. Patnaik JL, Byers T, Diguiseppi C, et al. The influence of comorbidities
on overall survival among older women diagnosed with breast cancer. J Natl
Cancer Inst. 2011;103:1101-1111.
554
6. Doyle JJ, Neugut AI, Jacobson JS, et al. Chemotherapy and cardiotoxicity in older breast cancer patients: a population-based study. J Clin Oncol.
2005;23:8597-8605.
7. Pinder MC, Duan Z, Goodwin JS, et al. Congestive heart failure in older
women treated with adjuvant anthracycline chemotherapy for breast cancer.
J Clin Oncol. 2007;25:3808-3815.
8. Perez EA, Suman VJ, Davidson NE, et al. Cardiac safety analysis of
doxorubicin and cyclophosphamide followed by paclitaxel with or without
trastuzumab in the North Central Cancer Treatment Group N9831 adjuvant
breast cancer trial. J Clin Oncol. 2008;26:1231-1238.
9. Guarneri V, Lenihan DJ, Valero V, et al. Long-term cardiac tolerability
of trastuzumab in metastatic breast cancer: the M.D. Anderson Cancer
Center experience. J Clin Oncol. 2006;24:4107-4115.
10. Shenoy C, Klem I, Crowley AL, et al. Cardiovascular complications of
breast cancer therapy in older adults. Oncologist. 2011;16:1138-1143.
Recent Advances in Cardiotoxicity of

Anticancer Therapies
By Marianne Ryberg, MD
Overview: The treatment of two major diseases in the Western world, cancer and heart disease, has improved significantly in recent years. Today, many more cancers are curable
than in previous years. Cancer treatment often consists of
chemotherapy, radiation therapy, and now also targeted therapy. All three types of treatment can lead to an increased risk
of developing or of worsening a pre-existent cardiovascular
disease either during the treatment, immediately afterward,
or several years after cessation of therapy. Anthracyclines, a
class of drugs that are also known as anthracycline antibiotics, and the drug cisplatin have contributed to the success
of cancer treatment. However, these agents can cause cardiovascular disease during treatment, and studies have shown
that the risk of disease persists for many years after treatment
stops. Irradiation contributes significantly to this risk when
the cardiovascular system is part of the radiation field. If
the targeted therapy also inhibits the genes responsible for
maintaining the function of the cardiovascular system, development of cardiovascular symptoms is inevitable. Therefore, it
is essential to have a cardiovascular endpoint in trials with
targeted therapy. When treatment stops, however, the effect
on the cardiovascular system appears to cease, but it is not
known whether the long-term risk of developing cardiovascular disease increases. Combined, these factors indicate that
close cooperation between oncologists and cardiologists is
essential to optimally benefit patients with cancer.
cardiotoxicity, Van Hoff 4 showed that the risk of developing

cardiac failure increased exponentially when the doxorubicin dose was increased. Another anthracycline, epirubicin,
apparently has a linearly increased risk of approximately
40% for each 100 mg/m2.5 This indicates that the cardiac
myocytes will be affected as of the very first treatment with
epirubicin. However, several studies show that risk increased with serious comorbidity, the various ages of the
different patients, previous irradiation (including of the
heart), and concomitant antitumor therapy.4-6 Furthermore
the upper limit of dose for doxorubicin and epirubicin is
based on retrospective studies in patients with metastatic
diseases. The problem with using this subset of patients to
establish upper limits of dose is that some of these patients
will die from cancer before they develop symptoms of heart
failure. Furthermore, a predisposing genetic variant involved in the oxidative stress or metabolism and transport of
anthracycline can contribute to heart failure.7 One paradox
is that an increase of the dose of the anthracycline appears
to improve the survival rate for those with metastatic
diseases.5 This calls for a more personalized approach for
treatment of patients with metastatic disease using an
anthracycline. Furthermore, the risk of developing cardiotoxicity in the long term after an anthracyline-based adjuvant treatment is not addressed by the aforementioned
studies.
A prominent feature of the risk of developing cardiotoxicity is subcellular damage or even necrosis in the heart
muscle cells during treatment, which seems to continue even
after treatment has stopped.8 This is believed to be caused
by an increase in oxidative stress in cells. Reactive oxygen
species (ROS) are formed when the quinine moiety of anthracyclines is reduced to semiquinone, thus initiating a
cascade of free-radical formation. Furthermore a highly
reactive ROS is created when an anthracycline uncouples
YTOTOXIC DRUGS, targeted therapy, and radiation

therapy have changed the final outcome for many
patients with cancer, many of whom are now cured of
cancers there were considered fatal just a few years ago.
Instead of a death sentence, their diagnosis is comparable to
that of a chronic disease with periods of remission, exacerbation, and re-treatment. This means that treating patients
with cancer involves new challenges because modern cancer
treatments can have severe side effects, not only during
treatment but also in the long run. One side effect is the
increased risk of cardiovascular disease, either transient or
permanent depending on the nature of the injury to the
cardiovascular system.1-3 Some of the types of cardiovascular disease that presumably result from cancer treatment
are heart failure, ischemia, hypertension, hypotension, arrhythmias, conductive disorder, thromboembolic events, and
QTc prolongation. The risk of developing cardiovascular
disease from cancer treatment may be underestimated because of the process drugs undergo to be approved for
general use. To be approved for general use, positive phase
I through phase III trials must be conducted. Patients
participating in these trials are not always representative
of the average patient with cancer because patients with
significant comorbidityincluding cardiovascular disease,
children, and the elderlyare excluded from the trials. This
means that the risk of cardiac disease is apparently greater
than shown in the trials. In recognition of these issues,
cooperation between oncologists and cardiologists is crucial.
Anthracyclines
Some of the classic cytostatic drugs, such as anthracyclines, cyclophosphamid, 5-fluorouracil, and cisplatin, are
known to cause cardiovascular disease. Anthracyclines are
one of the most feared of these treatments because of their
ability to cause heart failure, not only in connection with the
treatment but also in the long term. Because they are also
highly active, widely used drugs, many long-term survivors
of cancer treatment (including children) have been successfully treated with anthracycline-based chemotherapy. Anthracylines remain an indispensable option in treating
cancer today, both for children and adults. Doxorubicin, an
anthracycline and one of the most active anticancer drugs
used today, was quickly recognized as having serious side
effects, including heart failure. In a retrospective study of
From the Department of Oncology, Herlev Hospital, University of Copenhagen, Denmark.

Address reprint requests to Marianne Ryberg MD, Department of Oncology, Herlev
Hospital, University of Copenhagen, Herlev Ringvej 75, DK-2730 Herlev, Denmark; email:
mary@heh.regionh.dk.
1092-9118/10/1-10
555
MARIANNE RYBERG
the electron transport chain in the mitochondria. ROS have

a deleterious effect on the cells, thus leading to cell death
and to organ damage.9,10 A recent experimental study in
rats demonstrates that anthracycline-caused cardiomyopathy can be mediated by depletion of the cardiac stem cell
pool. When exposed to an anthracycline, cardiac progenitor
cells (CPCs) appeared to be even more vulnerable to damaged ROS. Additionally, CPC cells were arrested in the
G2/M cell cycle, resulting in a depletion of the CPC cell pool
in the myocardium. The final result is heart failure in the
animal. To establish whether the delivery of syngeneic
contaminating tumor cells (CTCs) could oppose the progression of anthracycline cardiotoxicity, the syngeneic CTCs
were injected in the failing myocardium. This improved the
regeneration of cardiac myocytes and of vascular structure,
consequently improving ventricular performance and the
rate of animal survival.11 These encouraging results can
perhaps lead to the possibility of implementation of stem cell
transplantation as part of treatment with anthracyclines.
occur not only during treatment but also subsequently. Our

knowledge about the risk of cardiac disease after cessation of
treatment is obtained by studies of testicular cancer survivors. The treatment consisted of a cisplatin-based chemotherapy for patients generally diagnosed between 20 to 40
years of age, and nearly all of them were cured. In all
likelihood, they will survive for years to come, but they will
have an increased risk of developing hypertension, hypercholesterolemia, myocardial ischemia, and MI.14 The increased risk for MI was approximately 5.7% in a median
observation time of 19 years (1319 years) for 990 long-term
survivors treated with cisplatin-based chemotherapy. One
reason for this delayed risk is that the patients had an
increased plasma level of endothelia and inflammatory
marker protein years later that might eventually progress to
more severe endothelia dysfunction and overt atherosclerosis.15 This increase may be because cisplatin can still be
measured in the blood 20 years after treatment.16
5-Fluorouracil/Capecitabine
Timing of Combination Chemotherapy
The story behind the introduction of taxanes in the treatment of patients with breast cancer is a lesson in how
important timing is in order to avoid cardiotoxicity.12 In an
early phase II study doxorubicin and paclitaxel were given
simultaneously for metastatic breast cancer, resulting in
an impressive response but with approximately one-fifth of
the patients experiencing cardiotoxicity.12 Pharmacokinetics studies show that the presence of paclitaxel increases the
plasma level area under the curve (AUC) of doxorubicin by
30%, compared with a 24-hour delay in administration of a
paclitaxel infusion. A later phase III study that separated
the administration of doxorubicin and paclitaxcel by 24
hours did not show an increase in cardiotoxicity.13 These
important results have been translated to benefit the adjuvant treatment of patients with breast cancer. Part of the
adjuvant anthracycline-based treatment has been replaced
with a taxane-based treatment so that the cumulative dose
of an anthracycline recommended in adjuvant treatment has
been reduced.
Cisplatin
During treatment, some chemotherapy agents can cause

ischemia syndrome (i.e., chest pains), palpitations, and in
rare cases a lethal myocardial infarction (MI). This can
happen with cisplatin, for example, where the syndrome can
KEY POINTS
556
The risk for development of cardiac disease is substantial after treatment with chemotherapy, radiation therapy, or with targeted therapy.
In order to avoid cardiotoxicity, timing of combination
therapies is essential.
The cardiac risk in the long term for patients treated
with targeted therapy is not known.
Preclinical and clinical trials must be designed to
evaluate a cardiac risk.
Cooperation between oncologists and cardiologists is
crucial.
5-Fluorouracil and its prodrug, capecitabine, can cause

ischemic syndrome during treatment, but the symptoms
disappear when treatment stops. The mechanism is not well
established, but may be a result of coronary vasospasm. It
has also been reported to cause Takotsubo cardiomyopathy.17 These two drugs are the foundation of the adjuvant
treatment of colorectal cancer. When needed, the treatment
can be repeated despite the occurrence of previous cardiac
events. The treatment must be given in close cooperation
with cardiologists and at lower doses. Capecitabine is also
used in palliative setting for many types of cancer, including
breast cancer. The true incidence of cardiac events is not
known but is believed to be approximately 1% to 18%.18,19
Compared with cisplatin, no increased risk of cardiotoxicity
has been reported subsequently.
Radiotherapy
Radiotherapy has been and is still an essential modality in

the curative treatment of many types of cancer. This applies
to both children and adults with cancer. Radiation therapy
can cause coronary artery disease (CAD), diffuse myocardial
fibrosis, and the more seldom-occurring pericarditis and
pericardial fibrosis. The most common side effect, however,
is CAD. A radiation dose of greater than 30 GY is definitively
known to cause damage to the heart. However, it is unknown whether a threshold dose exists that does not incur
risk to the heart.20 Because of a prolonged latency of 10 to 15
years before the occurrence of cardiac disease, it is difficult
to be certain as to whether the increased incidence is
because of the irradiation or whether it is caused by a
comorbidity due to aging. A study of 4,122 survivors treated
for cancer in their childhoods has increased understanding
of the extent of the risk. Compared with the general population, patients in this study were generally shown to have 5
times greater risk of dying of cardiac disease. The authors
have retrospectively estimated that the mean radiation dose
delivered to the heart and the risk of dying of cardiac disease
increased linearly when the average radiation dose exceeded
5 Gy.21
Patients with breast cancer (BC) who underwent irradiation to the right breast had a lower risk for MI than those
patients treated for left-sided BC.20 This was demonstrated
CARDIOTOXICITY OF ANTICANCER THERAPIES
by doing a retrospective estimation of the radiation dose to

the heart in patients with BC. Researchers have defined
hotspots for radiation, which included the proximal right
coronary artery, the mid- and distal descending artery, and
the distal diagonal artery.20 A Swedish group examined the
distribution of coronary stenosis in a coronary angiography
in patients treated with adjuvant radiotherapy for BC. The
patients were compared with otherwise healthy woman
referred for a coronary angiography, and the incidence of
coronary stenosis was similar. The anatomic location of the
stenosis was substantially different. In patients who were
treated for left-sided BC, stenosis increased substantially
in the areas deemed to be hotspots. The relative risk for
high-risk compared with low-risk patients (no radiation
therapy) was 1.90 for stenosis, grades 35 (95% CI: 1.11 to
3.24).22
The biologic processes set in motion by radiation therapy
continue after the end of therapy, and side effects might not
be visible until decades later. Radiation decreases the capillary density with the passage of time, thus causing a loss of
the flow reserves territory of myocardium in the radiation
portal. This leads to ischemia of the myocardium. Radiation
also seems to accelerate the development of age-related
atherosclerosis in patients, thus causing stenosis/thrombosis in the arteries in already weakened areas. This also
explains the additive effect of anthracycline-based treatment, because of its ability to weaken the myocardium.
In order to decrease the radiation dose to the heart, use of
radiation-reducing techniques are essential. The introduction of a three-dimensional computerized planning system
makes dose delivery estimation possible. By defining specific
cardiac substructures, such as arteries and the entire heart,
the dose can be estimated. As a result, a dose relationship
can be established between the risk of cardiac disease and
the dose delivered. In the case of hotspots, another field
technique can be considered, if possible. A promising new
technique is breath holding, where the dose is delivered
when the heart is out of the irradiation field. Modern
techniques such as this one are important because cancer
survival rates are improving, and awareness of the longterm effects of irradiation can help decrease risks.
Targeted Therapies
A highly promising new way to treat cancer is targeted

therapy. The main principle is to use drugs designed to
inhibit the dysregulated genes that drive malignant transformation. Targeted cancer therapies currently available are
monoclonal antibodies (mAbs), which target growth factor
receptors, and other small molecules, which are inhibitors of
kinases (mostly of tyrosine kinase inhibitors, TKIs). Kinases
are enzymes that catalyze the transfer of phosphate from
ATP, usually to a serine, threonine, or to tyrosine residue on
protein substrates. These reactions act as critical mediators
of cellular signal transduction, and thereby regulate cellular
processes including cell cycle, progression, metabolism,
transcription, and apoptosis. Mutation in their genes contributes to the malignant transformation in normal cells in
many cancers. Inhibition of the mutated genes can also
interfere with the function of the genes in a normal cell,
including the function of the cells in the cardiovascular
system.23 Cardiac toxicity caused by targeted therapies can
be divided into on-target and off-target toxicity. On-target
cardiac toxicity means that the inhibition of the mutation in
the specific kinase responsible for the malignant transformation also has a decisive influence on the maintenance of
normal function in cardiac myocytes. As a result, the risk of
cardiotoxicity is inevitable when this therapy is used.
An example of on-target toxicity involves trastuzumab,
which is a human monoclonal antibody directed against the
extracellular region of the HER2/ErbB2 receptor. The normal HER2 gene is involved in cell proliferation, angiogenesis, and cell survival. Thus, a deficiency in the HER2 gene
will result in the downstreaming of these processes and
cause a malignant transformation of the cells. Furthermore,
an overexpression or amplification of tyrosine kinase epidermal growth factor HER2/ErbB2 is associated with an aggressive clinical phenotype of BC with a poor survival rate.24
The monoclonal antibody trastuzumab was introduced in
the treatment of HER2-positive BC in both the adjuvant and
metastatic setting, and it produced impressive results.25
The preclinical trials showed no sign of cardiotoxicity, but it
soon became clear that the drug could cause cardiotoxicity.
Five percent to 7% of the patients developed cardiac dysfunction in monotherapy, and 28% did so when trastuzumab
was given concomitantly with an anthracycline.24 Trastuzumab was then given sequentially with a remarkable
reduction of the cardiac dysfunction in both the metastatic
and adjuvant settings. The HER2 gene and its ligand
neuregulin are also involved in normal cardiomyocyte proliferation during development and survival throughout
life.26 Thus, trastuzumab, by inhibiting HER2, also inhibits
the myocardial survival pathway; its influence on the function of the cardiac myocyte is inevitable. This may result in
cardiac dysfunction and thus increases the risk for heart
failure. Meanwhile, the myocardiac biopsies show that the
myocyte structure is normal without myofibrillar loss or
ultra-structural disturbance, which is in contrast to anthracyline.24 There are indications that the biologic process that
trastuzumab sets in motion may be reversible. The ability of
cardiac myocytes to regenerate may decrease when an
anthracycline and trastuzumab are given concomitantly.13
The HERA trial, which had a 3.6-year follow-up, showed
that when trastuzumab was given after treatment with an
anthracycline, the majority of cardiac events occurred during the treatment. The incidence was found to be 5%. The
cardiac dysfunction improved in 80% of these patients, and
only 20% had progressive dysfunction during follow-up.27 A
meta-analysis of 11,882 patients from 10 randomized trials
has been published and showed that the addition of trastuzumab to anthracycline-based chemotherapy significantly
increased the risk of cardiac dysfunction (relative risk 4.27;
CI, 2.75 6.61, p 0.00001).28 The risk seems to increase
with old age, hypertension, and obesity. In contrast to
anthracycline treatment, treatment with trastuzumab can
be repeated, but data on long-term follow-up is not available
yet.
Antiangiogenese Inhibitors
Antiangiogenese inhibitors are designed to target the

vascular endothelial growth factor (VEGF), which may lead
to an increased risk of developing hypertension, cardiotoxicity, and to the occurrence of thomboembolic events. The
most prominent types of this kind of inhibitor are bevacizumab, sunitinib, and sorafenib. Bevacizumab is a recombinant humanized monoclonal antibody that binds VEGF and
prevents it from binding to its receptors (VEGFR-1 and 2).
557
MARIANNE RYBERG
The drug has been approved in treatment of many types of

cancer (e.g., colorectal, lung, and breast cancers). Hypertension is the adverse event correlated most strongly to bevacizumab, with an incidence ranging from 16% to 47%.29 From
the results of prospective trials, it seems that higher doses
of bevacizumab result in higher incidence of hypertension.
However, the exact dose that causes hypertension is not
known.30 This effect appears to be dose dependent, but the
exact dose that causes hypertension is not known. Given in
combination with sorafenib, the incidence of hypertension
was 67%.31 This is because VEGF/VEGFR signaling affects
the adaptive response of the heart to blood pressure stress.
In addition, signal pressure is inhibited, which leads to
heart failure. The incidence of ischemic heart disease and
arterial thromboembolic events is as high as 3.3%.30 Approved for the treatment of patients with metastatic renal
cancer, the TKIs sunitinib and sorafenib are multikinases
inhibitors. Sunitinib is also used in the treatment of gastrointestinal stromal tumors, and sorafenib is used in the
treatment of advanced hepatocellular carcinoma. They are
inhibitors of growth factor receptors, the most important of
which are VEGF, platelet-derived growth factor, the stem
cell factor KIT receptor, and the Von Hippel-Lindau
hypoxia-inducible gene pathway. Furthermore, sorafenib
also affects the RAS-RAF-MEK-ERK pathway by inhibiting
the intracellular kinases RAF and BRAF. This inhibition
substantially affects tumor angiogenesis and/or cell proliferation. These kinases also play an important role in the
maintenance of the normal function of the vascular system.
Therefore, inhibition may also lead to dysfunction in the
cardiovascular system. An observational single-center study
analyzed 86 patients who received TKI treatment for metastatic renal carcinoma and found that 18% had experienced
life-threatening cardiovascular events. The cancer therapy
was interrupted and treatment of the cardiovascular symptoms was initiated, after which all patients were able to
continue treatment. This highlights the reversibility of cardiac symptoms and the importance of cardiac monitoring
and intervention for these patients.32
The previously mentioned inhibitors are only the beginning of a new era, in which many more are in development
for the treatment of cancer and other diseases. As a result,
developing preclinical systems that can detect whether a
potential drug is cardiotoxic is imperative. It is equally
important that clinical trials be designed to assess the risk
of cardiotoxicity, if there is any suspicion of it during
preclinical trials.
Conclusion
The reason why cardiovascular symptoms develop during

and after the treatment of cancer depends on the treatment
modality and on the patients general status. As a result,
increasing physician knowledge about the various treatment
regimens and their short- and long-term effects on the heart
is crucial. Furthermore, as many heart diseases are treatable, close collaboration between cardiologists and oncologists is imperative. The number of new therapies available
for treating patients with cancer will continue to grow,
which is why cataloging and updating side effects, including
long-term side effects, are essential to keeping individual
specialists abreast of the newest treatment developments.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Marianne Ryberg*
REFERENCES
1. Albini A, Pennesi G, Donatelli F, et al. Cardiotoxicity of Anticancer
drugs: The need for Cardio-Oncological Prevention. J Natl Cancer Inst.
2010;102(1):14-25.
2. Carver JR, Shapiro CL, Ng A, et al. American Society of Clinical
Oncology Clinical Evidence Review on the Ongoing Care of Adult Cancer
Survivors: Cardiac and Pulmonary Late Effect. J Clin Oncol. 2007;25:39914008.
3. Minotti G, SalvatorelliE, Menna P. Pharmacological Foundations of
Cardio-Oncology. J Pharmacol Exp Ther. 2010;334:2-8.
4. von Hoff DD, Layard MW, Basa P, et al. Risk Factors for Doxorubicininduced congestive heart failure. Ann Intern Med. 1970;91:710-717.
5. Ryberg M, Nielsen D, Cortese G, et al. New insight to epirubicin cardiac
toxicity. Competing risks analysis of 1097 breast cancer patients. J Natl
Cancer Inst. 2008;100:1-10.
6. Swain SM, Whaley FS, Ewer MS. Congestive Heart Failure in Patients
Treated with Doxorubicin. A retrospective Analysis of Three Trials. Cancer.
2003;97:2869-2879.
7. Blanco JG, Leisenring WM, Gonzalez-Covarrubias VM, et al. Genetic
Polymorphisms in the carbonyl reductase 3 gene CBR3 and the NAD(P)H:
Quinone Oxioreductase 1 gene NQO1 in patients who developed anthracycline-related congestive heart failure after childhood cancer. Cancer. 2008;
112:2789-2895.
8. Zhou S, Palmeira CM, Wallace KB. Doxorubicin-induced persistent
oxidative stress to cardiac myocytes. Toxicol Lett. 2011;121:151-157.
9. Minotti G, Menna P, Salvatorelli E, et al. Anthracyclines: Molecular
advances and pharmacologic development in antitumor activity and cardiotoxicity. Pharmacol Rev. 2004;56:185-229.
558
10. Wallace KB. Adriamycin-induced interference with cardiac mitochondrial calcium homeostasis. Cardiovasc Toxicol. 2007;7:101-107.
11. De Angelis A, Piegari E, Cappetta D, et al. Anthracycline cardiomyopathy is mediated by depletion of the cardiac stem cell pool and is rescued by
restoration of progenitor cell function. Circulation. 2010;121:276-292.
12. Gianni L, Munzone E, Capri G, et al. Paclitaxel by 3-hour infusion in
combination with bolus doxorubicin in women with untreated metastatic
breast cancer: high antitumor efficacy and cardiac effects in a dose-finding
and sequence-finding study. J Clin Oncol. 1995;23:2688-2699.
13. Gianni L, Salvatorelli E, Minotti G. Anthracycline cardiotoxicity in
breast cancer patients: Synergism with trastuzumab and taxanes. Cardiovasc
Toxicol. 2007;7:67-71.
14. Hauges HS, Wethal T, Aass N, et al. Cardiovascular Risk Factors and
Morbidity in Long-Term Survivors. J Clin Oncol. 2010;28:4649-4657.
15. Meinardi MT, Gietema JA, van der Graaf WT, et al. Cardiovascular
Morbidity in Long-Term Survivors of Metastatic Testicular cancer. J Clin
Oncol. 2000;18:1725-1732.
16. Giordano SH, Kuo YF, Freeman JL, et al. Risk of cardiac death after
adjuvant radiotherapy for breast cancer. J Natl Cancer Inst. 2005;97:419-424.
17. Grunwald MR, Howie L, Diaz LA Jr. Takotsubo Cardiomyopathy and
Fluorouracil: Case Report and Review of the Literature. J Clin Oncol.
2012;30:e11-e14. Epub 2011 Dec 5.
18. Dalzell JR, Samuel LM. The spectrum of 5-fluorouracil cardiotoxicity.
Anti-cancer Drugs. 2009;20:79-80.
19. Jensen SA, Srensen JB. Risk Factors and prevention of cardiotoxicity
induced by 5-fluorouracil or capecitabine. Cancer Chemother Pharmacol.
2006; 58:487-493. Epub 2006 Jan 18.
CARDIOTOXICITY OF ANTICANCER THERAPIES

20. Taylor CW, Nisbet A, McGale P, et al. Cardiac exposures in breast
cancer radiotherapy: 1950s-1990s. Int J Radiat Oncol Biol Phys. 2007;69:
1489-1495.
21. Tukenova M, Guibout C, Oberlin O, et al. Role of Cancer Treatment in
Long-Term Overall and Cardiovascular Mortality After Childhood Cancer.
J Clin Oncol. 2010;28:1308-1315. Epub 2010 Feb 8.
22. Nilsson G, Holmberg L, Garmo H, et al. Distribution of Coronary Artery
Stenosis After Radiation for Breast Cancer. J Clin Oncol. 2012;30:380-386.
Epub 2011 Dec 27.
23. Force T, Kerkela R. Cardiotoxicity of the new cancer therapeutics
mechanisms of, and approaches to, the problem. Drug Discov Today. 2008;
13:778-784. Epub 2008 Sep 2.
24. Ewer SM, Ewer MS. Cardiotoxicity Profile of Trastuzumab. Drug Saf.
2008;31:409-467.
25. Slamon DJ, Leyland-Jones B, Shak S, et al. Use of chemotherapy plus
a monoclonal antibody against HER2 for metastatic breast cancer that over
expresses HER2. N Engl J Med. 2001;344:783-792.
26. Stortecky S, Suter T. Insights into cardiovascular side-effects of modern anticancer therapeutics. Curr Opin Oncol. 2010;22:312-317.
27. Procter M, Suter T, de Azambuja E, et al. Longer-Term Assessment of

Trastuzumab-Related Cardiac adverse Events in the Herceptin Adjuvant
(HERA) Trial. J Clin Oncol. 2010;28:3422-3428.
28. Chen T, Xu T, Li Y, et al. Risk of cardiac dysfunction with trastuzumab
in breast cancer patients: A meta-analysis. Cancer Treat Rev. 2011;37:312320. Epub 2010 Oct 16.
29. Zambelli A, Della Porta MG, Eleutri E, et al. Predicting and preventing
cardiotoxicity in the era of breast cancer targeted therapies. Novel molecular
tools for clinical issues. Breast. 2011;20:170-80. Epub 2010 Dec 13.
30. Vaklavas C, Lenihan D, Kurzrock R, et al. Anti-vascular Endothelial
Growth Factor Therapies and Cardiovascular Toxicity: What are the Important Clinical Markers to Target? Oncologist. 2010;15:130-141. Epub 2010
Feb 5.
31. Azad NS, Posadas EM, Kwirkowski VE, et al. Combination targeted
therapy with sorafenib and bevacizumab results in enhanced toxicity and
antitumor activity. J Clin Oncol. 2008;26:3709-3714.
32. Schmidinger M, Zielinski CC, Vogl UM, et al. Cardiac Toxicity of
Sunitinib and Sorafenib in patients with metastatic renal cell carcinoma.
J Clin Oncol. 2008;26:5204-5212. Epub 2008 Oct 6.
559
WHEN CANCER BECOMES PERSONAL: THE

PHYSICIANS VIEW OF PATIENT CARE
CHAIR
Teresa Gilewski, MD
New York, NY
SPEAKERS
Peter Bach, MD
New York, NY
Martin Raber, MD
Houston, TX
George W. Sledge Jr., MD
Indianapolis, IN
The Oncologist as the Patient with Cancer or

Relative
By Teresa Gilewski, MD, Martin Raber, MD, and George W. Sledge Jr., MD
Overview: To an extent, physicians are familiar with the

consequences of illness through their interactions with patients. However, when cancer becomes personal, the physician has an opportunity to gain greater insight into the
intricacies of medical care, including its humanistic elements.
Physicians who encounter cancer in themselves or in a relative may deepen their understanding of the patient experience.
ERSONAL EXPERIENCES may influence an individuals thoughts and actions. It is plausible that a physicians perspective of patient care may be affected by an
encounter with illness on a personal level. Any human being
who is sick or who has experienced illness along with a
family member or close friend has the opportunity to gain
considerable insight into the complexities of medical care.
This may be especially true for physicians who have the
unique position of viewing patient care from the perspective
of the provider as well as the recipient. In particular, the
humanistic aspects of medicine may become more apparent
and more meaningful. Human interactions form the core of
patient care, and optimal patient care requires adequate
observation and self-reflection of these interfaces.
Teresa Gilewski, MD: The Humanistic Side of

MedicineThe Impact from a Personal
Cancer Experience
The practice of medicine is a complex interplay of science,

humanism, business and social policy. During different
periods in history, the emphasis on these various elements
has shifted. For example, many of the basic moral and
ethical tenets of medicine were established by Hippocrates
and others in ancient Greece at a time when there was
limited scientific knowledge.1 With the onset of new scientific discoveries, a greater focus on technology and science
was inevitable. In 1910, the noted educator Abraham Flexner emphasized the importance of both a scientific and
clinical research basis in medical education.2 His report
helped to restructure medical school curriculum to one that
was more research focused. However, 15 years later he felt
that scientific medicine was . . . sadly deficient in cultural
and philosophic background.3 Finding a balance between
the scientific and the human components of medicine remains a challenge to the present.
In addition, now there is an ever-increasing presence, on a
daily basis, of the business and social aspects of medical
care. The humanistic aspects of medicine have often become
secondary to these other more easily measurable facets of
medicine. However, in the last decade medical groups, such
as the American Board of Internal Medicine and Institute of
Medicine, have underscored the importance of practicing
medicine in a humanistic manner.4,5 The association between professionalism and humanism has become clearer.6,7
Yet, there are many challenges, both on institutional and
personal levels, that may hinder the incorporation of a
humanistic approach into patient care. These include the
emphasis on research and greater clinical productivity as
well as the stresses associated with caring for ill patients.3,8
Their views provide a unique perspective, on the basis of the

convergence of their medical knowledge and personal reaction to illness. They also confront distinct challenges specific
to their work environment. An enhanced recognition of their
viewpoints provides valuable information in the quest to
alleviate patient suffering and explore the fundamentals of
patient care.
The medical literature contains heartfelt essays that focus

on the importance of human interactions in medicine.9,10 In
particular, the sections A Piece of My Mind in the Journal
of the American Medical Association and the Art of Oncology in the Journal of Clinical Oncology provide physicians
an opportunity to contemplate various aspects of the personal impact of illness.11-13 Narrative medicine utilizes the
practice of writing about patients experiences and the
writers reflection on those experiences to foster empathy.14
Physicians usually develop a general awareness of the
difficulties that patients confront, but not necessarily a true
understanding of those struggles. However, physicians who
encounter cancer in themselves or in a family member
develop a new intimacy with illness.15-17 This altered relationship with disease may result in a heightened awareness
of the consequences of sickness.18 Once there is a personal
connection with cancer, the day-to-day routine of the medical system may assume a fresh look. What may have seemed
trivial and of little significance may become momentous.
Some of the usual inconsequential basic human interactions
between a patient and the physician suddenly become memorable for their compassion or lack thereof.
Experiencing cancer on a personal level may provide the
physician with new points of reference that have the potential to influence ones perspective of patient care. These
experiences may originate not only in adulthood but in
childhood as well. In childhood, the effect of caring for a sick
relative and observing the interactions of adult relatives and
physicians coping with illness may be long remembered.
This child who later becomes a physician may use some of
these observations toward the care of patients. Of course,
each physician brings a unique personality and other life
experiences to every situation. However, exposure to personal illness has the potential to enhance the physicians
appreciation of the fragility and uncertainty of life as well as
the value of kindness.
The physicians appreciation of the human experience
surrounding illness may be heightened, not only in regards
to patients but in regards to colleagues. Perhaps on the basis
of this greater awareness, a deeper recognition of the essence of physicians may develop. Specifically, that they are
From the Memorial Sloan-Kettering Cancer Center, New York, NY; M. D. Anderson
Cancer Center, Houston, TX; University of Indiana, Indianapolis, IN.
Address reprint requests to Teresa Gilewski, MD, Memorial Sloan-Kettering Cancer
Center, 300 East 66th St., New York, NY 10065; email: gilewskt@mskcc.org.
1092-9118/10/1-10
561
GILEWSKI, RABER, AND SLEDGE
imperfect human beings, who in their relationships with

patients may at times fall short of their own best intentions
and at other times reach profound moments of harmony.
Does experiencing illness personally make one a more
humanistic physician? Perhaps for some the answer is yes,
for others no, but for all it is a possibility. Regardless, it is
important to generate dialogue about the humanistic element of medicine because it is fundamental to patient care.
As the noted physician William Osler (1849 1919) eloquently stated, Care more for the individual patient than
for the special features of the disease . . . Put yourself in his
place . . . The kindly word, the cheerful greeting, the sympathetic lookthese the patient understands.
Martin Raber, MD: The Oncologist as the Patient
Without question, being a physician has influenced my

experience as a patient and being a patient has influenced
my experience as a physician. I am not sure that it has made
me a better physician, but I know that it has made me a
different physician. It has also given me a different view of
the clinic, the hospital, and the medical teams. At the same
time that I have been a patient, I have also tried to maintain
a semblance of an academic career and continue to see
patients in my area of expertise. This has been possible only
because my institution and my colleagues have been willing
to make substantial accommodations to limit my responsibilities and to cover me whenever I was unable to work.
I dont think anything prepares you for being sick, and the
experience of entering the hospital as a patient is so incredibly different from that of entering it as a physician that, at
first, one is totally disoriented. As physicians, we tend to
think of the hospital and clinic experience as centered on us.
Although it is true that the physicians decisions and comments are what drives the delivery of care, in many respects
the patient experience is driven by the ancillary personnel.
KEY POINTS
562
A personal experience with illness provides a unique

opportunity for the physician to gain considerable
insight into the humanistic elements of medicine that
are at the core of patient care.
The patient experience is very much dependent on
interactions with the ancillary staff; how they treat
each other and how you treat them is as important as
how they treat the patient.
Physician-patient communication is more difficult
than it seems, particularly in areas in which we do
not have lab tests, images, or clinical findings to
explain the situation.
Caring for patients while also being a patient is a
challenge; it requires substantial accommodation on
the part of ones colleagues, team members, and
patients.
A physician who is the relative of a patient with
cancer may struggle with distinctive challenges that
relate to the intersection of the physicians medical
knowledge and personal emotions with family
dynamics.
Patients spend most of their time with clinic and chemotherapy nurses, clerks, radiology and laboratory technologists,
patient transportation personnel, and others that we tend
not to consider as central to the patient experience. If they
dont deal with the patient in a positive way, the patient
experience is not good. This has to do with not only how they
deal with the patient but also with each other, and how you
as the physician deal with them. Good communication skills
are important not only for the physician but also for everyone who interacts with the patient.
Patients see and hear everything that goes on. We sit in
the waiting rooms and observe the interactions and process
what we see. Even in the examining room we hear the
conversations in the hallway between physician and nurse or
trainee. Sitting in a waiting room (which is what we patients
do a lot of) I am always amazed at how perceptive my fellow
patients are about the staff and the clinic operations.
Another surprise to me as a physician was the difficulty
that I have had at times communicating with my medical
team. This related most often to symptoms or situations for
which there was not a good laboratory or radiologic corollary. I have come to believe that the doctor-patient visit in
the clinic is much more stereotyped than we think. Although
the physician wants to find out how the patient is feeling,
assess the patients condition, and give the patient information about his disease and plans, and the patient wants to
tell the physician how he feels and understand his condition
and the plans, there is tremendous opportunity for misunderstanding. Patient and physician have somewhat different
goals and expectations of the clinic visit. They also have a
conversation in which many words are used without prior
agreement as to their meaning. Good, bad, fair, tired,
and alright are examples of words that may mean very
different things to the patient and the physician. That
realization caused me to substantially change the way that
I interview patients in my clinic, and the way I speak to my
physicians in their clinic.
Trying to maintain a medical career at the same time one
is facing serious illness, and undergoing cancer therapy, is a
challenge. Early in the illness I had what I call doctorpatient confusion, and was unable to practice at all. Later,
after I recovered from some devastating complications, I
found that I had passed through that phase, and once again
began to see patients. In this period I have come to realize
that my illness has had a major effect on my colleagues who
have often been simultaneously my colleagues and physicians, on my team, and on my patients, all of whom have had
to adjust to the realities of my health problems. As have I.
George Sledge, MD: The Oncologist as the Relative
of a Patient with Cancer
Because cancer is common, and because cancer doctors

have relatives with cancer, cancer professionals regularly
come face to face with cancer in relatives. Their unique
perspective on cancera perspective that has both intellectual and emotional components derived from years of caring
for patients with cancer clearly affects how they interact
with those relatives, both for better and worse.
Physicians are routinely taught to avoid taking care of
relatives, an admonition that is both wise and rarely
completely respected. Wise, because physicians need to
maintain emotional distance from their patients. Every
relationship with a relative is fraught with family history,
THE ONCOLOGIST AS THE PATIENT OR RELATIVE
and that history is not always either happy or uncomplicated. That history, and ongoing family dynamics, can
complicate recommendations, and the recommendations
themselves can easily become the source of future family
discord as well as personal feelings of grief or guilt.
At the same time, physician-relatives of patients with
cancer are in the enviable position of being able to point the
patient with cancer in the direction of the best care, or at
least of the best caregiver, available. Although much of
cancer care is standard, not all cancer professionals are
equivalent in either expertise, competence, or compassion, a
fact that physicians are well aware of: a doctors doctor is
often defined in terms of whom a physician is willing to have
care for a relative.
A physician faces additional challenges as the relative of a
patient with cancer. Physicians are routinely asked what
would you do if I was your . . . [sister, brother, daughter, son,
mother, father]? This question is more difficult to dodge or
ignore when it comes from a relative, and the responsibility
attached to the answer differs qualitatively from that encountered in usual practice, rightly or wrongly.
Physician-relatives also face challenges related to their
interactions with caregivers and hospital systems. If a
physician disagrees with the advice given to a relative, what
is that physician to do? If an interaction with the health care
system is a negative one, does the physician-relative claim
special attention for the patient? Under normal circumstances, most physicians accept that their colleagues deserve substantial autonomy in decision making and
therapeutic recommendations, and recognize that the health
care system is imperfect. Indeed, no health care system
would long survive intrusive oversight by colleagues.
Yet this dynamic frequently changes when a physicians
relative enters the health care system, particularly that part
of the system in which the physician-relative practices. This
may result in excessive diagnostic testing and overtreatment of the relative, often to that patients detriment, a
situation akin to the defensive medicine practiced by physicians concerned with malpractice. Physician-relatives who
are aware of this dilemma may be caught between the Scylla
and Charybdis of this dynamic, wishing the best care for a
relative but also wishing to make the best use of a colleagues expertise and wisdom.
The physician is also faced with the problem of knowing
too much. This is frequently the case when the physicians
relative has a poor-prognosis cancer. In this setting, relatives may explicitly encourage false hope, leaving the physician in the emotionally precarious position of balancing
reality and optimism. These discussions may have long-term
consequences, not just for the patient, but for the family as
a whole: emotional wounds that never completely heal.
Under normal circumstances, the physician can go home
and unwind after a hard day at work, but when home is the
source of stress there may be no place to turn.
With all patients with advanced disease, there comes a
time when active therapy is no longer appropriate, and in
which a focus on quality-of-life measures, advanced care
planning, and hospice care become reasonable. What is the
role of the physician-relative in selecting that moment?
Indeed, is there a role?
Finally, the physician-relative must deal with his or her
own emotional needs, both during the relatives treatment
arc and after a relatives death. Physicians are often excellent and compassionate communicators when dealing with
patients to whom they are unrelated; but it is the rare
physician who is capable of expressing his or her own emotional trauma, or who is capable of instituting the healing
process we all deserve in our own most profound times of grief
and loss. Dealing with a relatives death reminds us of our own
mortality, of that end to which we all must go.
Author
Teresa Gilewski*
Martin Raber*
George W. Sledge Jr.*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Greek Medicine, History of Medicine, U.S. National Library of Medicine,
National Institutes of Health. http://www.nlm.nih.gov/hmd/. Accessed March
7, 2012.
2. Flexner A. Medical education in the United States and Canada: A report
to the Carnegie Foundation for the Advancement of Teaching. New York:
Carnegie Foundation for the Advancement of Teaching, 1910.
3. Cooke M, Irby DM, Sullivan W, et al. American medical education 100
years after the Flexner report. N Engl J Med. 2006;355:1339-1344.
4. American Board of Internal Medicine. Project Professionalism. Philadelphia, Pa, America Board of Internal Medicine, 2001;4.
5. Hewitt M, Herdman R, Holland J (eds). Meeting Psychosocial Needs of
Women with Breast Cancer. Institute of Medicine and National Research
Council. Washington, DC, National Academies Press, 2004.
6. Cohen JJ. Linking professionalism to humanism: What it means, why it
matters. Acad Med. 2007;82:1029-1032.
7. Swick HM. Professionalism and humanism beyond the academic health
center. Acad Med. 2007;82:1022-1028.
8. Whippen DA, Canellos GP. Burnout syndrome in the practice of oncology: Results of a random survey of 1,000 oncologists. J Clin Oncol. 1991;9:
1916-1921.
9. Walshe FMR. Humanism, history, and natural science in medicine.

BMJ. 1950;August;12:379-384.
10. Pickering WG. Kindness, prescribed and natural in medicine. J Medical Ethics. 1997;23:116-118.
11. Young RK. A Piece of My Mind. John Wiley and Sons, Inc. Hoboken,
New Jersey, 2000.
12. Loprinzi CL. A new addition to the J Clin Oncol: The Art of Oncology
When the Tumor is Not the Target. J Clin Oncol. 2000;18:3.
13. Steensma DP. Art of Oncology: New voices wanted. J Clin Oncol.
2011;29:3343-3344.
14. Charon R. Narrative medicine-a model for empathy, reflection, profession, and trust. JAMA. 2001;286:1897-1902.
15. Biro D. One Hundred Days: My Unexpected Journey from Doctor to
Patient. New York: Random House; 2000.
16. Mullan F. Seasons of survival: reflections of a physician with cancer.
N Engl J Med. 1985;313:270-273.
17. Liberman L. I Signed As the Doctor: Memoir of a Cancer Doctor
Surviving Cancer. Port Charlotte, FL: Booklocker.com, Inc.; 2009.
18. Gilewski T. The physician as the patient [video]. New York: Memorial
Sloan Kettering Cancer Center; 2007.
563
ADVANCES IN INFECTION MANAGEMENT IN

PEDIATRIC ONCOLOGY
CHAIR
Andrew Y. Koh, MD
University of Texas Southwestern Medical Center
Dallas, TX
SPEAKERS
Lillian Sung, MD, PhD
The Hospital for Sick Children
Toronto, ON, Canada
Theoklis Zaoutis, MD
Childrens Hospital of Philadelphia
Philadelphia, PA
Prolonged Febrile Neutropenia in the

Pediatric Patient with Cancer
By Andrew Y. Koh, MD
Overview: Infectious diseases continue to be major causes

of morbidity and mortality in pediatric patients with cancer.
Yet not all pediatric patients with cancer with fever and
neutropenia are at equal risk for substantial morbidity or
mortality from infection. Patients at highest risk for developing infectious complications are those with severe and prolonged neutropenia, substantial medical comorbidity, and
hematologic malignancy, or recipients of stem-cell transplantation. These high-risk patients also have concomitant host
immune deficits as well: severe mucositis, lymphopenia, hypogammaglobulinemia, and gut microbial dysbiosis. Because
bacterial and fungal infections are the most common infectious complications, continuation of empirical antibacterial
antibiotics that were initiated at the onset of febrile neutropenia and prompt initiation of empirical antifungal therapy in
LTHOUGH THE advances of infectious diseases supportive care during the last several decades have
permitted patients to successfully recover from the impact of
cytotoxic cancer chemotherapy, infectious diseases continue
to be major causes of morbidity and mortality in pediatric
patients with cancer. This review will address the host
immune deficits in pediatric patients with cancer, riskstratification of patients with febrile neutropenia, general
management guidelines for patients with prolonged neutropenia, and potential strategies for preventing infectious
diseases in patients with prolonged neutropenia.
Host Defenses: Innate and Adaptive Immunity
To best manage or prevent infectious complications in the

patient with cancer with prolonged neutropenia, it is essential to understand the host immune deficits that result from
the underlying disease and/or cancer chemotherapy.
the setting of prolonged fever and neutropenia continue to be

the standard of care. In high-risk patients, antibiotic therapy
should be maintained until neutrophil counts have recovered.
Adjunctive therapies have been shown to be ineffective (e.g.,
colony-stimulating factors) or necessitate further study (e.g.,
granulocyte infusions or keratinocyte growth factor treatment
to heal mucositis). Prophylactic use of antibacterial and
antifungal antibiotics in high-risk patients has shown promise
but the fear of inducing antimicrobial-resistant strains remains a deterrent. Finally, the novel concepts of manipulating
the host gut microbiota and/or augmenting GI mucosal immunity to prevent invasive bacterial and fungal infections in
pediatric patients with cancer offers great promise, but more
definitive studies need to be performed.
The relationship between neutrophil numbers and the risk

of infectious complications in patients with leukemia was
first described in the seminal work of Bodey and colleagues
in 1966.2 The investigators concluded that 1) the risk of
infection was directly related to the absolute neutrophil
count (ANC), severe infections being more prevalent when
the ANC fell below 100 cells/mm3; 2) relapse of leukemia
was associated with higher rates of infection than was
remission; and 3) duration of neutropenia was the single
most important factor in predicting risk of infection, with
severe neutropenia that lasted longer than 3 weeks being
associated with 100% risk of infection and the highest
mortality rates. The rate of decline of circulating polymorphonuclear lymphocytes (PMNs) is also critical. For example, patients with rapidly declining ANCs after therapy for
acute leukemia seem to be at greater risk for infectious
complications than are patients with chronic neutropenia
(e.g., aplastic anemia).3
Mucocutaneous Barriers and Commensal Gut Microbiota
Pediatric oncology patients sustain disruptions of the

mucocutaneous integrity because of their underlying cancer
and its treatment. The disruption of mucocutaneous barriers
provides a key portal of entry for bacterial and fungal
pathogens (Sidebar 1). Colonization by normal bacterial
flora (aerobic gram-positive bacteria and a variety of anaerobic bacteria with relatively low virulence) provides a competitive microbiologic barrier to colonization by extrinsically
acquired bacterial and fungal organisms. However, within
24 hours of hospitalization, seriously ill patients undergo a
change in their indigenous microflora toward one of aerobic
gram-negative organisms.1 Approximately one-half of the
responsible pathogens causing infections are acquired by
oncology patients after initial admission to the hospital, and
more than 80% of the microbiologically documented infections that occur in adult patients with acute myelogenous
leukemia (AML) are caused by organisms that were a
component of the endogenous mucosal organisms.1
Phagocytic Cells
By far the most important risk factor in the development

of infections in children with cancer remains neutropenia.
Risk Assessment in Patients with Cancer with Fever

and Neutropenia
Not all patients with fever and neutropenia are at equal

risk for substantial morbidity or mortality resulting from
infection. The identification of a low-risk subset may allow
for modifications of therapy in this group, with goals of
reduced therapy-related toxicity, improved quality of life,
and decreased cost of treatment. Although there is no
definitive consensus about the criteria used to prospectively
distinguish high risk from low risk, the following key factors
that may raise the risk of infectious complications can be
surmised from studies that have been conducted: (1) anticipated duration of neutropenia4; (2) substantial medical
comorbidity4; (3) cancer status and cancer type; (4) docu-
From the University of Texas Southwestern Medical Center, Dallas, TX.

Address reprint requests to Andrew Koh, MD, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390-9063; email: Andrew.koh@
utsouthwestern.edu.
1092-9118/10/1-10
565
ANDREW Y. KOH
Sidebar 1. Predominant Microbial Pathogens

Infecting Pediatric Patients with Cancer and
Prolonged Neutropenia
Bacteria
Gram-negative enteric organisms
Escherichia coli, Klebsiella pneumoniae, Enterobacter spp., Citrobacter spp., Pseudomonas aeruginosa, Bacteroides spp.
Gram-positive
Staphylococci: coagulase-negative, coagulasepositive
Streptococci: group D, -hemolytic, anaerobic
Clostridia
Fungi
Candida spp. (C. albicans, C. tropicalis, other
species)
Aspergillus spp. (A. fumigatus, A. flavus)
mented infection on presentation; (5) evidence of bone marrow recovery5; and (6) magnitude of fever (Sidebar 2).5
By using these risk criteria, patients with acute lymphoblastic leukemia (ALL) or AML in induction, relapsed ALL
or AML, Burkitts lymphoma in induction, and hematopoietic stem cell transplantation would all be considered highrisk patients. These patients not only have severe and
prolonged neutropenia but also have profound lymphopenia,
hypogammaglobulinemia, severe mucositis, and gut microbial dysbiosis. Thus, these high-risk patients might benefit
from implementation of prophylactic strategies to prevent
infectious complications.
KEY POINTS
566
Infectious diseases, particularly bacterial and fungal

infections, continue to be major causes of morbidity
and mortality in pediatric patients with cancer.
Pediatric patients with cancer at highest risk for
developing infectious complications are those with
severe and prolonged neutropenia; substantial medical comorbidity, and specific cancer types and status
(i.e., hematologic malignancy, relapse).
Besides severe and prolonged neutropenia, highrisk patients also have concomitant host immune
deficits: severe mucositis, lymphopenia, hypogammaglobulinemia, and gut microbial dysbiosis.
Continuation of empirical antibacterial antibiotics
that were initiated at the onset of febrile neutropenia
and prompt initiation of empirical antifungal therapy
in the setting of prolonged fever and neutropenia
continue to be the standard of care for pediatric
patients with cancer with prolonged neutropenia.
Adjunctive therapies (i.e., granulocyte transfusion or
keratinocyte growth factor treatment for mucositis)
and prophylactic use of antibacterial and antifungal
antibiotics in high-risk patients have shown promise,
but more definitive studies need to be done.
Sidebar 2. Risk Assessment for Cancer Patients

with Fever and Neutropenia
High Risk
1) Anticipated prolonged ( 7 days in duration)
and profound neutropenia (absolute neutrophil count 100 cells/L after cytotoxic chemotherapy)4,6
2) Significant medical comorbid conditions, including hypotension, pneumonia, new-onset
abdominal pain, or neurologic changes4,6
3) Cancer type (i.e., hematologic malignancy,
Burkitts lymphoma)
4) Cancer status (e.g., relapse)
5) Documented infection on presentation
Low Risk
1) Anticipated brief ( 7 days in duration) neutropenic periods4
2) No or few comorbidities4
Standard Management and Presumptive Therapy
The single most important advance in infectious diseases

oncology supportive care leading to improved survival has
been the prompt initiation of empirical antibacterial antibiotics when the neutropenic patient with cancer becomes
febrile. Before this approach was instituted in the early
1970s, the mortality rate from gram-negative infections,
especially that of P. aeruginosa, E. coli, and K. pneumoniae,
approached 80%, but with the widespread use of effective
empirical antibiotics, the overall mortality rate has declined
to approximately 10% to 40%.
Approximately 85% to 90% of pathogens that are documented to be associated with new fevers in neutropenia
patients are gram-positive and gram-negative bacteria. Several empirical regimens have been devised over the last
several decades. Many of the regimens studied are equivalent in their efficacy, although study designs and definitions
of success have not been uniform. The Infectious Diseases
Society of America has published general guidelines for use
of antimicrobial agents in neutropenic patients with unexplained fever.6
Duration of Therapy
The management of high-risk patients with prolonged

neutropenia is addressed in a series of prospective clinical
studies that stratified according to those who defervesced
after the initiation of broad-spectrum antibiotics or who
remained persistently febrile.7 Among patients with prolonged neutropenia who defervesced while undergoing
therapy, 41% again became febrile within 3 days of stopping antibiotics on day 7; new bacterial isolates from those
with documented infections were susceptible to the antibiotics that had been withdrawn. No subsequent infections
were observed among patients who continued receiving
antibiotics.
The duration of antibiotic therapy depends on several
factors, including isolation of the presumed pathogen, clinical identification of a presumed infectious process, duration
of both fever and neutropenia, and the patients schedule for
chemotherapy. Traditionally, broad-spectrum antibiotic
FEBRILE NEUTROPENIA IN PEDIATRIC CANCER
therapy is continued until the ANC has returned to 500

cells/mm3. For patients in whom infection has been documented and resolution of fever and neutropenia has been
prompt, the course of antibiotic treatment should be appropriate for the infection identified, parenteral antibiotics
being preferred for patients with serious infections. For
patients without a defined site of infection who show signs of
bone marrow recovery, antibiotics generally can be discontinued, even before the ANC reaches 500/mm3. But even
when a bacterial pathogen is isolated from a febrile, neutropenic child, there is evidence that broad-spectrum antibiotic
therapy should be continued.8
Empirical Antifungal Therapy
Fungi have emerged as an important cause of superinfections in patients with prolonged neutropenia, and may affect
9% to 31% of this population. In a randomized clinical trial,
56% of patients with unexplained fever who remained febrile after receiving empirical antibiotics developed complications within 3 days of stopping therapy.9 Strikingly, 31%
of these patients eventually developed invasive fungal infections. Neutropenic patients who remain febrile despite a 4to 7-day trial of broad-spectrum antimicrobial therapy are
particularly prone to fungal disease.9
Traditionally, amphotericin B was the only available systemic antifungal agent. Its use in empirical regimens for
prolonged or recurrent fever in neutropenic patients reduced
the incidence of documented fungal infections and attributable mortality.9 Considerable interest has been generated
by preparations of liposomal amphotericin because of its
lower toxicity. In the only randomized, double-blind trial
comparing liposomal amphotericin with conventional amphotericin B as empirical antifungal therapy, the outcomes
were similar with respect to survival, resolution of fever, and
discontinuation of study drug because of toxic effects or lack
of efficacy.10 Liposomal amphotericin B was associated with
fewer breakthrough fungal infections, less infusion-related
toxicity, and less nephrotoxicity.10
The search for alternative antifungal agents has been
prompted by the potential toxicity of amphotericin B and
emergence of fungi resistant to it. The azoles represent a
less toxic class of antifungal agents. Although fluconazole
has been reported to be as effective as amphotericin in the
treatment of candidemia in patients without neutropenia or
other major immunodeficiency condition,11 the picture is
less clear in febrile and neutropenic patients with cancer.
The azoles, however, may be less active than amphotericin B
against some species. Voriconazole compared favorably with
liposomal amphotericin B in adult patients with cancer with
fever and neutropenia.12 In 2006, the U.S. Food and Drug
Administration (FDA) approved the use of posaconazole for
prophylaxis against the development of invasive Aspergillus
and Candida infections in immunocompromised patients 13
years of age and older. Posaconazole is distinct in having
been successfully used in salvage treatment of infections
caused by zygomycetes. But posaconazoles efficacy as an
empirical antifungal agent in febrile and neutropenic patients with cancer has not been evaluated.
The newest class of antifungal agents are the echinocandins: large lipopeptide molecules that are inhibitors of
-(1,3)-glucan synthesis, which is essential for fungal cell
wall synthesis. Both in vitro and in vivo, the echinocandins
are rapidly fungicidal against most Candida spp and fungi-
static against Aspergillus spp., but they are not active

against Zygomycetes, Cryptococcus neoformans, or Fusarium
spp. Because the drug target is not present in mammalian
cells, adverse events are generally mild, including (for caspofungin) local phlebitis, fever, abnormal liver function tests,
and mild hemolysis. Oral bioavailability is suboptimal,
thereby limiting use to the intravenous route. In a prospective randomized, double-blind trial comparing the efficacy
and safety of caspofungin with that of liposomal amphotericin B, caspofungin was found to be as effective and generally better tolerated than liposomal amphotericin B when
administered as empirical antifungal therapy in patients
with persistent fever and neutropenia.13 The newer echinocandins, micafungin and anidulafungin, also appear to be
well-tolerated in pediatric oncology patients.
Thus, amphotericin B, voriconazole, and caspofungin have
been well characterized for empirical antifungal therapy for
persistent fever in high-risk neutropenic patients. For patients who remain neutropenic, antifungal therapy should
be continued until the resolution of neutropenia. Persistence
or recrudescence of fever should prompt a meticulous investigation for nonfungal infectious causes (e.g., bacterial or
viral superinfections) or for a fungus that is resistant to
initial empirical antifungal coverage.
Recombinant Human Colony-Stimulating Factors
The use of hematopoietic growth factors, such as

granulocyte-macrophage colony-stimulating factor (GMCSF) and granulocyte colony-stimulating factor (G-CSF),
has been implemented widely in the management of neutropenia in patients who have undergone cancer and bone
marrow transplantation. Both G-CSF and GM-CSF have
been evaluated as adjuncts to chemotherapy to assist bone
marrow reconstitution and to prevent neutropenia and reduce infectious complications. Primary use of G-CSF has
been shown to lower the incidence of febrile neutropenic
episodes by approximately 50% in three randomized studies
in which the incidence of fever and neutropenia in the
control group was greater than 40%. Primary administration of CSFs should be reserved for patients who are expected to experience rates of fever and neutropenia ( 40%)
that are comparable with or greater than those seen in the
control patients in these randomized trials.
Use of GM-CSF has been less consistently helpful and has
been associated with more adverse effects than those of
G-CSF. No study to date has demonstrated an advantage in
rates of tumor response, fatal infections, or overall survival.
Secondary use of G-CSF or GM-CSF in subsequent cycles of
chemotherapy has not demonstrated disease-free or overall
survival benefits when the dose of chemotherapy was maintained. Therefore, reduction in chemotherapy dose should be
considered the primary therapeutic option in a patient who
experiences neutropenic fever or severe or prolonged neutropenia after the previous cycle of treatment. Finally, if G-CSF
is administered within the first few days of chemotherapy
for the initial induction or first postremission course for
acute lymphoblastic leukemia (ALL), the duration of neutropenia of less than 1,000/mm3 can be shortened by approximately 1 week. These studies and guidelines for the use of
hematopoietic growth factors are summarized in a consensus paper by the American Society of Clinical Oncology.14
Clinical trials have reported conflicting results when attempting to evaluate whether the addition of CSFs to
567
ANDREW Y. KOH
antibiotics in the treatment of febrile neutropenia improves

patient outcomes. A meta-analysis of 13 studies revealed
that the use of CSFs in patients with established fever and
neutropenia reduces the amount of time spent in the hospital and the neutrophil recovery period. Overall mortality
was not influenced by use of CSFs, but a marginally significant decrease in infection-related mortality was noted.15
Granulocyte Transfusions
Although the principal of granulocyte transfusions in

neutropenic patients with refractory infections is physiologically sound, the data supporting this clinical practice have
been the subject of considerable controversy. Some early
randomized studies demonstrated a clinical benefit for the
PMN transfusion group compared with controls,16 whereas
others showed no overall benefit although demonstrating
efficacy in certain subgroups of patients,17 and still others
reported no benefit at all.18 Furthermore, the administration of granulocyte transfusions carries risks of alloimmunization and respiratory distress.
Recent technical advances in the ability to collect substantially larger numbers of PMNs per pheresis have opened
new potential for the therapeutic benefit of this adjunctive
modality. The use of G-CSF to mobilize granulocytes in
normal healthy donors has been shown to be safe and
effective, allowing for the collection of substantially more
PMNs per cycle of pheresis.19 In addition to total number of
cells transfused, human leukocyte antigen (HLA) match is
also thought to be an important factor in the efficacy of
granulocyte transfusions. Given the technical advances in
G-CSF mobilization, WBC collection, and donor matching, a
randomized trial is warranted to investigate the use of
granulocyte transfusions in neutropenic patients with refractory infections. However, the challenges posed by standardization of collection methods, HLA matching, dose of
granulocytes, and type of patients enrolled are daunting to a
multicenter trial. Until such studies are available, selection
of candidate patients to receive granulocyte transfusions
must be made by an individual assessment of risk and
benefit.
Prophylactic Antimicrobial Therapy
Many clinical trials have focused on use of prophylactic

antibiotics to prevent infections in neutropenic patients. The
oral fluoroquinolone antibiotics have been investigated because of their good bioavailability and broad activity against
aerobic bacteria. A meta-analysis of published randomized
controlled trials of quinolone prophylaxis (18 trials with
1,408 patients) found that quinolone prophylaxis substantially reduced the incidence of gram-negative bacterial infections, microbiologically documented infections, total
infections, and fevers, but did not alter the incidence of
gram-positive infections or infection-related deaths.20 Two
recent distinct studies evaluating the use of prophylactic
oral levofloxacin (500 mg daily) in patients receiving chemotherapy for either solid tumors or lymphoma21 or hematologic malignancies22 both showed a reduction in documented
infections. Finally, a recent meta-analysis of antibiotic prophylaxis in neutropenic patients with cancer (95 trials
performed between 1973 and 2004) concluded that antibiotic
prophylaxis (various antibiotic regimens) substantially decreased the risk for death compared with placebo or no
568
23
treatment. When trials that used quinolone prophylaxis

(52 trials) were separately analyzed, there was a substantial
reduction in the risk for all-cause mortality, as well as
infection-related mortality, fever, clinically documented infections, and microbiologically documented infections. Although these results are very encouraging, many of these
studies reported increasing rates of antimicrobial resistance
to quinolones. Thus, if quinolone prophylaxis is implemented, vigilant monitoring of the incidence of bacteremia
(specifically gram-negative bacteremia) is mandatory to first
detect the loss of efficacy of fluoroquinolone prophylaxis.
The use of fluconazole as antifungal prophylaxis has been
conducted in adult patients with leukemia who have undergone bone marrow transplantation.24 Although a decrease in
fungal colonization and superficial infections was noted, a
reduction in systemic fungal infections and associated mortality was identified only in the patients undergoing bone
marrow transplantation.
Augmentation of Mucocutaneous Barriers and
Commensal Gut Microbiota
Keratinocyte growth factor, a member of the family of

fibroblast growth factors, has been shown to reduced the
duration and severity of oral mucositis after intensive chemotherapy.25 A preliminary analysis of blood-borne infections showed a lower incidence among patients receiving
palifermin than among those receiving placebo. Future
investigations, however, are merited to assess the ability of
palifermin therapy to decrease the complications caused by
systemic infection.
Recent studies have shown that the GI microbiota plays a
crucial role in preventing overgrowth of pathogenic microbes by directly inhibiting the growth of specific pathogens and/or by stimulation of GI mucosal effectors (e.g.,
antimicrobial proteins) that act to clear invading pathogens.
Although probiotic bacterial strains have been used to treat
inflammatory bowel disease, necrotizing enterocolitis, and
enteric infections human patients, no studies have been
performed in pediatric patients with cancer. The major
concern is that introduction of enteric bacteria may result in
dissemination of the introduced strain. The importance of
the microbiota in driving protective immune responses during GI infection is best illustrated by the finding that
restoring signaling through innate immune receptors in
antibiotic-treated mice can protect from intestinal infections
(i.e., lipopolysaccharide activation of TLR4 or flagellin stimulation of TLR5 to prevent vancomycin-resistant Enterococci
dissemination.).26
Conclusion
Pediatric patients with cancer with prolonged neutropenia

have increased risk for severe, recurrent, or new bacterial
and fungal infections. Although prompt initiation of empirical antibacterial antibiotics when the neutropenic patient
with cancer becomes febrile has lead to substantial improvement in morbidity and mortality associated with bacterial
and fungal infections, infectious complications still persist.
Currently, appropriate administration of antibacterial and
antifungal antibiotics remains standard of care. Adjunctive
use of agents that promote healing of mucosal damage, use
of probiotics or prebiotics to reestablish gut microbiota
homeostasis, and granulocyte infusions offer promise, but
more definitive studies are necessary.
FEBRILE NEUTROPENIA IN PEDIATRIC CANCER
Author
Andrew Y. Koh*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Schimpff SC, Young VM, Greene WH, et al. Origin of infection in acute
nonlymphocytic leukemia. Significance of hospital acquisition of potential
pathogens. Ann Intern Med. 1972;77:707-714.
2. Bodey GP, Buckley M, Sathe YS, et al. Quantitative relationships
between circulating leukocytes and infection in patients with acute leukemia.
Ann Intern Med. 1966;64:328-340.
3. Pizzo PA. After empiric therapy: what to do until the granulocyte comes
back. Rev Infect Dis. 1987;9:214-219.
4. Talcott JA, Siegel RD, Finberg R, et al. Risk assessment in cancer
patients with fever and neutropenia: a prospective, two-center validation of a
prediction rule. J Clin Oncol. 1992;10:316-322.
5. Rackoff WR, Gonin R, Robinson C, et al. Predicting the risk of bacteremia in childen with fever and neutropenia. J Clin Oncol. 1996;14:919-924.
6. Freifeld AG, Bow EJ, Sepkowitz KA, et al. Clinical practice guideline for
the use of antimicrobial agents in neutropenic patients with cancer: 2010
Update by the Infectious Diseases Society of America. Clin Infect Dis.
52:427-431.
7. Pizzo PA, Robichaud KJ, Gill FA, et al. Duration of empiric antibiotic
therapy in granulocytopenic patients with cancer. Am J Med. 1979;67:194200.
8. Pizzo PA, Ladisch S, Ribichaud K. Treatment of gram-positive septicemia in cancer patients. Cancer. 1980;45:206-207.
9. Pizzo PA, Robichaud KJ, Gill FA, et al. Empiric antibiotic and antifungal
therapy for cancer patients with prolonged fever and granulocytopenia. Am J
Med. 1982;72:101-111.
10. Walsh TJ, Finberg RW, Arndt C, et al. Liposomal amphotericin B for
empirical therapy in patients with persistent fever and neutropenia. National
Institute of Allergy and Infectious Diseases Mycoses Study Group. N Engl
J Med. 1999;340:764-771.
11. Rex JH, Bennett JE, Sugar AM, et al. A randomized trial comparing
fluconazole with amphotericin B for the treatment of candidemia in patients
without neutropenia. Candidemia Study Group and the National Institute.
N Engl J Med. 1994;331:1325-1330.
12. Walsh TJ, Pappas P, Winston DJ, et al. Voriconazole compared with
liposomal amphotericin B for empirical antifungal therapy in patients with
neutropenia and persistent fever. N Engl J Med. 2002;346:225-234.
13. Walsh TJ, Teppler H, Donowitz GR, et al. Caspofungin versus liposomal amphotericin B for empirical antifungal therapy in patients with
persistent fever and neutropenia. N Engl J Med. 2004;351:1391-1402.
14. Smith TJ, Khatcheressian J, Lyman GH, et al. 2006 update of recommendations for the use of white blood cell growth factors: an evidence-based
clinical practice guideline. J Clin Oncol. 2006;24:3187-3205.
15. Clark OA, Lyman GH, Castro AA, et al. Colony-stimulating factors for
chemotherapy-induced febrile neutropenia: a meta-analysis of randomized
controlled trials. J Clin Oncol. 2005;23:4198-4214.
16. Herzig RH, Herzig GP, Graw RG, Jr., et al. Successful granulocyte
transfusion therapy for gram-negative septicemia: a prospectively randomized controlled study. N Engl J Med. 1977;296:701-705.
17. Alavi JB, Root RK, Djerassi I, et al. A randomized clinical trial of
granulocyte transfusions for infection in acute leukemia. N Engl J Med.
1977;296:706-711.
18. Winston DJ, Ho WG, Gale RP. Therapeutic granulocyte transfusions
for documented infections. A controlled trial in ninety-five infectious granulocytopenic episodes. Ann Intern Med. 1982;97:509-515.
19. Bensinger WI, Price TH, Dale DC, et al. The effects of daily recombinant human granulocyte colony-stimulating factor administration on normal
granulocyte donors undergoing leukapheresis. Blood. 1993;81:1883-1888.
20. Engels EA, Lau J, Barza M. Efficacy of quinolone prophylaxis in
neutropenic cancer patients: a meta-analysis. J Clin Oncol. 1998;16:11791187.
21. Cullen M, Steven N, Billingham L, et al. Antibacterial prophylaxis after
chemotherapy for solid tumors and lymphomas. N Engl J Med. 2005;353:988998.
22. Reuter S, Kern WV, Sigge A, et al. Impact of fluoroquinolone prophylaxis on reduced infection-related mortality among patients with neutropenia
and hematologic malignancies. Clin Infect Dis. 2005;40:1087-1093.
23. Gafter-Gvili A, Fraser A, Paul M, et al. Meta-analysis: Antibiotic
prophylaxis reduces mortality in neutropenic patients. Ann Intern Med.
2005;142:979-995.
24. Goodman JL, Winston DJ, Greenfield RA, et al. A controlled trial of
fluconazole to prevent fungal infections in patients undergoing bone marrow
transplantation. N Engl J Med. 1992;326:845-851.
25. Spielberger R, Stiff P, Bensinger W, et al. Palifermin for oral mucositis
after intensive therapy for hematologic cancers. N Engl J Med. 2004;351:
2590-2598.
26. Kinnebrew MA, Ubeda C, Zenewicz LA, et al. Bacterial flagellin
stimulates Toll-like receptor 5-dependent defense against vancomycinresistant Enterococcus infection. J Infect Dis. 2010;201:534-543.
569
Initial Management of Low-Risk Pediatric

F e v e r a n d N e u t r o p e n i a : E f fi c a c y a n d
Safety, Costs, Quality-of-Life Considerations,
and Preferences
By Lillian Sung, MD, PhD
Overview: Initial management options for pediatric low-risk

fever and neutropenia (FN) include outpatient compared with
inpatient management and oral compared with intravenous
therapy. Single-arm and randomized trials have been conducted in children. Meta-analyses provide support for the
equivalence of outpatient and inpatient approaches. Outpatient oral management may be associated with a higher risk of
readmission compared with outpatient intravenous management in children with FN, although other outcomes such as
treatment failure and discontinuation of the regimen because
of adverse effects were similar. Importantly, there have been
no reported deaths among low-risk children treated as outpatients or with oral antibiotics. Costs, whether derived directly
or through cost-effectiveness analysis, are consistently reduced when an outpatient approach is used. Quality of life
EVER AND neutropenia (FN) is a common and potentially fatal complication of intensive chemotherapy in
children with cancer.1 Over the last 4 decades, outcomes of
FN have improved dramatically related to hospitalization
and prompt initiation of empiric antibiotic therapy.2,3 However, children with FN are heterogeneous.4 Based on characteristics at presentation, some of these children are
predicted to be at low risk of mortality and adverse events,
and these children may be managed less intensively.5 Currently, there are several validated prediction rules designed
to identify the low-risk child with FN.6 These rules vary in
terms of their specific characteristics, but all rules identify
children with an anticipated short duration of neutropenia.
The ability to identify such patients has led to the consideration and design of clinical trials to evaluate less intensive
strategies for low-risk FN and, ultimately, to their implementation in clinical practice.
The most relevant lesser intensity strategies relate to site
of care (outpatient compared with inpatient) and mode of
antibiotic administration (oral compared with intravenous).
In adult patients with low-risk FN, oral therapy is associated with similar outcomes to intravenous treatment,7-9 and
there is increasing evidence that outpatient management of
this population is a safe approach.10,11 As a result, adultbased guidelines for the management of FN now recommend
outpatient management and oral antibiotics for selected
low-risk patients.12,13 However, there has been reluctance to
adopt these strategies for children with FN.14 The following
sections summarize the existing literature related to outpatient management and oral antibiotic therapy in children
with FN with respect to efficacy and safety, costs, and QoL
and preferences.
Efficacy and Safety of Outpatient Management and
Oral Antibiotic Administration
Outpatient Compared with Inpatient Management
Successful outpatient management of children with FN is

predicated on several important considerations related to
570
(QoL) and preferences should be considered in order to

evaluate different strategies, plan programs, and anticipate
uptake of outpatient programs. Using parent-proxy report,
child QoL is consistently higher with outpatient approaches,
although research evaluating child self-report is limited. Preferences incorporate estimated QoL, but, in addition, factor in
issues such as costs, fear, anxiety, and logistical issues. Only
approximately 50% of parents prefer outpatient management.
Future research should develop tools to facilitate outpatient
care and to measure caregiver burden associated with this
strategy. Additional work should also focus on eliciting child
preferences for outpatient management. Finally, studies of
effectiveness of an ambulatory approach in the real-world
setting outside of clinical trials are important.
the child, family, and local infrastructure. First, the child

must be identified as having low-risk features using one of
the validated clinical prediction rules.6 Family considerations include a history of compliance with other medical
procedures, rapid accessibility to hospital, ability to communicate reliably by telephone, and ability to provide continual
observation and assessment of the child. The local health
care team must also have the infrastructure to support
outpatient management, which includes clinics that can
assess outpatients and professionals who can be in contact
with families on a regular basis and who are readily available if problems arise.
Outpatient therapy can be initiated at the onset of FN
or after a short period of inpatient treatment followed by
discharge to the home (step-down management). Outpatient
programs may deliver antibiotics by different routes of
administration, including entirely oral administration, entirely intravenous administration, or step-down management in which treatment begins with intravenous and then
transitions to oral therapy. In the outpatient setting,
follow-up assessments may occur in clinics, in the home, by
telephone, or a combination of approaches. The frequency of
follow-up assessments may vary considerably depending on
the specific outpatient model of care.
There has been a single meta-analysis of randomized
controlled trials (RCTs) that compared outpatient compared
with inpatient management of FN.15 Six studies were included: four were adult studies and two were pediatric
studies. Two studies consisted of an entirely outpatient
approach whereas four consisted of a step-down approach in
From the Division of Haematology/Oncology, and Program in Child Health Evaluative

Sciences, The Hospital for Sick Children, Toronto, ON, Canada.
Address reprint requests to Lillian Sung MD, PhD, Division of Haematology/Oncology,
The Hospital for Sick Children, 555 University Avenue, Toronto, ON M5G1X8; email:
lillian.sung@sickkids.ca
1092-9118/10/1-10
TREATMENT APPROACHES IN CHILDREN
which the patient was admitted to hospital and then discharged early while still receiving empiric antibiotics for FN.
Outpatient management was associated with a similar risk
of treatment failure compared with inpatient management
(rate ratio [RR] 0.81, 95% CI 0.55 to 1.28, p 0.28) where
RR 1 favored inpatient care. In interpreting this RR, it is
important to realize that failure was biased against outpatient care because readmission, a criterion for failure, is only
applicable to outpatients. There was no difference in mortality (RR 1.11, 95% CI 0.41 to 3.05, p 0.83). In a stratified
analysis of the two pediatric studies,16,17 results were similar to the overall analysis.
Although this review provides evidence for the safety of
outpatient management, it includes only two studies in
children, and, in total, 278 pediatric subjects were studied.
Consequently, this analysis may have had insufficient power
to evaluate outpatient care in pediatric FN. Thus, we subsequently conducted a systematic review in which we evaluated data from all prospective trials in pediatric FN that
studied a homogeneous, initial antibiotic regimen.18 There
were 16 trials in which either outpatient or inpatient management were established within the first 24 hours of
treatment-initiation in low-risk FN. There was no increase
in treatment failure (including modification of antibiotics)
with outpatient management in comparison with inpatient
management (15% compared with 27%, p 0.04). The rate
of adverse events leading to antibiotic discontinuation was
similar (1% compared with 2%, p 0.39). Among the 953
outpatients, there were no infection-related deaths.
In summary, when combining data from prospective observational and randomized trials in pediatric patients,
similar outcomes were observed between those treated as
outpatients or inpatients. However, outpatient management
should only be instituted if the child can be confidently
classified as low-risk, and if the social circumstances and
local infrastructure support ambulatory management.
Consequently, if the child, family, and health care facility
characteristics support outpatient care, then ambulatory
management may be offered. Outpatient FN programs in
KEY POINTS
Prospective single-arm and randomized trials have

provided evidence for the efficacy and safety of outpatient management and oral antibiotic administration for low-risk children with fever and neutropenia.
Outpatient oral management may be associated with
a higher risk of readmission in children, although
other outcomes such as treatment failure and discontinuation of the regimen because of adverse effects
are similar.
Outpatient management is cost-saving when compared to an inpatient or an early discharge strategy.
Anticipated child quality of life as reported by parents is higher with outpatient strategies.
When comparing inpatient with outpatient management of low-risk fever and neutropenia, the most
preferred option is variable, and frequently, parents
and children will prefer inpatient care.
children vary considerably in terms of frequency of follow up

and nature of follow up (for example, telephone contact
compared with a clinic visit). Future work should consider
minimal and optimal approaches to follow outpatient children with FN.
Oral versus Parenteral Antibiotic Administration
Regardless of whether low-risk children with FN are

treated in the outpatient or inpatient setting, another question relates to the optimal mode of antibiotic administration.
Oral antibiotic administration may be desirable because it
facilitates outpatient management, is more convenient, and
is usually less expensive compared with intravenous antibiotic administration. However, there are several unique issues with oral medication administration in young children.
Issues to consider include the likelihood that the child can
and will accept oral therapy in available formulations given
the childs level of cooperation. First, very young children
may not be able to swallow pills or capsules, and, thus, oral
antibiotics are only feasible in this age group if the medication is available in a liquid formulation. Second, taste of oral
medication is a much more important issue in young children compared to adults, as an unpalatable drug may be
refused. Third, children may refuse all oral medications
regardless of taste, particularly if they feel unwell. Furthermore, the likelihood that the child will accept oral therapy
may change if nausea, vomiting, or mucositis are present.
Oral antibiotic administration may be initiated at the onset
of the FN episode or following a short period of parenteral
administration before transitioning to oral administration
(step-down management).
Two meta-analyses of RCTs evaluated oral and intravenous antibiotic administration for FN. One study of 2,770
patients included both outpatients and inpatients7 whereas
the second study of 1,595 patients was restricted to outpatients.15 Neither study restricted their review to low-risk FN
patients. Both reviews showed similar results with no difference in treatment failure (including modification), overall
mortality, or adverse effects of antibiotics. These findings
were demonstrated both among combined adult and pediatric analyses, as well as in stratified analyses of pediatric
studies alone. However, a stratified analysis of five pediatric
RCTs demonstrated that intravenous outpatient management was associated with a lower rate of readmission
compared with oral outpatient management (RR 0.52, 95%
CI 0.24 to 1.09, p 0.08).15
More information about the efficacy and safety of oral
antibiotic administration was derived from an analysis of
prospective pediatric trial data comparing oral and parenteral antibiotic therapy initiated within 24 hours of treatment initiation in low-risk FN.18 Oral antibiotics used in
these studies were fluoroquinolone monotherapy (7 studies
with 581 patients), fluoroquinolone and amoxicillinclavulanate (3 studies with 159 patients), and cefixime
(1 study with 45 patients). The review found no difference in
treatment failure (including modification) among children
who received oral compared with intravenous antibiotic
therapy (20% compared with 22%, p 0.68). The rate of
antibiotic discontinuation because of adverse events was
similar between the oral and intravenous regimens (2%
compared with 1%, p 0.73). There were no infectionrelated deaths among the 676 children given oral antibiotics.
To summarize, there were more readmissions among chil-
571
LILLIAN SUNG
dren treated with oral therapy compared with intravenous

therapy in the outpatient setting, although other outcomes,
including treatment failure and adverse events, were similar; and there were no infection-related deaths among lowrisk pediatric patients.
Consequently, if oral antibiotics can confidently be administered, oral antibiotic therapy may be used in selected
low-risk children with FN, either in the outpatient or inpatient settings. One approach to outpatient oral antibiotic
administration may be to administer one dose of oral antibiotics during the initial health care encounter and only
discharge home with oral therapy if that dose can be
ingested. Costs, feasibility, and patient/family preferences
should influence the route of antibiotic administration for
children with low-risk FN in both the outpatient and inpatient settings.
Costs
The adult literature has consistently demonstrated that

outpatient management is substantially cheaper than inpatient management of FN, primarily related to the costs of
hospitalization.19-21 Limited data are available on costs in
the pediatric setting.16,17 However, the literature supports a
similar cost-saving associated with ambulatory management in children.22,23
A pediatric cost-utility model has been created that includes event probabilities, costs, and QoL estimates.24 Outpatient management was the most cost-effective approach,
with outpatient intravenous antibiotic administration being
more cost-effective than outpatient oral antibiotic administration because of the higher rate of readmission among
those who receive oral antibiotics and better QoL seen
with intravenous therapy (see below). However, within the
plausible range of some inputs into the model, outpatient
management with oral therapy could become the most
cost-effective strategy. Entire treatment in hospital with
intravenous antibiotics was the least cost-effective strategy;
it was more expensive and less effective than an early
discharge approach.
These data suggest that from a cost perspective, outpatient management with either intravenous or oral antibiotics should be used if the infrastructure is in place to support
this model of care, and, if the child, family, and system
factors permit an ambulatory approach.
QoL and Preferences
QoL considerations and preferences become more important when strategies such as outpatient management of FN
have advantages and trade-offs. The advantages of outpatient therapy include substantial cost-saving for the health
care system and reduction in nosocomial infections. However, the trade-offs include increased responsibility and
burden for parents and other care providers, the need to
monitor children remotely from the health care center, and
the need for follow-up visits. It is important to evaluate QoL
and preferences for several reasons. First, this knowledge
allows strategies to be compared using approaches such as
cost-effectiveness analyses. Second, this information can
help to develop outpatient programs and may influence how
programs are structured. In other words, this information
may be used to determine whether or which outpatient FN
programs might be successful in a given context or practice
setting.
572
QoL and preferences are two distinct constructs, although

QoL may influence preferences. QoL is complex in pediatric
medicine because QoL considerations may include the index
child, siblings, and parents/providers. Outpatient management of children with FN may be associated with improved
QoL for children25 although the effect on parents is unclear.26 In a cross-sectional study in which we interviewed
parents of children with cancer and health care providers
who care for these children, we asked respondents to compare just two options, namely oral outpatient management
compared with inpatient intravenous management.26 Both
parents and health care professionals believed child QoL
would be better at home compared with the hospital setting.
However, health care professionals, when compared to parents, overestimated QoL for children at home and underestimated QoL for parents in the hospital setting.
In order to gain more insight, we asked parents of children
with cancer and the children themselves to anticipate child
QoL in four different scenarios, namely outpatient oral
therapy, outpatient intravenous therapy, early discharge,
and inpatient intravenous management.27 The outpatient
regimen in these studies always included three times weekly
clinic visits as the standard scenario because almost all
published studies of outpatient management in pediatric
cancer include at least this frequency of clinic visits. Among
parent respondents, early discharge and outpatient management with intravenous antibiotics were associated with
higher anticipated child QoL (score of 5.9 each on a 10-point
visual analog scale), whereas outpatient oral therapy was
associated with the lowest anticipated child QoL (score of
4.7). In other words, parents were concerned about the effect
of oral medication administration on the childs QoL. For
child self-respondents, all of whom were 12 years or older,
early discharge was anticipated to be associated with the
highest child QoL. It is important to note that there are very
little data about child self-report of QoL in different FN
health states.
Preference will likely include QoL considerations but
also may include factors such as fear and anxiety, costs,
and logistics. We have taken several approaches to the
elicitation of preferences. The simplest approach is one in
which we have asked parents and children their preferred
management strategy. Using this approach, consistently,
approximately 50% of parents preferred outpatient management compared with inpatient management.26,27 When
asked to rank the most preferred strategy among the four
options, we found that the most common top-ranked choice
was inpatient intravenous management among both parents
and children.27
We have also used other techniques to gain insight into
parent decision making in the context of management options for low-risk FN. More specifically, we evaluated preferences using a threshold technique and conjoint analysis.
With the threshold technique, one measures strength of
preference for a treatment option by systematically changing one attribute of a treatment option (for example, increasing or decreasing the number of clinic visits per week
associated with outpatient care) until the respondent gives
up their initially preferred option. Using this technique, we
described strength of preferences for inpatient parenteral
compared with outpatient oral management.26 We found
that, in general, in order to accept outpatient oral management, parents would not tolerate clinic visits more than
TREATMENT APPROACHES IN CHILDREN
three times per week or a rate of readmission of more than

15%. We also found that stronger preference for outpatient
therapy was associated with higher anticipated QoL for the
parent and child at home relative to hospital.26
We also used a conjoint analysis technique to assess
preferences and decision making for outpatient management of low-risk FN.28 Conjoint analysis is an emerging
approach to the measurement of preferences in the face of
multiple trade-offs in health care. In contrast to the threshold technique, conjoint analysis allows evaluation of multiple attributes concurrently or conjointly.29 Using this
technique, we quantified the relative importance of attributes associated with two treatment options outpatient
oral compared with inpatient intravenous management.
Parents would be willing to accept only 2.1 (95% CI 1.1 to
3.2) clinic visits weekly in order to accept outpatient management. With clinic visits three times weekly and a 7.5%
chance of readmission, the probability of parents accepting
an outpatient approach was only 43% (95% CI 39% to
48%).28
Finally, we used a qualitative approach to better understand parental perspectives and how this can influence
preferences.30 The major themes identified when choosing
between outpatient oral and inpatient intravenous therapy
included convenience/disruptiveness for the family, concerns
related to physical health of the child, emotional well-being
for the child, and modifiers of parental decision making.
Thus, we demonstrated that many parents and children
prefer inpatient management, although anticipated QoL on
the aggregate level was higher with early discharge and
outpatient parenteral strategies. It is relatively straightforward to explain these differences. Preferences for a treatment option in this context incorporate considerations other
than child QoL, such as parent QoL, convenience, costs,
safety, and anxiety.
Implementation of Outpatient FN
We have used these findings to develop an outpatient

low-risk FN program in our own health care setting. We
have selected a very low-risk population, based on previ-
ously validated clinical prediction rules, to ensure that the

risk of readmission is very low and to build confidence in the
outpatient program among families and health care professionals. We administer empiric antibiotics orally because we
have found it difficult to initiate empiric antibiotics intravenously in a timely fashion. We administer a single dose of
oral antibiotic in the emergency room or clinic to ensure that
the child can ingest the oral antibiotic to be used in the
outpatient setting before discharge to home. The current
structure includes clinic visits two or three times weekly. A
system of home visits by nursing staff has not shown to be
feasible, and, consequently, the patients are monitored between clinic visits with daily phone calls by a health care
professional. There are plans for evaluation of the program
on a regular basis, although it is too soon to provide such a
report.
Conclusion
In summary, a series of prospective observational studies

and RCTs have been conducted in pediatrics that support
the efficacy and safety of outpatient care and oral antibiotic
administration as initial treatment for children with lowrisk FN. Costs are clearly lower with an ambulatory approach. However, QoL is more complicated because child
and parent QoL considerations are important, and incremental QoL at home compared with the hospital may not be
the same for both respondent types. Preferences are also
important to evaluate because this information may be used
to plan outpatient programs and to anticipate uptake of
these programs.
Future work may include the development of tools to ease
care in the outpatient setting and to measure caregiver
burden associated with this therapy. Additional work should
also focus on eliciting child QoL and preferences for outpatient management of low-risk FN given that these may differ
substantially from parent-proxy responses.31,32 Finally, the
study of the effectiveness of an ambulatory approach in the
real-world setting outside of clinical trials is important to
fully understand the effect of different management strategies for initial treatment of low-risk pediatric FN.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Lillian Sung*
REFERENCES
1. Bodey GP, Buckley M, Sathe YS, et al. Quantitative relationships
between circulating leukocytes and infection in patients with acute leukemia.
Ann Intern Med. 1966;64:328-340.
2. Schimpff S, Satterlee W, Young VM, et al. Empiric therapy with
carbenicillin and gentamicin for febrile patients with cancer and granulocytopenia. N Engl J Med. 1971;284:1061-1065.
3. Pizzo PA. Management of fever in patients with cancer and treatmentinduced neutropenia. N Engl J Med. 1993;328:1323-1332.
4. Rondinelli PI, Ribeiro Kde C, de Camargo B. A proposed score for
predicting severe infection complications in children with chemotherapyinduced febrile neutropenia. J Pediatr Hematol Oncol. 2006;28:665-670.
5. Ammann RA, Bodmer N, Hirt A, et al. Predicting adverse events in
children with fever and chemotherapy-induced neutropenia: the prospective
multicenter SPOG 2003 FN study. J Clin Oncol. 2010;28:2008-2014.
6. Phillips B, Wade R, Stewart LA, et al. Systematic review and meta-
analysis of the discriminatory performance of risk prediction rules in febrile

neutropaenic episodes in children and young people. Eur J Cancer. 2010;46:
2950-2964.
7. Vidal L, Paul M, Ben-Dor I, et al. Oral versus intravenous antibiotic
treatment for febrile neutropenia in cancer patients. Cochrane Database Syst
Rev. 2004;4:CD003992.
8. Freifeld A, Marchigiani D, Walsh T, et al. A double-blind comparison of
empirical oral and intravenous antibiotic therapy for low-risk febrile patients
with neutropenia during cancer chemotherapy. N Engl J Med. 1999;341:305311.
9. Kern WV, Cometta A, De Bock R, et al.Oral versus intravenous empirical antimicrobial therapy for fever in patients with granulocytopenia who are
receiving cancer chemotherapy. International Antimicrobial Therapy Cooperative Group of the European Organization for Research and Treatment of
Cancer. N Engl J Med. 1999;341:312-318.
573
LILLIAN SUNG
10. Carstensen M, Sorensen JB. Outpatient management of febrile neutropenia: time to revise the present treatment strategy. J Support Oncol.
2008;6:199-208.
11. Talcott JA, Yeap BY, Clark JA, et al. Safety of early discharge for
low-risk patients with febrile neutropenia: a multicenter randomized controlled trial. J Clin Oncol. 2011;29:3977-3983.
12. Klastersky J, Paesmans M, Rubenstein EB, et al. The Multinational
Association for Supportive Care in Cancer risk index: A multinational scoring
system for identifying low-risk febrile neutropenic cancer patients. J Clin
Oncol. 2000;18:3038-3051.
13. Freifeld AG, Bow EJ, Sepkowitz KA, et al. Clinical practice guideline
for the use of antimicrobial agents in neutropenic patients with cancer: 2010
update by the Infectious Diseases Society of America. Clin Infect Dis.
2011;52:e56-93.
14. Ammann RA, Simon A, de Bont ES. Low risk episodes of fever and
neutropenia in pediatric oncology: Is outpatient oral antibiotic therapy the
new gold standard of care? Pediatr Blood Cancer. 2005;45:244-247.
15. Teuffel O, Ethier MC, Alibhai SM, et al. Outpatient management of
cancer patients with febrile neutropenia: a systematic review and metaanalysis. Ann Oncol. 2011;22:2358-2365.
16. Ahmed N, El-Mahallawy HA, Ahmed IA, et al. Early hospital discharge
versus continued hospitalization in febrile pediatric cancer patients with
prolonged neutropenia: A randomized, prospective study. Pediatr Blood
Cancer. 2007;49:786-792.
17. Santolaya ME, Alvarez AM, Aviles CL, et al. Early hospital discharge
followed by outpatient management versus continued hospitalization of
children with cancer, fever, and neutropenia at low risk for invasive bacterial
infection. J Clin Oncol. 2004;22:3784-3789.
18. Manji A, Beyene J, Dupuis LL, et al. Outpatient and oral antibiotic
management of low-risk febrile neutropenia are effective and safe in childrena systematic review of prospective trials. Poster presented at: 2011
Annual Pediatric Oncology Group of Ontario Symposium; November 2011;
Toronto, ON, Canada.
19. Liou SY, Stephens JM, Carpiuc KT, et al. Economic burden of haematological adverse effects in cancer patients: a systematic review. Clin Drug
Investig. 2007;27:381-396.
20. Lathia N, Mittmann N, DeAngelis C, et al. Evaluation of direct medical
costs of hospitalization for febrile neutropenia. Cancer. 2010;116:742-748.
574
21. Hendricks AM, Loggers ET, Talcott JA. Costs of home versus inpatient
treatment for fever and neutropenia: analysis of a multicenter randomized
22. Klaassen RJ, Allen U, Doyle JJ. Randomized placebo-controlled trial of
oral antibiotics in pediatric oncology patients at low-risk with fever and
neutropenia. J Pediatr Hematol Oncol. 2000;22:405-411.
23. Wiernikowski JT, Rothney M, Dawson S, et al. Evaluation of a home
intravenous antibiotic program in pediatric oncology. Am J Pediatr Hematol
Oncol. 1991;13:144-147.
24. Teuffel O, Amir E, Alibhai SM, et al. Cost-effectiveness of outpatient
management for febrile neutropenia in children with cancer. Pediatrics.
2011;127:e279-286.
25. Speyer E, Herbinet A, Vuillemin A, et al. Agreement between children
with cancer and their parents in reporting the childs health-related quality of
life during a stay at the hospital and at home. Child Care Health Dev.
2009;35:489-495.
26. Sung L, Feldman BM, Schwamborn G, et al. Inpatient versus outpatient management of low-risk pediatric febrile neutropenia: measuring parents and healthcare professionals preferences. J Clin Oncol. 2004;22:39223929.
27. Cheng S, Teuffel O, Ethier MC, et al. Health-related quality of life
anticipated with different management strategies for paediatric febrile neutropaenia. Br J Cancer. 2011;105:606-611.
28. Sung L, Alibhai SM, Ethier MC, et al. Discrete choice experiment
produced estimates of acceptable risk of therapeutic options in cancer patients with febrile neutropenia. J Clin Epidemiol. 2012. In press.
29. Ryan M, Farrar S. Using conjoint analysis to elicit preferences for
health care. BMJ. 2000;320:1530-1533.
30. Diorio C, Martino J, Boydell KM, et al. Parental perspectives on
inpatient versus outpatient management of pediatric febrile neutropenia.
J Pediatr Oncol Nurs. 2011;28:355-362.
31. Ungar WJ. Economic Evaluation in Child Health. Oxford: Oxford
University Press; 2010.
32. Upton P, Lawford J, Eiser C. Parent-child agreement across child
health-related quality of life instruments: a review of the literature. Qual Life
Res. 2008;17:895-913.
GENETIC COUNSELING OF THE PATIENT WITH

PEDIATRIC CANCER
CHAIR
Joshua D. Schiffman, MD
University of Utah
Salt Lake City, UT
SPEAKERS
Kim E. Nichols, MD
Philadelphia, PA
Sara Knapke, MS
Cincinnati Childrens Hospital Medical Center
Cincinnati, OH
I d e n t i fi c a t i o n , M a n a g e m e n t , a n d E v a l u a t i o n
of Children with Cancer-Predisposition
Syndromes
By Sara Knapke, MS, Kristin Zelley, MS, Kim E. Nichols, MD, Wendy Kohlmann, MS,
and Joshua D. Schiffman, MD
Overview: A substantial proportion of childhood cancers are

attributable to an underlying genetic syndrome or inherited
susceptibility. Recognition of affected children allows for
appropriate cancer risk assessment, genetic counseling, and
testing. Identification of individuals who are at increased risk
to develop cancers during childhood can guide cancer surveil-
ORE THAN 12,060 children and adolescents will be

diagnosed with cancer this year in North America.1
Of these patients, it is estimated that up to 10% will develop
their cancer because of the presence of an underlying cancerpredisposing condition.2-5 However, this 10% estimate has
been extrapolated primarily from adults whose cancers were
associated with germline mutations6 and the actual percentage of childhood cancers caused by underlying genetic mutations remains unclear. In fact, a recent study reported that
up to 29% of children seen in a pediatric oncology survivorship program qualified for consideration of referral to a
cancer genetics clinic based on a significant family history
of cancer, a history of specific tumors in the surviving child,
or the presence of unique physical findings.7 These data
suggest that more than one in four children or adolescents
with a history of cancer may have a genetic cancerpredisposing condition and could therefore potentially benefit from further evaluation and management.8 Mounting
evidence now demonstrates that early cancer detection in
this population may lead to improved survival and treatment outcomes.5,6,9-11 In this chapter, we will outline for the
practicing oncologist how to identify these high-risk pediatric patients based on specific tumor presentation and family
history patterns, and discuss the issues involved in managing children and adolescents with a known genetic predisposition to cancer. Ethical and psychosocial considerations
related to genetic testing for cancer risk in minors and
suggestions for family-centered and multidisciplinary care
will also be explored.
Identification of Children with Cancer Predisposition
There are several ways in which a patient with an underlying cancer-predisposing condition can be identified. Specifically, tumor type and laterality, family cancer history,
and presence of other physical features or medical conditions must be taken into consideration when determining
whether a child has a possible genetic predisposition to
cancer.
lance and clinical management, which may improve outcomes

for both the patient and other at-risk relatives. The information provided through this article will focus on the current
complexities involved in the evaluation and management of
children with cancer-predisposing genetic conditions and
highlight remaining questions for discussion.
be a feature of some syndromes. For example, over 70% of

patients with early or bilateral Wilms Tumor may have
an inherited or de novo mutation in the WT1 gene.14,15 An
example of specific tumor types associated with an underlying cancer-predisposing gene mutation include adrenocortical tumors and choroid plexus carcinomas, which may be
caused by germline TP53 mutations in 80%16 or 35% to
100%17-19 of patients, respectively. Based on these data, it is
now recommended that all children with adrenocortical
tumors and choroid plexus tumors be offered genetic testing
for TP53 mutations (Li-Fraumeni Syndrome [LFS]).20 In
addition, children with rhabdomyosarcoma younger than
age 3 should be considered for TP53 mutation screening.21
(Some oncologists also will consider testing for TP53 mutations in children with osteosarcoma younger than ages 5 to
10 at presentation based on the strong association between
LFS and osteosarcoma22). All children with bilateral retinoblastoma and up to 20% of children with unilateral retinoblastoma will have heritable retinoblastoma and most will
have a detectable germline mutation of the RB1 gene.23
More than 70% of pediatric patients with malignant paragangliomas (PGL) or pheochromocytomas (PCC) have underlying SDHB mutations (Familial PGL/PCC Syndrome),
and the percentage of affected patients is even higher when
considering other PGL- or PCC-related genes.24,25 Nearly a
third of patients with rhabdoid tumor (soft tissue or brain)
will harbor a SMARCB1/INI1 mutation (Rhabdoid Tumor
Syndrome).26 At least 12 percent of patients with pediatric
(i.e., succinate dehydrogenase [SDH]-deficient) gastrointestinal stromal tumors (GISTs) lacking mutations in the KIT
and PDGFRA genes have germline mutations in the SDHassociated genes.27 Even patients with hepatoblastoma have
a reported 10% risk of having a germline APC mutation
(Familial Adenomatous Polyposis [FAP] Syndrome).28 Unusual cancers diagnosed in children such as colorectal or
thyroid cancer also can indicate an inherited cancerpredisposition syndrome and should trigger the pediatric
Presence of Specific Tumor Patterns
As demonstrated in Table 1, a hereditary pediatric cancer

syndrome can be suggested by the presence of a specific
pattern of cancer presentation or tumor type in an individual. Bilateral tumors in paired organs, multifocal tumors,
and multiple primary cancers in a single individual often
indicate an underlying or inherited genetic cause of cancer.6,10,12,13 In addition, an earlier age of diagnosis can also
576
From the Cincinnati Childrens Hospital Medical Center, Cincinnati, OH; Childrens
Hospital of Philadelphia, Philadelphia, PA; Huntsman Cancer Institute, Salt Lake City,
UT; Center for Childrens Cancer Research (C3R), Huntsman Cancer Institute, Salt Lake
City, UT.
Address reprint requests Joshua Schiffman, MD, Division of Pediatric Hematology/
Oncology, Center for Childrens Cancer Research (C3R), Huntsman Cancer Institute, 2000
Circle of Hope, Salt Lake City, UT 84112.
1092-9118/10/1-10
CANCER PREDISPOSITION IN CHILDHOOD

Table 1. Differential Genetic Diagnosis by Tumor Type
Tumor Type
Differential Genetic Diagnosis
Gene(s)
Adrenocortical carcinoma
Li-Fraumeni syndrome (LFS)
TP53
Atypical teratoid and malignant rhabdoid tumor
Rhabdoid tumor syndrome
SMARCB1/INI1
Basal cell carcinoma
Nevoid Basal Cell Carcinoma Syndrome (NBCCS)/Gorlin
PTCH1
Choroid plexus carcinoma
LFS
TP53
Cystic nephroma (familial)
Pleuropulmonary blastoma (PPB) family tumor and dysplasia syndrome
DICER1
Desmoid tumors
Familial Adenomatous Polyposis (FAP)
APC
Endolymphatic sac tumors
Von-Hippel Lindau syndrome
VHL
Glioblastoma multiforme
Turcot syndrome/Lynch syndrome

LFS
MLH1, MSH2, MSH6, PMS2

TP53
Hemangioblastoma (retinal/cerebellar)
Von-Hippel Lindau syndrome
VHL
Hepatoblastoma
FAP
Beckwith-Weidemann syndrome
APC
CDKN1C/11p15
Medulloblastoma
Turcot syndrome/FAP syndrome

NBCCS/Gorlin syndrome
APC
PTCH1
Medullary thyroid cancer
Multiple Endocrine Neoplasia type 2
RET
Neuroblastoma (bilateral or multifocal)
Familial Neuroblastoma
ALK, PHOX2B
Optic pathway tumor
Neurofibromatosis type 1
NF1
Ovarian sex cord-stromal tumors
PPB family tumor and dysplasia syndrome

Peutz-Jeghers syndrome
DICER1
STK11
Paraganglioma/Pheochromocytoma
Familial PGL/PCC syndrome

Von Hippel Lindau syndrome
Multiple Endocrine Neoplasia type 2
SDHB/C/D/AF2
VHL
RET
NF1
Retinoblastoma
Hereditary Retinoblastoma
RB1
Pleuropulmonary blastoma
PPB family tumor and dysplasia/DICER1 syndrome
DICER1
LFS
LFS
TP53
DICER1
TP53
Schwannoma (acoustic or vestibular)
NF2
Sertoli-Leydig cell tumor
DICER1
Wilms tumor (bilateral)
Wilms tumor syndrome
WT1
Sarcomas
Rhabdomyosarcoma
Osteosarcoma
oncologist to consider a cancer genetics referral.29-32 Pediatric patients presenting with the tumors listed in Table 1
should be considered for referral to a cancer genetics clinic
where they can be seen by a team of specialists including
genetic counselors, oncologists, and geneticists with expertise in cancer predisposition. We believe that the more an
oncologist looks for these presentations, the more pediatric
hereditary cancer syndromes will be identified in patients.8
If recognized, at-risk patients and their families can benefit
from genetic testing, cancer screening, and enrollment in
research protocols.
KEY POINTS
There exists a growing number of hereditary cancer

syndromes with implications in childhood.
Identification of cancer predisposition in childhood
can substantially influence the choice of appropriate
screening and management options for the child and
other relatives.
Genetic risk assessment, counseling, and testing are
critical elements in the care of children and families
with cancer-predisposing conditions.
This report focuses on predisposition to solid tumors,

where surveillance is most likely to influence treatment and
outcome. However, it is important to recognize that a number of genetic conditions predispose the patient to the
development of hematopoietic malignancies, myelodysplastic syndrome, or both. The reader is directed to the Seif
reference33 for an in-depth discussion of these conditions.
Family History as a Screening Tool
Evaluation of all children with cancer should include

assessment of the family history. Features of the family
medical history that may indicate the presence of a hereditary cancer syndrome include, but are not limited to (1)
multiple family members on the same side of the family with
the same or related type of cancers, or individuals with
distinct cancer types that are known to group together in
specific cancer-predisposition syndromes (e.g., rhabdomyosarcoma, osteosarcoma, adrenocortical carcinoma, brain tumors, leukemia, and early-onset breast cancer, as is seen in
families with LFS); (2) close relatives with early, multiple, or
bilateral cancers; and (3) a pattern suggestive of generationto-generation transmission (i.e., autosomal dominant inheritance). Therefore, clinicians should collect information on
age of cancer onset, type of cancer and laterality in at least
first- (parents and siblings), second- (grandparents, aunts,
and uncles), and third- (cousins and great grandparents)
577
KNAPKE ET AL
degree relatives and consider referral for any concerning

features in the family history.
There are a number of potential limitations to the use of
family history as a screening tool, such as adoption of the
patient or another family member without information
about biologic relatives, small family size, limited or no
contact with relatives, and early death from noncancer
related causes. In addition, for some hereditary cancer
syndromes, there are a number of de novo or new dominant
mutations or alternative inheritance patterns (i.e., autosomal recessive) and, in these cases, family history would
not be fully informative. As family histories evolve over
time, it is also possible that members may not have yet
demonstrated features of a particular hereditary cancer
syndrome. This is especially true in pediatrics where parents will be younger than those of adult patients. For this
reason, frequent updating of the family history is indicated.
Many cancer syndromes are also characterized by reduced
penetrance and may not exhibit a clear inheritance pattern.
Medical History and Physical Features Concerning for a Hereditary
Cancer Syndrome
Medical history and physical examination are important

components in the evaluation of children for a cancerpredisposing syndrome. Several hereditary cancer syndromes are characterized not only by the development of
cancer, but also by the presence of certain physical or
developmental manifestations that might provide important
clues to the diagnosis (Table 2). These may include congenital anomalies or dysmorphic features, along with developmental delays, intellectual disabilities, autism, or autism
spectrum disorders. Other associated features may include
skin findings, macrocephaly, and benign tumors or polyps.
In many cases, one or more of these features precede the
development of cancer and serve as an early clue to the
diagnosis. The presence of one or more of these features in a
child diagnosed with cancer should raise suspicion for a
hereditary cancer syndrome and prompt referral to a geneticist or genetic counselor. Referral to other specialties may
also be necessary to evaluate and manage the physical or
neurocognitive issues related to these genetic conditions.
Management of Children with Cancer Predisposition
Surveillance. Recognition of hereditary cancer syndromes

in children is necessary to facilitate appropriate surveillance
and management of individuals who are predisposed to
malignancies. Surveillance and management guidelines exist or are being developed for several hereditary cancer
syndromes with onset in childhood (Table 2). The primary
goal of cancer surveillance is to detect cancer at the earliest
and most curable stage with the least amount of complications and late effects from treatment. For this reason, cancer
surveillance protocols are most suited for solid tumors, such
as hepatoblastoma and Wilms tumor, where outcome is
directly linked to stage at diagnosis. By detecting cancer at
an early stage, cancer surveillance protocols aim to improve
overall survival and decrease morbidity in individuals at
increased risk for tumor development. Cancer surveillance
has the potential benefits of increasing the cure rate for
cancers and reducing or eliminating the need for chemotherapy, radiation therapy, or both, which can have substantial
side effects. However, cancer surveillance also can result in
increased worry about cancer, an increased rate of biopsies
578
or invasive procedures because of false-positive results, and

increased cost of care. Benefits and disadvantages of surveillance may differ from one genetic syndrome to another, and
the benefits of each syndrome-specific protocol must outweigh the disadvantages. We recommend an open discussion
with patients and their families about both the advantages
and the risks involved in early cancer screening.
Several factors must be considered in the development
and implementation of an effective cancer surveillance protocol. First, there must be a benefit to early cancer detection
in individuals with cancer-predisposing conditions. Second,
the age-specific cancer risks for the hereditary cancer syndrome must be known and considered high enough to
warrant surveillance. This information is needed to know to
whom surveillance should be offered, when to initiate surveillance, and the duration of surveillance. The surveillance
method and interval must also be carefully considered in
light of the specific cancer risks associated with a given
syndrome. Ideally, surveillance methods should be readily
available, safe, and have high sensitivity and specificity.
Whenever possible, screening tools that involve no or minimal radiation should be used, as patients with hereditary
cancer syndromes may be at increased risk for developing
radiation-induced cancers. Finally, there must be effective
treatment methods available for the cancer(s) identified
through screening.
For some hereditary cancer syndromes, such as hereditary
retinoblastoma, surveillance and management guidelines
are well established and have been shown to improve outcomes for affected individuals.34,35 For other syndromes,
such as LFS, there is ongoing debate about the optimal
surveillance protocol. The development of an effective surveillance protocol for individuals diagnosed with or at risk
for LFS has been complicated by the variability in type,
location, and age of onset of tumors, as well as the increased
risk for multiple primary tumors and radiation-induced
cancers.36 This raises many questions about the optimal
method(s), interval, and duration of surveillance. Several
groups have begun to offer a combined modality surveillance
protocol for LFS utilizing rapid whole-body magnetic resonance imaging (MRI), brain MRI, abdominal ultrasound,
complete blood count, and biochemical markers. Recent data
have shown that this surveillance protocol may indeed be
effective and improves the survival of patients with LFS
through early tumor detection.11 Further prospective studies will be necessary to validate the effectiveness of this
surveillance protocol in both children and adults.
It is important to note that surveillance protocols may
change over time as new evidence becomes available. Therefore, clinicians should refer to the literature and resources
such as the National Comprehensive Cancer Network
(NCCN) for the most up-to-date surveillance guidelines.
Risk-Reduction Strategies
For some pediatric cancer syndromes, early identification

of high-risk children allows for the elimination or dramatic
reduction of risk through prophylactic surgery. For example,
multiple endocrine neoplasia type 2 (MEN2) and FAP are
classic examples of when surgery to remove the at-risk
organ may be considered early in the patients lifetime to
prevent the development of cancer. MEN2 is caused by
mutations in RET and is associated with nearly a 100%
lifetime risk for medullary thyroid cancer (MTC), as well as

Table 2. Hereditary Cancer Syndromes with Solid-Tumor Risk and Manifestations in Childhood: Genetics, Risks, Features, and Surveillance
Syndrome
Gene(s)
Inheritance
Cancer/Tumor Risks
Other Features
Cancer Surveillance
Beckwith-Wiedemann
syndrome/idiopathic
hemihypertrophy52,59
Chromosome 11p15
methylation defects;
UPD; CDKN1C
Imprinting/AD
Hepatoblastoma
Wilms tumor
Hemihypertrophy
Macroglossia
Omphalocele
Umbilical hernia
Neonatal hypoglycemia
Ear pits/creases
- Abdominal ultrasound every 3 mo from birth until age

4y
- Serum AFP level every 612 wk from birth until
age 4 y
- Renal ultrasound every 3 mo from 48 y
Biallelic Mismatch Repair

Gene syndrome53
MLH1, MSH2, MSH6,

PMS2
AR
Lymphoma
Leukemia
Brain tumors
Colorectal
Small bowel
Associated benign tumors:
- GI polyps (adenomas)
Caf au lait macules

Axillary and inguinal
freckling
Lisch nodules
- Colonoscopy annually beginning at age 3 or at

diagnosis
- Upper endoscopy (EGD) and video capsule endoscopy
annually
- Ultrasound of head at birth and MRI of brain every
6 mos
- CBC, erythrocyte sedimentation rate, lactate
dehydrogenase every 4 mos
- Urinary and uterine ultrasound annually in adulthood
Familial Adenomatous
Polyposis39
APC
AD
Colon
Small bowel
Pancreatic
Thyroid - papillary
Hepatoblastoma
CNSmedulloblastoma
Bile duct
Gastric
FAP-associated benign tumors:
- Soft tissue tumors (Gardners
fibromas, desmoids tumors)
- Osteomas
- GI polyps (adenomas)
Epidermoid cysts
Supernumerary/missing
teeth
Congenital hypertrophy of
the retinal pigmented
epithelium (CHRPE)
- Colonoscopy every1224 mo from age 1012 y until

colectomy
- Upper endoscopy (EGD) every 1236 mo starting at
age 18, lifelong
- Thyroid exam and consider thyroid ultrasound every
12 mo starting at age 18, lifelong
- Consider hepatoblastoma screening:
- Abdominal ultrasound every 3 mo from birth until age
4y
- Serum AFP level every 612 wk from birth until
age 4 y
Familial Neuroblastoma
ALK, PHOX2B
AD
Neuroblastoma
Hirschsprung disease
(PHOX2B only)
No published guidelines available
Familial PGL/PCC
Syndrome54
SDHB/C/D/AF2
AD
Paraganglioma
Pheochromocytoma
Possible associations: Renal,
Thyroid
None
- Urine or plasma metanephrines and catecholemines

every 12 mo starting at age 10 y, lifelong
- MRI of neck, chest, abdomen, and pelvis every 12 mo
starting at age 10 y, lifelong
Heritable
Retinoblastoma55
RB1
AD
Retinoblastoma
Pinealomas
Sarcomas
Melanoma
None
- Brain MRI every 6 mo from birth until age 5 y

- Eye exam, frequency determined by ophthalmologist,
from birth until age 5 y
- Thorough annual physical exam and careful attention
to development of any lumps or lesions (because of
the high rate of second cancers, particularly among
those who had RX treatment)
Juvenile Polyposis
syndrome39
SMAD4, BMPR1A
AD
Colon
Gastric
Small intestine
Pancreatic
- Juvenile GI polyps
Arteriovascular
malformations (SMAD4
only)
- Monitor for rectal bleeding, lifelong

- Colonoscopy/EGD every 1236 mo starting at
age 15 or if symptoms
- CBC every 12 mo starting in early childhood
- If SMAD4 mutation, surveillance for Hereditary
Hemorrhagic Telangiectasia
Li-Fraumeni syndrome56
TP53
AD
Adrenocortical carcinoma
Choroid plexus carcinoma
Bone and soft tissue sarcomas
Leukemia
Breast
Numerous other
None
- Physical exam (with careful skin and neurologic exam)

every 12 mo, lifelong
- Self breast exam every 1 mo starting at age 18 y,
lifelong
- Clinical breast exam every 612 mo starting at
age 2025 y, lifelong
- Mammogram and breast MRI every 12 mo starting at
age 2025 y, lifelong
- Colonoscopy (suggested) every 25 y starting no later
than age 25 y, lifelong
- Consider additional screening with abdominal
ultrasound, brain MRI, whole body MRI, and
biochemical markers11
Multiple Endocrine
Neoplasia type 150,51
MEN1
AD
Parathyroid adenoma
Gastrinoma
Insulinoma
Anterior pituitary
Forgut carcinoid
Numerous other
MEN1-associated benign tumors:
- angiofibromas
- collagenomas
- lipomas
None
- Annual serum concentration of prolactin from

age 5 y
- Annual fasting total serum calcium concentration
(corrected for albumin) and/or ionized-serum calcium
concentration from age 8 y
- Annual fasting serum gastrin concentration from
age 20 y
579
KNAPKE ET AL
Table 2. Hereditary Cancer Syndromes with Solid-Tumor Risk and Manifestations in Childhood: Genetics, Risks, Features, and Surveillance (Contd)
Syndrome
Gene(s)
Inheritance
Cancer/Tumor Risks
Other Features
Cancer Surveillance
- To be considered: annual fasting serum
concentration of intact (full-length) PTH
- Head MRI from age 5 y every 35 y
- Abdominal CT or MRI from age 20 y every 35 y
- To be considered: yearly chest CT, SRS octreotide
scan
Multiple Endocrine
Neoplasia type 237
RET
AD
Thyroidmedullary
Pheochromocytoma
MEN2-associated benign
tumors (MEN2B only):
- Mucosal neuromas
- Ganglioneuromas
Hyperparathyroidism
Marfanoid habitus (MEN2B only)
- Serum calcitonin every 12 mo

- PTH every 12 mo
- Urine/plasma metanephrines and catecholamines
every 12 mo
- Age at which screening is recommended to begin
varies based on the specific genetic mutation. See
American Thyroid Association Guidelines for
Management of Medullary Thyroid Cancer for
detailed recommendations.
Neurofibromatosis
type 151
NF1
AD
Schwannoma
Pheochromocytoma
Optic pathway tumor
Neurofibromas
JMML
AML
Caf au lait macules

Axillary and inguinal
freckling
Lisch nodules
Tibial bowing
Developmental delay/
intellectual disability/
autism
- Annual physical exam by a physician who is familiar

with the individual and with the disease
- Annual ophthalmologic exam in early childhood, less
frequent exam in older children and adults
- Regular developmental assessment by screening
questionnaire (in childhood)
- Regular blood pressure monitoring
- Other studies only as indicated based on clinically
apparent signs or symptoms
- Monitoring of those who have abnormalities of CNS,
skeletal system, or cardiovascular system by an
appropriate specialist
Neurofibromatosis
type 251
NF2
AD
Vestibular schwannoma
Meningioma
Schwannoma
Glioma
Neurofibroma
Posterior subcapsular
lenticular opacities
Cataract
Hearing loss
Focal weakness
Tinnitus
Balance dysfunction
Seizure
Focal sensory loss
Blindness
- Cranial MRI annually beginning at age 1012 y until

at least fourth decade of life
- Routine complete eye exam
- Hearing evaluation including BAER testing
Nevoid Basal Cell

Carcinoma (Gorlin)
syndrome51
PTCH1
AD
Basal cell carcinoma

Medulloblastoma
Jaw keratocysts
Macrocephaly
Palmar/plantar pits
Bifid ribs
Calcification of the falx
- Monitoring of head circumference through childhood

- Developmental assessment and physical exam every
6 mo
- Orthopantogram every 1218 mo starting at 8 y
- Skin exam at least annually
Peutz-Jeghers syndrome56
STK11
AD
Colorectal
Gastric
Small intestine
Pancreatic
Breast
Ovarian
Cervix
Uterus
Testes
Lung
PJS-associated benign tumors:
- Peutz-Jeghers GI polyps
- Sex cord tumors with annular
tubules (SCTAT)
Mucocutaneous pigmentation
- Clinical breast exam every 6 mo starting at age 25 y,

lifelong
- Mammogram and breast MRI every 12 mo starting at
age 25 y, lifelong
- Colonoscopy and EGD every 23 y starting in late
teenage years, lifelong
- MRCP and/or endoscopic ultrasound every 1224
mo, starting at age 2530 y, lifelong
- CA 199 every 1224 mo starting at age 2530 y,
lifelong
- Small bowel visualization, baseline at age 810 y,
follow-up based on findings
- Pelvic exam and pap smear every 12 mo starting at
age 1820 y
- Consider transvaginal ultrasound starting at age
1820 y
- Testicular exam every 12 mo starting at age 10 y
- Education about symptoms of lung cancer and
smoking cessation
Pleuropulmonary blastoma
(PPB) Family Tumor
and Dysplasia
syndrome/DICER1
syndrome
DICER1
AD
PPB
Rhabdomyosarcoma
Thyroid
Ovarian Sertoli-Leydig cell
tumors and dysgerminoma
Testicular seminoma
Other gonadal germ cell tumors
Leukemia
- Multinodular goiter
- Cystic nephroma
Pulmonary cysts
580

Table 2. Hereditary Cancer Syndromes with Solid-Tumor Risk and Manifestations in Childhood: Genetics, Risks, Features, and Surveillance (Contd)
Syndrome
Gene(s)
Inheritance
Cancer/Tumor Risks
Other Features
Cancer Surveillance
PTEN Hamartoma Tumor

syndrome56
PTEN
AD
Breast
Thyroid
Endometrial
Renal
- Multinodular goiter
- Cystic nephroma
PTEN-associated benign tumors:
- Lipomas
- Thyroid nodules/goiter
- Hamartomatous GI polyps
Macrocephaly
Arteriovascular malformations
autism
- Physical exam every 12 mo starting at age 18 y,

lifelong
- Thyroid ultrasound every 12 mo starting at age
18 y, lifelong
- Self breast exam every 1 mo starting at age 18 y,
lifelong
- Clinical breast exam every 612 mo starting at
age 25 y, lifelong
- Mammogram and breast MRI every 12 mo starting
at age 3035 y (or 510 y before earliest age of
diagnosis in family, whichever comes first)
- Patient education about signs and symptoms of
endometrial cancer and encourage prompt followup if issues arise
- Counsel about risk-reducing mastectomy and
hysterectomy on a case-by-case basis
- Colonoscopy (suggested) every 510 y, more
frequently if symptoms or polyps, starting at age
35 y, lifelong
Rhabdoid syndrome
SMARCB1/INI1
AD
Rhabdoid tumors
Schwannomatosis
None
von-Hippel Lindau
syndrome57
VHL
AD
Hemangioblastoma (retinal/
cerebellar)
Renal Cell Carcinoma
Pancreaticneuroendocrine
Pheochromocytoma
Endolymphatic sac tumors
Epididymal tumors
Renal and Pancreatic cysts
- Ophthalmologic screening every 12 mo starting at

age 1 y, lifelong
- Physical exam, blood pressure monitoring, and
neurologic assessment every12 mo starting at age
2 y, lifelong
- Urine/plasma catecholamine metabolites every
12 mo starting at age 2 y, lifelong
- Abdominal ultrasound every 12 mo from age
820 y
- Abdominal CT or MRI every 24 mo, to be
alternated with annual abdominal ultrasound
starting at age 20 y
- Brain/spine MRI every 12 mo starting at puberty
- Audiology assessment every 23 y from ages
210 y, and then as symptoms arise
Wilms tumor syndromes58

WAGR
Denys-Drash
Familial Wilms
Frasier
WT1
AD
Wilms tumor
Aniridia (WAGR syndrome)

Genitourinary defects
autism
- Renal ultrasound every 3 mo from birth until age

8y
Abbreviations: AD, autosomal dominant; AML, Acute myeloid leukemia; AR, autosomal recessive; AFP, alpha-fetoprotein; BAER, brainstem auditory evoked response;
CA 19-9, carbohydrate antigen 19-9; CBC, complete blood count; CHRPE, congenital hypertrophy of the retinal pigmented epithelium; CNS, central nervous system;
CT, computed tomography; EGD, esophagogastroduodenoscopy; GI, gastrointestinal; JMML, juvenile myelomonocytic leukemia; MEN2, multiple endocrine neoplasia
type 2; mo, month (s); MRCP, magnetic resonance cholangiopancreatography; MRI, magnetic resonance imaging; PPB, pleuropulmonary blastoma; PTH, parathyroid
hormone; SCTAT, sex cord tumors with annular tubules; SRS, somatostatin receptor scintigraphy; WAGR, Wilms tumor, aniridia, genitourinary anomalies, and mental
retardation syndrome; wk, week(s); XRT, radiation; y, year(s).
an increased risk for hyperparathyroidism and pheochromocytoma. Although all individuals with a RET mutation have
an increased risk for MTC, there are direct genotypephenotype correlations that predict the age of onset and
determine timing for prophylactic thyroidectomy.37 Studies
have found that genetic identification of children at risk for
MEN2 and prophylactic thyroidectomy has greatly reduced
the likelihood of developing MTC in this population.38
Similarly, the identification of the APC gene as the cause
of FAP has allowed for genetic testing to identify children
with this condition. Individuals with APC mutations associated with a classic FAP phenotype develop hundreds to
thousands of colonic polyps beginning in adolescence and
have a risk of colon cancer that approaches 100% if untreated. Current guidelines recommend beginning surveillance for colon polyps at age 10. When polyps become too
numerous to follow endoscopically, colectomy is recommended.39 Use of genetic testing to evaluate at-risk children
for a familial APC mutation is more cost-effective than
relying on colon examinations to determine whether a child
has inherited this condition.40 Because of the morbidities
associated with colectomy, alternatives to surgery are being
sought. Nonsteroidal anti-inflammatory inhibitors, such as

sulindac, and Cox-2 inhibitors have been found to reduce
polyp development in adults with FAP.41,42 Studies have
now begun to look at the potential utility of these medications in children43 and an international clinical trial is
currently enrolling children with a molecular diagnosis of
FAP in a 5-year trial comparing the rate of polyp development with celecoxib compared with placebo.44
Genetic Risk Assessment and Counseling
Elements of an appropriate cancer genetic evaluation

include collection of a thorough personal and family medical
history, genetic risk assessment through pedigree analysis
and published literature, genetic testing when appropriate
for specific cancer syndromes, informed consent, results
disclosure, and psychosocial assessment.45 This process is
often complex and may also require other essential components including medical records ascertainment and review,
health insurance preauthorization for testing, and facilitating communication with other at-risk family members about
complex results.
Benefits of cancer genetic risk assessment, counseling,
581
KNAPKE ET AL
and testing include identification of at-risk individuals and

families. In cases where a causative gene mutation can be
identified, mutation-specific testing can identify those individuals in the family who have inherited the genetic risk
factor and warrant high-risk screening and management.
Single-site mutation testing can also identify those individuals in the family who did not inherit the condition and,
therefore, are not predicted to be at increased risk and can
forego additional measures. Many family members often
have increased anxiety and worry about the risk for cancer
in the family, and appropriate risk assessment and identification of a specific cause can provide accurate information
and, in some cases, help to empower family members and
alleviate emotional burden.
There are also potential disadvantages of and obstacles to
cancer genetics evaluations. For example, although the
sensitivity and utility of genetic testing continues to improve, a causative gene mutation cannot be identified in
some cases even when there is a high suspicion of a specific
diagnosis. Therefore, clinical judgment and expertise must
be applied in these cases to develop an acceptable screening
and management plan for the patient as well as at-risk
family members. In addition, there are also some cases with
striking features of a hereditary cancer syndrome that do
not fit a specific diagnosis or may represent a previously
undescribed syndrome. Research studies as well as advances
in gene finding and exome sequencing may be beneficial in
such cases. However, these newer genomic technologies may
lead to the identification of more variants of unknown
significance for which it can be difficult to counsel the
patient and his or her family. Advancements in genetic
information and testing continue to change at a rapid pace.
Therefore, there needs to be a clear expectation in the
pediatric cancer genetics clinic for periodic follow-up and
recontact with families in the event that new information is
obtained. The results of genetic testing should not stand
alone in risk assessment but rather be one tool in the genetic
cancer risk assessment process.
Ethical and psychosocial considerations remain critical in
the assessment and care of children with potential cancerpredisposition syndromes. For example, with respect to the
informed consent process, clinicians need to consider the
childs capacity for autonomy as well as participation in
assent/consent. In addition, much historic debate has occurred about genetic testing in minors.46-49 A distinction
between diagnostic and presymptomatic or predictive testing is relevant. Although diagnostic testing is generally
acceptable in children with features of a genetic condition,
predictive testing is generally reserved for those conditions
for which clinical management would be altered during
childhood.48 Regarding the psychosocial effects on the family
experience, it is important to explore implications on emotional well-being, family dynamics, risk perception, influ-
582
Table 3. Cancer Genetic Services Resources

Resource
Web site
National Society of Genetic Counselors

Find a Genetic Counselor Tool
Cancer Genetic Services Directory
www.nsgc.org
http://cancer.gov/cancertopics/genetics/
directory
ence on other siblings, reproductive decision-making, and

financial consequences.
Because of all of these potential risks and benefits, it is
important that discussions on genetic testing are done in a
sensitive, comprehensive, and inclusive manner by fully
trained specialist health professionals, such as genetic counselors and clinical geneticists, in a relaxed and comfortable
environment.48 After identifying individuals and families
who might be at increased risk for cancer, referral to a
program with expertise in childhood cancer predisposition is
indicated. Many health care systems may have genetic
counselors or other specialists with genetic expertise on site.
Others may need to seek out and establish appropriate
referral practices to another organization in their area.
Resources for finding local genetic specialists can be found in
Table 3. If cancer genetics services are not available nearby,
an increasing number of programs also offer their services
through a telemedicine service model. When possible, it also
may be beneficial to establish relationships with cancer
genetics programs that practice through a multidisciplinary
approach to care. A growing number of cancer genetics
programs have established specific clinics related to pediatric cancer predisposition and integrate expertise from clinical geneticists, pediatric oncologists, and other relevant
subspecialists.
Conclusion
In an era of personalized medicine, identification of disease susceptibility is no longer solely for academic interest
but is becoming an accepted and clinically relevant element
in the current management of patients. Therefore, it is
imperative for clinicians to recognize those children and
families who will benefit most from a cancer genetics referral, and assist in the follow-up and management of these
individuals. Although a great deal of knowledge about
cancer-predisposing conditions affecting children now exists, the scope of genomic information is expanding at a
rapid pace and the future of this field will become increasingly complex. These new genetic data must be carefully
examined through clinical, translational, and basic research
protocols to ensure their effective translation to the optimized care of children at increased genetic risk for cancer.
Acknowledgements
K.E.N. acknowledges support in part by the Grundy Vision of
Life Fund. W.K. and J.D.S. acknowledge the use of the Genetic
Counseling Shared Resource supported by P30 CA042014
awarded to Huntsman Cancer Institute.
Author
Sara Knapke*
Kristin Zelley*
Kim E. Nichols*
Wendy Kohlmann
Joshua Schiffman*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Myriad
REFERENCES
1. American Cancer Society. Cancer Facts & Figures 2012. Atlanta: American Cancer Society, 2012.
2. Narod SA, Stiller C, Lenoir GM. An estimate of the heritable fraction of
childhood cancer. Br J Cancer. 1991;63:993-999.
3. Pakakasama S, Tomlinson GE. Genetic predisposition and screening in
pediatric cancer. Pediatr Clin North Am. 2002;49:1393-1413.
4. Plon SE, Nathanson K. Inherited susceptibility for pediatric cancer.
Cancer J. 2005;11:255-267.
5. Strahm B, Malkin D. Hereditary cancer predisposition in children:
genetic basis and clinical implications. Int J Cancer. 2006;119:2001-2006.
6. Garber JE, Offit K. Hereditary cancer predisposition syndromes. J Clin
Oncol. 2005;23:276-292.
7. Knapke S, Nagarajan R, Correll J, et al. Hereditary cancer risk assessment
in a pediatric oncology follow-up clinic. Pediatr Blood Cancer. 2012;58:85-89.
8. Schiffman JD. Hereditary cancer syndromes: if you look, you will find
them. Pediatr Blood Cancer. 2012;58:5-6.
9. American Society of Clinical Oncology policy statement update: genetic
testing for cancer susceptibility. J Clin Oncol. 2003;21:2397-2406.
10. Eng C, Hampel H, de la Chapelle A. Genetic testing for cancer
predisposition. Annu Rev Med. 2001;52:371-400.
11. Villani A, Tabori U, Schiffman J, et al. Biochemical and imaging
surveillance in germline TP53 mutation carriers with Li-Fraumeni syndrome:
a prospective observational study. Lancet Oncol. 2011;12:559-567.
12. Nichols KE, Malkin D, Garber JE, et al. Germ-line TP53 mutations
predispose to a wide spectrum of early-onset cancers. Cancer Epidemiol
13. Qualman SJ, Bowen J, Erdman SH. Molecular basis of the brain
tumor-polyposis (Turcot) syndrome. Pediatr Dev Pathol. 2003;6:574-576.
14. Royer-Pokora B, Beier M, Henzler M, et al. Am J Med Genet A.
2004;127:249-257.
15. Ruteshouser EC, Huff V. Familial Wilms tumor. Am J Med Genet C
Semin Med Genet. 2004;129C:29-34.
16. Varley JM, McGown G, Thorncroft M, et al. Are there low-penetrance
TP53 alleles? Evidence from childhood adrenocortical tumors. Am J Hum
Genet. 1999;65:995-1006.
17. Gonzalez KD, Noltner KA, Buzin CH, et al. Beyond Li Fraumeni
syndrome: clinical characteristics of families with TP53 germline mutations.
J Clin Oncol. 2009;27:1250-1256.
18. Gozali AE, Britt B, Shane L, et al. Choroid plexus tumors; management, outcome, and association with the Li-Fraumeni syndrome: The Childrens Hospital Los Angeles (CHLA) experience, 1991-2010. Pediatr Blood
Cancer. Epub 2011 Oct 11.
19. Tabori U, Shlien A, Baskin B, et al. TP53 alterations determine clinical
subgroups and survival of patients with choroid plexus tumors. J Clin Oncol.
2010;28:1995-2001.
20. Tinat J, Bougeard G, Baert-Desurmont S, et al. 2009 version of the
Chompret criteria for Li Fraumeni syndrome. J Clin Oncol. 2009;27:e108-109.
21. Diller L, Sexsmith E, Gottlieb A, et al. Germline p53 mutations are
frequently detected in young children with rhabdomyosarcoma. J Clin Invest.
1995;95:1606-1611.
22. Calvert G, Randall LR, Jones KB, et al. At risk populations for
osteosarcoma: the syndromes and beyond. Sarcoma. In press.
23. Draper GJ, Sanders BM, Brownbill PA, et al. Patterns of risk of
hereditary retinoblastoma and applications to genetic counseling. Br J
Cancer. 1992;66:211-219.
24. King KS, Prodanov T, Kantorovich V, et al. Metastatic pheochromocytoma/paraganglioma related to primary tumor development in childhood or
adolescence: significant link to SDHB mutations. J Clin Oncol. 2011;29:41374142.
25. Schiffman JD. No child left behind in SDHB testing for paragangliomas
and pheochromocytomas. J Clin Oncol. 2011;29:4070-4072.
26. Eaton KW, Tooke LS, Wainwright LM, et al. Spectrum of SMARCB1/
INI1 mutations in familial and sporadic rhabdoid tumors. Pediatr Blood

Cancer. 2011;56:7-15.
27. Janeway KA, Kim SY, Lodish M, et al. Defects in succinate dehydrogenase in gastrointestinal stromal tumors lacking KIT and PDGFRA mutations. Proc Natl Acad Sci U S A. 2011;108:314-318.
28. Aretz S, Koch A, Uhlhaas S, et al. Should children at risk for familial
adenomatous polyposis be screened for hepatoblastoma and children with
apparently sporadic hepatoblastoma be screened for APC germline mutations? Pediatr Blood Cancer. 2006;47:811-818.
29. Dinauer C, Francis GL. Thyroid cancer in children. Endocrinol Metab
Clin North Am. 2007;36:779-806.
30. Moore SW, Appfelstaedt J, Zaahl MG. Familial medullary carcinoma
prevention, risk evaluation, and RET in children of families with MEN2.
J Pediatr Surg. 2007;42:326-332.
31. Walker M, OSullivan B, Perakath B, et al. Selecting patients with
young-onset colorectal cancer for mismatch repair gene analysis. Br J Surg.
2007;94:1567-1571.
32. Lwiwski N, Greenberg CR, Mhanni AA. Genetic testing of children at
risk for adult onset conditions: when is testing indicated? J Genet Couns.
2008;17:523-525.
33. Seif AE. Pediatric leukemia predisposition syndromes: clues to understanding leukemogenesis. Cancer Genetics. 2011;204:227-244.
34. Imhof SM, Moll AC, Schouten-van Meeteren AY. Stage of presentation
and visual outcome of patients screened for familial retinoblastoma: nationwide registration in the Netherlands. Br J Ophthalmol. 2006;90:875-878.
35. Kivelo T. Trilateral retinoblastoma: a meta-analysis of hereditary
retinoblastoma associated with primary ectopic intracranial retinoblastoma.
J Clin Oncol. 1999;17:1829-1837.
36. Varley JM, Evans DG, Birch JM. Li-Fraumeni syndrome: A molecular
and clinical review. Br J Cancer. 1997;76:1-14.
37. Kloos RT, Eng C, Evans DB, et al. Medullary thyroid cancer: management guidelines of the American Thyroid Association. Thyroid. 2009;19:565612.
38. Skinner MA, Moley JA, Dilley WG, et al. Prophylactic thyroidectomy in
multiple endocrine neoplasia type 2A. N Engl J Med. 2005;353:1105-1113.
39. National Comprehensive Cancer Network. Colorectal cancer screening.
NCCN Clinical Practice Guidelines in Oncology. Version 2.2011. www.nccn.
org. Accessed February 6, 2011.
40. Bapat B, Noorani H, Cohen Z, et al. Cost comparisons of predictive
genetic testing versus conventional screening for familial adenomatous polyposis. Gut. 1999;44:698-703.
41. Boon EM, Keller JJ, Wormhoudt TA, et al. Sulindac targets nuclear
beta-catenin accumulation and Wnt signaling in adenomas of patients with
familial adenomatous polyposis and in human colorectal cancer cell lines.
Br J Cancer. 2004;90:224-229.
42. Steinbach G, Lynch PM, Phillips RK, et al. The effect of celecoxib, a
cyclooxygenase-2 inhibitor, in familial adenomatous polyposis. N Engl J Med.
2000;342:1946-1952.
43. Lynch PM, Ayers GD, Hawk E, et al. The safety and efficacy of celecoxib
in children with familial adenomatous polyposis. Am J Gastroenterol. 2010;
105:1437-1443.
44. ClinicalTrials.gov. Accessed March 23, 2012.
45. Riley BD, Culver JO, Skrzynia C, et al. Essential elements of genetic
cancer risk assessment, counseling and testing: updated recommendations of
the National Society of Genetic Counselors. J Genetic Counsel. Epub 2011 Dec
2.
46. Clayton EW. Genetic testing in children. J Med Philos. 1997;22:233251.
47. Malpas PJ. Predictive genetic testing of children for adult-onset diseases and psychological harm. J Med Ethics. 2008;34:275-278.
48. Tischkowitz M, Rosser E. Inherited cancer in children: practical/ethical
problems and challenges. Eur J Cancer. 2004;40:2459-2470.
583
KNAPKE ET AL
49. Nelson RM, Botkin JR, Kodish ED, et al. Ethical issues with genetic
testing in pediatrics. Pediatrics. 2001;107:1451-1455.
50. Brandi ML, Gagel RF, Angeli A, et al. Guidelines for diagnosis and
therapy of MEN type 1 and type 2. J Clin Endocrinol Metab. 2001;86:5658-5671.
51. GeneReviews. www.genetests.org. Accessed March 23, 2012.
52. Debaun MR. Screening protocol for Beckwith-Wiedemann syndrome.
http://www.beckwith-wiedemann.info/protocol.html. Accessed February 6,
2012.
53. Durno CA, Aronson M, Tabori U, et al. Oncologic surveillance for
subjects with biallelic mismatch repair gene mutations: 10-year follow-up of a
kindred. Pediatr Blood Cancer. Epub 2011 Dec 16.
54. Timmers HJ, Gimenez-Roqueplo AP, Mannelli M, et al. Clinical aspects
of SDHx-related pheochromocytoma and paraganglioma. Endocr Relat Cancer. 2009;16:391-400.
584
55. National Cancer Institute. Retinoblastoma treatment (Physician Data

Query). http://www.cancer.gov/cancertopics/pdq/treatment/retinoblastoma/
HealthProfessional.Accessed January 30, 2012.
56. National Comprehensive Cancer Network. Genetic/familial high-risk
assessment: breast and ovarian. NCCN Clinical Practice Guidelines in
Oncology. Version 1.2011. www.nccn.org. Accessed February 6, 2012.
57. VHL Family Alliance. Suggested protocol for screening. www.vhl.org.
Accessed February 6, 2012.
58. Scott RH, Walker L, Olsen OE, et al. Surveillance for Wilms tumor in
at-risk children: pragmatic recommendations for best practice. Arch Dis
Child. 2006;91:995-999.
59. Choyke PL, Siegel MJ, Craft AW, et al. Screening for Wilms tumor in
children with Beckwith-Wiedemann syndrome or idiopathic hemihypertrophy. Med Pediatr Oncol. 1999;32:196-200.
HOW TO MANAGE VERY RARE PEDIATRIC CANCERS

CHAIR
Alberto S. Pappo, MD
St. Jude Childrens Research Hospital
Memphis, TN
SPEAKERS
Fariba Navid, MD
Memphis, TN
Stephen Skapek, MD
Chicago, IL
Treating Rare Cancer in Children:

The Importance of Evidence
By Alberto S. Pappo, MD
Overview: The study of pediatric rare cancers, which account

for approximately 9% of all childhood malignancies, has been
hindered by their histologic heterogeneity and by their preferential occurrence in adolescents, a population that has been
underrepresented in clinical trials sponsored by the National
Cancer Institute.
The use of cooperative group and investigator-initiated
registries can help improve our ability to identify and select
populations of patients with rare cancers that can benefit from
single-arm studies, and incorporation of biologic aims and
tissue banking can accelerate our understanding of the biology of these cancers. These studies should be promoted
further through expansion of international outreach efforts.
HE STUDY of rare pediatric cancers is complicated by

the fact that pediatric cancer itself is a rare occurrence,
with only approximately 13,000 patients younger than age
20 diagnosed yearly in the United States.1 Several groups
have tried to define a pediatric rare cancer; for example the
European Italian Tumori Rari in Eta Pediatrica (TREP)
(the rare tumor project for pediatric patients) project uses
a definition of a cancer that has an annual incidence of up to
2 per million and for whom there is no available clinical
trial.2 However, on the basis of data from the Surveillance,
Epidemiology, and End Results (SEER) program of the
National Cancer Institute, this definition would exclude
tumors such as thyroid cancer and melanoma, which have
an increasing incidence in adolescents and young adults, but
are still considered rare. With the merger of the Pediatric
Oncology Group and the Childrens Cancer Group in 2000,
the newly formed Childrens Oncology Group (COG) was
poised to increase our understanding of rare cancers in
children. With this objective in mind, a Rare Tumor Committee was formed and the results of its initial experience
within the context of a cooperative group have been published and summarized.3 When initially conceived, the main
objectives of the COG Rare Tumor Committee were to
facilitate the study of rare cancers, using the infrastructure
available at the time that the two Groups merged. To better
estimate the actual numbers of rare tumors diagnosed in the
United States, we opted to define them as those tumors
(mostly carcinomas) mostly diagnosed in older patients.
These subset of tumors closely fit the description of other
malignant epithelial neoplasms and melanomas listed in
the International Classification of Childhood Cancer subgroup XI of the SEER database. These histologies include
nasopharyngeal carcinoma, adrenocortical carcinoma, melanoma, nonmelanoma skin cancers, and other unspecified
carcinomas. We then used the available COG registry to
capture cases of these rare cancers and compared the registration rates with the expected number of cases as calculated according to available through the SEER database. We
were surprised to see that only 7% of expected cases of
selected histologies such as melanoma, thyroid carcinoma,
adrenocortical carcinoma, and nasopharyngeal carcinoma
were registered within COG. We also noted that the registration rates varied according to histologic subtype and age,
586
Well-designed preclinical models that accurately recapitulate human disease offer an attractive alternative to the study
of rare cancers and may accelerate the process of target
identification and drug discovery and development.
The concept of specialized clinics for selected rare cancers
has proven to be very successful in pediatric gastrointestinal
tumors. This paradigm should be further explored in other rare
cancers because it offers an unprecedented opportunity to
collaborate closely with interested investigators. In addition,
it offers patients an opportunity to discuss their disease with
specialists, allows these patients to provide tissue for further
research, and ultimately can promote the development of
clinical trials that are unique for that specific disease.
with very low registration rates in patients with melanoma

and thyroid carcinoma and higher registration rates in
patients with retinoblastoma, adrenocortical carcinoma, and
nasopharyngeal carcinoma. These differences might be explained in part by referral patterns that are dependent on
the age of the patient. Younger patients with adrenocrtiocal
carcinoma and retinoblastoma require multidisciplinary
care and are likely to be referred to tertiary academic
institutions for treatment. In contrast, adolescents with
thyroid carcinoma or melanoma are treated often with
primary surgical resection by professionals outside of a
pediatric cancer program.
Our initial mandate also included increasing the number
of banked rare tumor biologic specimens available for
future research through a COG banking specimen protocol,
activated in October of 2003. The number of samples received by this repository over a 4-year period totaled 517
snap-frozen specimens, of which only 56 (11%) were considered to belong to a rare cancer. Finally, the COG Rare
Tumor Committee promoted the development of clinical
trials, and two COG-sponsored trials for the treatment of
nasopharyngeal carcinoma and adrenocortical carcinoma
were developed. The enrollment rates for both studies were
less than initially projected, and the statistical sections had
to be modified in order to ensure that these trials could be
completed in a timely manner.3 Thus our preliminary experience with the study of rare cancers within the context of a
pediatric cooperative group was somewhat disappointing
and highlighted the difficulties associated with the study
of this very challenging group of pediatric tumors. Other
opportunities or avenues for research that could potentially
increase our knowledge in rare cancers, as well as methods
for evidence gathering and interpretation must be considered.
From the Division of Oncology, St. Jude Childrens Research Hospital, Memphis, TN.
Address reprint requests to Albert S. Pappo, MD, St. Jude Childrens Research Hospital,
262 Danny Thomas Place, Memphis, TN 38105; email: alberto.pappo@stjude.org.
1092-9118/10/1-10
RARE CANCER IN CHILDREN

Registries
The use of registries to study rare cancers has become

commonplace within the setting of national and international cooperative groups.2-5 For example, TREP in Italy
was launched in 2000 and accrued more than 300 eligible
patients with a variety of histologies over a 6-year period.6
In their experience, thyroid carcinoma, carcinoma of the
appendix, and gonadal germ cell tumors were the most
common tumors registered, and significant underreporting
of various rare cancers was also found among adolescents.6
The TREP project has published valuable information regarding the natural history, treatment, and outcome of
selected rare cancers and has conducted a prospective trial
for the treatment of nasopharyngeal carcinoma.7 In an effort
to expand the registration efforts within the Childrens
Oncology Group, the Childrens Cancer Research Network
(CCRN) was launched in 2007. The purpose of this initiative
was to integrate the registration process for patients with
tumors at COG member institutions into the research of the
group, providing a research resource in North America. The
main purpose of this trial was to obtain informed consent
from patientsinfant and children, adolescent, and young
adults with newly diagnosed cancersand securely record
patient identification and contact information. In addition,
data concerning their cancer and consent for permission for
future contact with the patients for possible epidemiologic
studies is also obtained. In contrast with the CCRN, SEER
does not ascertain all cases of childhood cancer and does not
allow for future contact of research participants, limiting the
availability of epidemiologic studies in these patients. A
preliminary analysis of enrollments of rare cancers to the
CCRN study demonstrates that the registration rates for
adrenocortical carcinoma, thyroid carcinoma, nasopharyngeal carcinoma, and melanoma have not significantly
changed, with approximately two-thirds of the estimated
incidence cases of patients with adrenocortical carcinoma
being enrolled but only approximately 5% of patients with
melanoma being enrolled on the registry. In an attempt to
more efficiently define the natural history and biology or
rare cancers, various groups have developed independent
registries that have provided meaningful insights into the
biology of pediatric rare cancers. The International Pediatric
Adrenocortical Tumor Registry was founded in 1990 and has
KEY POINTS
Rare cancers in children are difficult to study.

Registries and single-arm trials can help develop
standards of care and increase the numbers of biologic specimens, which could further facilitate study
of these tumors.
Preclinical models offer a unique mechanism for
rapidly identifying novel druggable targets and for
prioritizing new therapies.
Specialized clinics can foster collaborations that will
improve the care and study of pediatric rare cancers.
International collaborative efforts must be expanded
in order to speed up progress in the field of pediatric
rare cancers.
described the clinical characteristics and outcome of 254 of

children with this rare disease.8 Their findings provided the
basis for the development of a collaborative cooperative
group trial between COG and Brazilian institutions. This
trial assesses the efficacy of surgery in stage I disease, the
role of retroperitoneal lymph node dissection in stage II
disease, and the role of multimodal chemotherapy with
mitotane and cisplatin-based regimens in advanced-stage
disease. In addition, the trial will continue to expand on the
initial observations of the registry that TP53 germ-line
mutations appear to occur in up to 70% of cases (one-third
being de novo) and that a unique TP53 Arg337 mutation
affects children in Southern Brazil. The latter does not show
the hereditary pattern seen in LiFraumeni syndrome and
has different transactivation activity that may disrupt the
function of p53 in a pH dependent manner.9,10 In another
independent effort, investigators of the pleuropulmonary
blastoma registry have identified heterozygous germ-line
mutations of DICER1, an endoribonuclease that is essential
for processing micro RNAs in 10 families with familial
pleuropulmonary blastoma.11 More recently, the mutational
spectrum of this gene has been expanded to include other
tumors and associations including ovarian Sertoli-Leydig
cell tumors, cystic nephroma, thyroid cysts, multinodular
goiter, and embryonal rhabdomyosarcoma.12-15
Preclinical Models
Well-designed comprehensive preclinical studies using

genetically engineered and orthotopic xenografts that
closely recapitulate the molecular, cellular, and genetic
features of human tumors can be an invaluable tool for
identifying novel targets and promising therapies for rare
cancers. In adults, the use of preclinical models successfully
identified everolimus and sunitinib as active agents for the
treatment of pancreatic neuroendocrine tumors.16,17 Retinoblastoma is a rare cancer of the retina that begins in utero
and is usually diagnosed during the first few years of life.
There are approximately 300 cases in the United States each
year and approximately 5,000 diagnosed world-wide. Preclinical studies have determined that the p53 gene is intact
but the pathway is silenced by overexpression of MDMX.18
Preclinical models have demonstrated that the subconjunctival administration of nutlin-3a, a small molecule inhibitor
of the MDMX-p53 interaction in combination with systemic
topotecan is effective in killing and reducing the tumor
burden of retinoblastoma cells in culture, as well as in
genetic and human orthotopic xenograft models of retinoblastoma.19 These studies provided direct evidence that
targeting the p53 pathway in vivo with MDMX inhibitors is
feasible and provides a potential novel therapy for this
disease. Similarly, using an integrated epigenetic and gene
expression analysis, SYK has been shown to be an important
oncogene in retinoblastoma. Furthermore, the use of SYK
inhibitors such as BAY-613606 and R406 cause cell death;
when combined with topotecan, this combination significantly improves the outcome of mice.20 These observations
suggest that well-designed preclinical studies can identify
novel targets and prioritize the development of new agents
for rare cancers.
Prospective Clinical Trials
The small number of patients available for enrollment in

pediatric rare cancer trials prevents the design of prospec-
587
ALBERTO S. PAPPO
tive randomized trials. However, small, well-designed, and

meaningful clinical trials can yield important information
about the biology, clinical presentation, and treatment options of rare diseases. Additionally, limited trials taht explore novel therapies such as ipilimumab and vemurafenib
for pediatric melanoma should be encouraged. Unfortunately, perhaps as a result of the current infrastructure
constraints, even single-arm cooperative group studies for
rare cancers have had suboptimal enrollment rates.3,21
Specialized Clinics
Highly specialized clinics for rare diseases is not a new

concept; however, rare cancer clinics that address specific
tumors such as the one developed at the NIH for pediatric
gastrointestinal tumors (GIST; www.pediatricgist.cancer.
gov) are unique and provide a novel model for multidisciplinary care of pediatric patients with rare cancers. In this
clinic, physicians and other health care professionals from
various disciplines including oncology, surgery, genetics,
and nutrition gather twice a year to study patients with
pediatric and wild-type GIST. Since its inception in 2008,
more than 90 patients have been seen, 38% of whom are
pediatric patients and 41% are young adults. Tissue has
been obtained from 61 patients and fresh frozen tumor in
five. The clinic has facilitated the study of rare diseases and
has contributed to the understanding of the unique biology
of these tumors. For example, pediatric GIST most commonly affect females, have epithelioid or mixed morphology,
arise in the stomach, have an indolent clinical course, and
rarely have activating mutations of KIT or PDGFR. In
addition the almost universal lack of SDHB expression in
tumor samples from these patients suggests that defects in
cellular respiration may play a pivotal role in the pathogenesis of the disease in younger patients.22
In summary, the study of rare pediatric cancers is challenging. Multiple mechanisms should be explored in order to
advance our understanding of these diseases. The use of
international registries can help identify the numbers of
patients at risk for a specific rare cancer and can aid in the
collection of biologic specimens in this population. Welldesigned, single-arm, collaborative clinical trials that incorporate banking and biologic endpoints can greatly advance
our understanding of these diseases and establish standards
of care for these patient. Preclinical models can more rapidly
aid in the identification of novel druggable targets. Finally,
specialized clinics offer the opportunity to study large numbers of patients by interested individuals facilitating the
study and specimen collection of pediatric rare cancers.
Author
Alberto S. Pappo
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Ziopharm
Oncology
REFERENCES
1. Howlader N, Krapcho M, Neyman N, et al (eds). SEER Cancer Statistics
Review, 1975-2008: Updated November 10, 2011. SEER web site. http://
seer.cancer.gov/csr/1975_2008/index.html. Accessed February 26, 2012.
2. Ferrari A, Bisogno G, De Salvo GL, et al. The challenge of very rare
tumours in childhood: The Italian TREP project. Eur J Cancer. 2007;43(4):
654-659.
3. Pappo AS, Krailo M, Chen Z, et al. Infrequent tumor initiative of the
Childrens Oncology Group: Initial lessons learned and their impact on future
plans. J Clin Oncol. 2010;28(33):5011-5016.
4. Bien E, Godzinski J, Dalligna P, et al. Pancreatoblastoma: A report from
the European cooperative study group for paediatric rare tumours (EXPeRT).
Eur J Cancer. 2011; 47(15):2347-2352.
5. Brecht IB, Graf N, Schweinitz D, et al. Networking for children and
adolescents with very rare tumors: Foundation of the GPOH Pediatric Rare
Tumor Group. Klin Padiatr. 2009;221(3):181-185.
6. Pastore G, De Salvo GL, Bisogno G, et al. Evaluating access to pediatric
cancer care centers of children and adolescents with rare tumors in Italy: The
TREP project. Pediatr Blood Cancer. 2009;53(2):152-155.
7. Casanova M, Bisogno G, Gandola L, et al. A prospective protocol for
nasopharyngeal carcinoma in children and adolescents: The Italian Rare
Tumors in Pediatric Age (TREP) project. Cancer. Epub 2011 Sep 14.
8. Michalkiewicz E, Sandrini R, Figueiredo B, et al. Clinical and outcome
characteristics of children with adrenocortical tumors: A report from the
International Pediatric Adrenocortical Tumor Registry. J Clin Oncol. 2004;
22(5):838-845.
9. DiGiammarino EL, Lee AS, Cadwell C, et al. A novel mechanism of
tumorigenesis involving pH-dependent destabilization of a mutant p53 tetramer. Nat Struct Biol. 2002;9(1):12-16.
10. Wasserman JD, Zambetti GP, Malkin D. Towards an understanding of
the role of p53 in adrenocortical carcinogenesis. Mol Cell Endocrinol. 2012;
351(1):101-110.
588
11. Hill DA, Ivanovich J, Priest JR, et al. DICER1 mutations in familial
pleuropulmonary blastoma. Science. 2009;325(5943):965.
12. Doros L, Yang J, Dehner L, et al: DICER1 Mutations in embryonal
rhabdomyosarcomas from children with and without familial PPB-tumor
predisposition syndrome. Pediatr Blood Cancer. Epub 2011 Dec 16.
13. Schultz KA, Pacheco MC, Yang J, et al. Ovarian sex cord-stromal
tumors, pleuropulmonary blastoma and DICER1 mutations: A report from
the International Pleuropulmonary Blastoma Registry. Gynecol Oncol. 2011;
122(2):246-250. Epub 2011 Apr 17.
14. Rio Frio T, Bahubeshi A, Kanellopoulou C, et al. DICER1 mutations in
familial multinodular goiter with and without ovarian Sertoli-Leydig cell
tumors. JAMA. 2011;305(1):68-77.
15. Heravi-Moussavi A, Anglesio MS, Cheng SW, et al. Recurrent somatic
DICER1 mutations in nonepithelial ovarian cancers. N Engl J Med. 2012;
366(3):234-242.
16. Yao JC, Shah MH, Ito T, et al. Everolimus for advanced pancreatic
neuroendocrine tumors. N Engl J Med. 2011;364(6):514-523.
17. Raymond E, Dahan L, Raoul JL, et al. Sunitinib malate for the
treatment of pancreatic neuroendocrine tumors. N Engl J M. 2011;364(6):
501-513.
18. Laurie NA, Donovan SL, Shih CS, et al. Inactivation of the p53
pathway in retinoblastoma. Nature. 2006;444:61-66.
19. Brennan RC, Federico S, Bradley C, et al. Targeting the p53 pathway in
retinoblastoma with subconjunctival Nutlin-3a. Cancer Res. 2011;71:4205-4213.
20. Zhang J, Benavente CA, McEvoy J, et al. A novel retinoblastoma
therapy from genomic and epigenetic analyses. Nature. 2012;481:329-334.
21. Casanova M, Bisogno G, Gandola L, et al. A prospective protocol for
nasopharyngeal carcinoma in children and adolescents: The Italian Rare
Tumors in Pediatric Age (TREP) project. Cancer. Epub 2011 Sep 14.
22. Janeway KA, Kim SY, Lodish M, et al. Defects in succinate dehydrogenase in gastrointestinal stromal tumors lacking KIT and PDGFRA mutations. Proc Natl Acad Sci U S A. 2011;108:314-318.
Genetic Alterations in Childhood Melanoma

By Fariba Navid, MD
Overview: Melanoma is rare in children. However, the clinical

features of the disease in this population have been welldocumented through single-institution experiences and
population-based analyses. Still, our understanding of the
etiologic factors in children remains unclear and diagnosis of
ESPITE BEING the most common skin cancer in

childhood, melanoma is a rare childhood cancer, making up less than 3% of all cancers seen in children. In the
United States, approximately 300 to 400 new cases per year
are diagnosed in patients younger than age 20; 80% of these
patients are between the ages of 15 and 19.1 The incidence of
melanoma in this older population is rising yearly.2 Most
information about the risk factors, natural history, treatment, and outcome of cutaneous melanoma in children has
been derived from retrospective reports of single-institution
experiences and population-based analyses (i.e., Surveillance Epidemiology and End Results [SEER] data).3-8 The
results of some of these studies seem to indicate that the
natural history and response to therapy of melanoma in
children and adolescents are similar to those in adults and
are stage dependent; however, those of other studies refute
these points and suggest that the biology of the disease
might be different in children than in adults.
Understanding the biologic similarities and differences
between adult and pediatric melanoma may provide insight
into the causative factors of the disease in children and
influence treatment decisions. Such knowledge might also
aid clinicians dealing with diagnostically challenging issues
such as distinguishing between Spitz nevi and spitzoid
melanoma or between malignant transformation and benign
proliferation in congenital nevi. Emerging evidence suggests
that molecular characterization of these lesions may be
helpful. This article reviews common genetic alterations in
pediatric melanoma and their potential utility in differentiating histologically challenging cases.
Common Genetic Alterations
During the past decade, several genetic alterations in

adult melanomas have been described.9-12 These changes
are summarized in Table 1. Characterization of these aberrations suggests that various distinct molecular pathways
are associated with the development of melanoma on the
basis of the tumor site, host phenotype, and amount of sun
exposure.13,14 These findings highlight the heterogeneity of
adult melanomas and raise the possibility that the same
genetic alterations may not be present in childhood melanoma. Large-scale genetic profiling studies of pediatric melanoma performed by using comparative genomic
hybridization (CGH) or genome-wide association techniques
have not been reported and may not be feasible because of
the tumors rarity. However, some genes have been evaluated in the pediatric population; findings about two that are
mutated in this setting are summarized below.
BRAF
The RAS/RAF/MAPK signaling pathway is the one most

commonly implicated in the pathogenesis of melanoma and
melanoma remains challenging in certain cases. This article

reviews emerging evidence indicating that molecular characterization of these lesions in children may be of diagnostic and
therapeutic value.
is the focus of recent drug development efforts. This pathway is dysregulated in approximately 50% of adult melanomas, frequently having mutations in the BRAF gene (less
frequently NRAS). The most common BRAF mutation in
melanoma is a substitution of glutamic acid for valine at
position 600 (BRAFV600E). This mutation accounts for more
than 90% of the BRAF mutations that occur in adult
melanomas and drives cell proliferation, invasion, and metastasis in these tumors. The U.S. Food and Drug Administration recently approved vemurafenib, an oral tyrosine
kinase inhibitor of mutated BRAF, because of its favorable
response rates in adult patients with BRAF-mutated melanoma.15,16 BRAFV600E mutations in pediatric melanoma
have only been reported from one small study showing the
mutation in five of 10 pediatric tumors.17 Finding
BRAFV600E and other aberrations of this pathway in more
samples would provide a rationale to test selective inhibitors
targeting this mutation and other RAS/RAF/MAPK signaling molecules in pediatric melanoma.
CDKN2A
CDKN2A, located on chromosome 9p21, encodes two distinct tumor suppressor genes, p16/INK4a and p14/ARF,
that play a key role in cell cycle regulation. The CDKN2A
locus is deleted in approximately 50% of sporadic adult
melanomas.13 Daniotti and colleagues17 found homozygous
CDKN2A deletions in nine of 14 pediatric melanomas (patients aged 2 to 19). Alterations in CDKN2A are also
implicated in the pathogenesis of 25% to 40% of familial
cutaneous melanomas.18 In these patients, melanoma tends
to develop when they are younger but not commonly when
they are younger than age 18. Whiteman and colleagues19
analyzed DNA from 31 children in the Queensland Cancer
registry who were younger than age 15. One patient among
the 10 with a family history of melanoma had a germ-line
mutation in CDKN2A; she had a history of two primary
melanomas before the age of 13. Among 147 adolescents in
the same registry who were aged 15 to 19, two had germ-line
alterations in CDKN2A; however, neither had a family
history of melanoma.20 In a retrospective review of 15
Swedish families with known germ-line CDKN2A mutations, the youngest family member to develop melanoma
was 18 years.21 The results of these studies suggest that
From the Division of Solid Malignancies, Department of Oncology, St. Jude Childrens
Research Hospital, Memphis, TN.
Address reprint requests to Fariba Navid, MD, Division of Solid Malignancies, Department of Oncology, St. Jude Childrens Research Hospital, 262 Danny Thomas Pl., Memphis,
TN 38105-2794; email: fariba.navid@stjude.org.
1092-9118/10/1-10
589
FARIBA NAVID
Table 1. Common Genetic Alterations in Adult Melanoma
Gene
Chromosomal
Locus
Alteration
Frequency (%)
Implicated in
Familial
Melanoma?
CDKN2A
CDK4
CCND1
BRAF
NRAS
MITF
MC1R
PTEN
APAF1
TP53
C-KIT
9p21
12q13
11q13
7q34
1p13
3p14
16q24
10q23
12q22
17p13
4q11
Mutated, deleted, silenced

Amplified or mutated
Amplified
Mutated
Mutated
Amplified or mutated
Mutated
Deleted or mutated
Silenced
Mutated, deleted
Mutated
3070
12
644
5070
1530
1016
?
520
40
10
2040
Yes
Yes
No
No
No
Yes
Yes
No
No
No
No
Abbreviations: APAF1, apoptotic protease-activating factor 1; BRAF, v-raf

murine sarcoma viral oncogene homolog B1; CCND1, cyclin D1; CDKN2A,
cyclin-dependent kinase inhibitor 2A; CDK4, cyclin-dependent kinase 4; C-KIT,
stem cell growth factor receptor; MC1R, melanocortin-1 receptor; MITF,
microphthalmia-associated transcription factor; NRAS, v-ras neuroblastoma RAS
viral oncogene homolog; PTEN, phosphatase and tensin homologue; TP53, tumor
protein p53.
CDKN2A alterations may play a role in a small subset of

melanomas that develop in those younger than 18 years.
Distinguishing between Spitz Nevi and
Spitzoid Melanoma
A Spitz nevus usually manifests in childhood and is

considered to be a benign melanocytic lesion. The classic
Spitz is described as a small, symmetric dome-shaped lesion;
it is typically nonpigmented and consists of epithelioid
and/or spindle-shaped cells with evidence of maturation at
the base. The presence of a high mitotic rate, pagetoid
spread, lack of maturation, pleomorphism and nuclear
atypia, ulceration, and giant cells have all been cited as
discriminating features between Spitz nevi and melanomas;
however, there is no consensus about these characteristics
among pathologists.22 Therefore, in certain cases, a specific
diagnosis of benign Spitz nevi versus malignant melanoma
is difficult to make on the basis of histologic features alone,
and no immunohistochemical markers reliably differentiate
KEY POINTS
590
The results of molecular studies may aid clinicians in

the diagnosis of melanocytic lesions that are histologically difficult to classify in children.
In contrast to malignant melanoma, Spitz nevi rarely
contain chromosomal aberrations.
Benign proliferations arising in congenital nevi have
a different pattern of chromosomal aberrations than
do proliferations arising in malignant melanoma.
Development of melanoma in individuals harboring
germ-line mutations in CDKN2A usually occurs in
adulthood and is uncommon in children and adolescents.
A better understanding of the similarities and differences between genetic alterations in pediatric melanomas and those in adult melanoma may accelerate
the use of promising or targeted therapies in the
pediatric population.
between the two neoplasms. Therefore, lesions are often

referred to as an atypical Spitz nevus or spitzoid tumor of
uncertain malignant potential.23
Chromosomal Aberrations
To better distinguish between melanoma and benign

melanocytic nevi, Bastian and colleagues24 performed a
CGH study of 54 benign nevi and 132 melanomas. Their
findings showed that 127 melanomas (96%) had chromosomal aberrations, whereas only seven (13%) of the benign
nevi had aberrations. All seven benign cases were Spitz nevi.
In six of these, the chromosomal abnormality was restricted
to a gain in the short arm of chromosome 11, an aberration
that was not present in any of the melanoma samples. In a
study of 102 Spitz nevi, fluorescent in situ hybridization
results showed that 12 had a gain in chromosome 11.
Furthermore, sequencing analysis showed that eight of the
12 had a mutated HRAS allele on the gained copy of
chromosome 11.25 Since this finding, two independent investigators have found that HRAS mutations are exclusive to
Spitz and atypical Spitz nevi and are not present in melanoma.26,27
RAS/RAF/MAPK Pathway Alterations
The RAS/RAF/MAPK signaling pathway, which can lead

to cell growth and division, is activated in Spitz nevi with or
without HRAS mutations. Despite the growth stimulatory
effects of this pathway, both lesions have a low proliferative
index as assessed by MIB-1 staining. The growth inhibitory
signal that prevents malignant transformation in these cells
is hypothesized to be a high expression of intact p16.28 In
line with this hypothesis, Dhaybi and colleagues29 detected
no p16 expression in six cases of spitzoid melanoma from
children aged 3 to 14; however, strong p16 expression was
detected in 18 Spitz nevi excised from children aged 1 to 14
and in 12 benign nevi.
Researchers have also investigated the RAS/RAF/MAPK
pathway genes BRAF and NRAS in Spitz nevi and spitzoid
melanoma. Gill and colleagues26 found no mutations in
either of these genes in nine spitzoid melanomas and 10
age-matched Spitz nevi from prepubescent children (aged 2
to 10). A follow-up study of 33 spitzoid melanomas in varying
age groups showed mutated BRAF in only one lesion and no
NRAS-mutated lesions.30 The authors concluded that the
absence of the frequently mutated BRAF and NRAS genes
in spitzoid melanoma suggests that these tumors are biologically different from conventional nonspitzoid melanomas.
However, although others have also reported the absence of
BRAF mutations in most Spitz nevi, BRAF and NRAS
mutations have been detected in spitzoid melanomas by
other researchers.27,31-33 The reasons for the conflicting
results are unclear. In these studies and others, mutations
in HRAS, NRAS, and BRAF appear to be mutually exclusive. Thus, what we can conclude from the mutational
analysis of melanocytic lesions is that a lesion with a HRAS
mutation is unlikely to be a melanoma; however, other
features of a lesion with a NRAS or BRAF mutation must be
taken into account to make a diagnosis of melanoma.
Congenital Nevi and Malignant Transformation
Congenital nevi are pigmented lesions that are present at

birth or appear shortly thereafter, having an overall preva-
GENETIC ALTERATIONS IN CHILDHOOD MELANOMA
lence of approximately 1%.34 The risk of malignant transformation of congenital nevi increases with the size of the
nevus. In a review of 289 published cases of large congenital
nevi (more than 20 cm in diameter), melanoma developed
within a congenital nevus in 34 patients (12%).35 The
clinical or molecular factors that lead to the transformation
of these lesions are unknown.
Secondary Proliferations
Secondary proliferations within a congenital nevus can

occur and are often difficult to histologically classify as being
malignant or benign. The results of CGH analysis of 29
congenital nevi and associated benign and malignant proliferations showed no aberrations in the typical congenital
nevi, in congenital nevi of increased cellularity, or in congenital nevi with benign proliferation. Additionally, seven of
nine cases of congenital melanocytic nevi with proliferations
simulating nodular melanoma predominately contained either gains or losses of entire chromosomes. This finding
differs from that in melanoma in which fragments of various
chromosomes are detected. The pattern of chromosomal
abnormalities between the nodular lesions and melanoma
was also different. In contrast, the six cases of melanoma
arising in congenital nevi had multiple cytogenetic aberrations in a pattern that was indistinguishable from that of
melanomas not associated with congenital nevi.36 These
findings suggest that the type of genomic instability in
atypical nodular proliferations differs from that predominat-
ing in melanoma and that CGH may be a useful tool in

distinguishing between the two entities.
RAS/RAF/MAPK Pathway Alterations
Some genes reported to be associated with melanoma,

including p53, p16, and CDK4, do not appear to be altered in
these proliferative lesions within a congenital nevi.37 NRAS
mutations, however, are frequent in congenital nevi: 26 of 32
cases of congenital nevi in one series and 10 of 17 cases in
another had NRAS mutations.37,38 Reports of BRAF mutations in congenital nevi are conflicting. For example, Yazdi
and colleagues32 and Pollock and colleagues39 detected
BRAF mutations in six of 13 congenital nevi and six of seven
congenital nevi, respectively. In contrast, Bauer and colleagues38 found no BRAF mutations in a series of 32 cases of
congenital nevi. They speculate that the discrepancy between their findings and others may be due to selection bias
because they chose only lesions present at birth, but others
may have also considered nevi appearing in the first year of
life. It is well-documented that more than 80% of acquired
nevi have BRAF mutations.39 The absence of BRAF mutations would support the hypothesis that nevi that develop in
utero, in the absence of ultraviolet exposure, are genetically
distinct from nevi that develop after birth.
A better understanding of the embryonic development of
congenital nevi and their transformation to benign and
malignant lesions may provide some insight into the etiology
of melanoma that develops in childhood in the absence of
these lesions.
Author
Fariba Navid
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Schering-Plough
Expert
Testimony
Other
Remuneration
REFERENCES
1. Bernstein L, Gurney, JG. Carcinomas and other malignant epithelial
neoplasms. In Ries LG, Smith M, Gurney J, et al (eds). Cancer Incidence and
Survival among Children and Adolescents: United States SEER Program
1975-1995. Bethesda, MD: National Cancer Institute, 1999;139-147.
2. Herzog C, Pappo AS, Bondy M, et al. Malignant melanoma. In Bleyer A,
OLeary M, Barr R, et al (eds). Cancer Epidemiology in Older Adolescents and
Young Adults 15 to 19 Years of Age, Including SEER Incidence and Survival:
1975-2000. Bethesda, MD: National Cancer Institute, 2006;53-63.
3. Moore-Olufemi S, Herzog C, Warneke C, et al. Outcomes in pediatric
melanoma: comparing prepubertal to adolescent pediatric patients. Ann Surg.
2011;253:1211-1215.
4. Paradela S, Fonseca E, Prieto VG. Melanoma in children. Arch Pathol
Lab Med. 2011;135:307-316.
5. Aldrink JH, Selim MA, Diesen DL, et al. Pediatric melanoma: a singleinstitution experience of 150 patients. J Pediatr Surg. 2009;44:1514-1521.
6. Strouse JJ, Fears TR, Tucker MA, et al. Pediatric melanoma: risk factor
and survival analysis of the surveillance, epidemiology and end results
database. J Clin Oncol. 2005;23:4735-4741.
7. Lange JR, Palis BE, Chang DC, et al. Melanoma in children and
teenagers: an analysis of patients from the National Cancer Data Base. J Clin
Oncol. 2007;25:1363-1368.
8. Livestro DP, Kaine EM, Michaelson JS, et al. Melanoma in the young:
differences and similarities with adult melanoma: a case-matched controlled
analysis. Cancer. 2007;110:614-624.
9. Gaudi S, Messina JL. Molecular bases of cutaneous and uveal melanomas. Patholog Res Int. 2011;2011:159421.
10. High WA, Robinson WA. Genetic mutations involved in melanoma: a
summary of our current understanding. Adv Dermatol. 2007;23:61-79.
11. Romano E, Schwartz GK, Chapman PB, et al. Treatment implications
of the emerging molecular classification system for melanoma. Lancet Oncol.

2011;12:913-922.
12. Yokoyama S, Woods SL, Boyle GM, et al. A novel recurrent mutation in
MITF predisposes to familial and sporadic melanoma. Nature. 2011;480:99103.
13. Curtin JA, Fridlyand J, Kageshita T, et al. Distinct sets of genetic
alterations in melanoma. N Engl J Med. 2005;353:2135-2147.
14. Bauer J, Buttner P, Murali R, et al. BRAF mutations in cutaneous
melanoma are independently associated with age, anatomic site of the
primary tumor, and the degree of solar elastosis at the primary tumor site.
Pigment Cell Melanoma Res. 2011;24:345-351.
2011;364:2507-2516.
17. Daniotti M, Ferrari A, Frigerio S, et al. Cutaneous melanoma in
childhood and adolescence shows frequent loss of INK4A and gain of KIT.
J Invest Dermatol. 2009;129:1759-1768.
18. Kefford RF, Newton Bishop JA, Bergman W, et al. Counseling and DNA
testing for individuals perceived to be genetically predisposed to melanoma: a
consensus statement of the Melanoma Genetics Consortium. J Clin Oncol.
1999;17:3245-3251.
19. Whiteman DC, Milligan A, Welch J, et al. Germline CDKN2A mutations in childhood melanoma. J Natl Cancer Inst. 1997;89:1460.
20. Youl P, Aitken J, Hayward N, et al. Melanoma in adolescents: a
case-control study of risk factors in Queensland, Australia. Int J Cancer.
2002;98:92-98.
21. Magnusson S, Borg A, Kristoffersson U, et al. Higher occurrence of
591
FARIBA NAVID
childhood cancer in families with germline mutations in BRCA2, MMR and
CDKN2A genes. Fam Cancer. 2008;7:331-337.
22. Barnhill RL, Argenyi ZB, From L, et al. Atypical Spitz nevi/tumors:
lack of consensus for diagnosis, discrimination from melanoma, and prediction of outcome. Hum Pathol. 1999;30:513-520.
23. Barnhill RL. The Spitzoid lesion: rethinking Spitz tumors, atypical
variants, Spitzoid melanoma and risk assessment. Mod Pathol. 2006;19:S21S33 (suppl 2).
24. Bastian BC, Olshen AB, LeBoit PE, et al. Classifying melanocytic
tumors based on DNA copy number changes. Am J Pathol. 2003;163:17651770.
25. Bastian BC, LeBoit PE, Pinkel D. Mutations and copy number increase
of HRAS in Spitz nevi with distinctive histopathological features. Am J
Pathol. 2000;157:967-972.
26. Gill M, Cohen J, Renwick N, et al. Genetic similarities between Spitz
nevus and Spitzoid melanoma in children. Cancer. 2004;101:2636-2640.
27. van Dijk MC, Bernsen MR, Ruiter DJ. Analysis of mutations in B-RAF,
N-RAS, and H-RAS genes in the differential diagnosis of Spitz nevus and
spitzoid melanoma. Am J Surg Pathol. 2005;29:1145-1151.
28. Maldonado JL, Timmerman L, Fridlyand J, et al. Mechanisms of
cell-cycle arrest in Spitz nevi with constitutive activation of the MAP-kinase
pathway. Am J Pathol. 2004;164:1783-1787.
29. Al Dhaybi R, Agoumi M, Gagne I, et al. P16 expression: a marker of
differentiation between childhood malignant melanomas and Spitz nevi. J Am
Acad Dermatol. 2011;65:357-363.
30. Lee DA, Cohen JA, Twaddell WS, et al. Are all melanomas the same?
592
Spitzoid melanoma is a distinct subtype of melanoma. Cancer. 2006;106:907913.

31. Palmedo G, Hantschke M, Rutten A, et al. The T1796A mutation of the
BRAF gene is absent in Spitz nevi. J Cutan Pathol. 2004;31:266-270.
32. Yazdi AS, Palmedo G, Flaig MJ, et al. Mutations of the BRAF gene in
benign and malignant melanocytic lesions. J Invest Dermatol. 2003;121:11601162.
33. Fullen DR, Poynter JN, Lowe L, et al. BRAF and NRAS mutations in
spitzoid melanocytic lesions. Mod Pathol. 2006;19:1324-1332.
34. Castilla EE, da Graca Dutra M, Orioli-Parreiras IM. Epidemiology of
congenital pigmented naevi: I. Incidence rates and relative frequencies. Br J
Dermatol. 1981;104:307-315.
35. DeDavid M, Orlow SJ, Provost N, et al. A study of large congenital
melanocytic nevi and associated malignant melanomas: review of cases in the
New York University Registry and the world literature. J Am Acad Dermatol.
1997;36:409-416.
36. Bastian BC, Xiong J, Frieden IJ, et al. Genetic changes in neoplasms
arising in congenital melanocytic nevi: differences between nodular proliferations and melanomas. Am J Pathol. 2002;161:1163-1169.
37. Papp T, Pemsel H, Zimmermann R, et al. Mutational analysis of the
N-ras, p53, p16INK4a, CDK4, and MC1R genes in human congenital melanocytic naevi. J Med Genet. 1999;36:610-614.
38. Bauer J, Curtin JA, Pinkel D, et al. Congenital melanocytic nevi
frequently harbor NRAS mutations but no BRAF mutations. J Invest Dermatol. 2007;127:179-182.
39. Pollock PM, Harper UL, Hansen KS, et al. High frequency of BRAF
mutations in nevi. Nat Genet. 2003;33:19-20.
Desmoid-Type Fibromatosis in Children:

A Step Forward in the Cooperative
Group Setting
By Natalie Pounds, MD, and Stephen X. Skapek, MD
Overview: Desmoid-type (aggressive) fibromatosis (desmoid

tumor) is a soft tissue neoplasm that can occur in both
children and adults. Although it is formally classified as an
intermediate-grade neoplasm because of its propensity for
locally invasive growth, it can lead to severe and lifethreatening problems. Because metastases do not arise from
desmoid tumor, therapeutic interventions have historically
focused on surgery or radiation to achieve local tumor control.
These approaches may be ineffective or impractical for some
children. In those cases, systemic therapy with cytotoxic or
noncytotoxic therapy has been used. Because of the relative
ESMOID TUMOR, also known as aggressive or

desmoid-type fibromatosis, represents a relatively rare
neoplasm that affects both children and adults. The overall
incidence of desmoid tumor is estimated to be two to four
new diagnoses per 1 million people per year.1 Desmoid
tumor incidence peaks in individuals from 6 to 15 years of
age and again between puberty and 40 years of age in
women. There seems to be a female predominance during
adolescence.2,3 Mortality from desmoid tumor is rare but has
been reported in children. Nonetheless, the disease poses
substantial problems related to diseases progression and
consequences from therapy.
The most common sites of origin of desmoid tumor are in
the abdominal wall, intra-abdominal or mesenteric sites, but
extra-abdominal sites include extremities, shoulder girdle,
chest wall, and inguinal region.1,4 Children tend to develop
desmoid tumors in similar extra-abdominal locations, but
both infants and young children have a higher propensity for
tumors in the head and neck region,2 a rare disease site for
adults. Patients with intra-abdominal desmoid tumors typically are asymptomatic or have symptoms associated with
an enlarging mass, including weight loss, cachexia, malaise,
renal failure, and small bowel compression or perforation.
Invasion of adjacent muscle, nerve, or vessels can cause
pain, limitation in joint movement, or contractures and
deformities. Desmoid tumor may be multifocal, but bona-fide
metastasis does not occur.1,4
Desmoid tumor poses a substantial clinical problem because it carries a propensity to recur after seemingly complete surgical resection. The reason for the high recurrence
rate is apparent from histologic studies. Desmoid tumor has
a benign histologic appearance, lacking nuclear and cytoplasmic features of a malignant tumor.1 Unlike most sarcomas, which are usually separated from adjacent structures
by a pseudocapsule, desmoid tumor often displays an irregular, infiltrating border. Several histologic differences have
been noted between desmoid tumors in children and adults.
First, higher mitotic rates are seen in childhood tumors.
Second, lesions tend to be more cellular.5 Whether rare
mitotic activity contributes to their relative resistance to
cytotoxic therapy is not clear.
Desmoid tumor is typically a sporadic disease with no
identified cause. However, it is clear that there may be an
underlying genetic component in disease pathogenesis. This
rarity of this neoplasm in children, knowledge on the use of

chemotherapy is based largely on anecdotal reports or retrospective series. Limited conclusions can be drawn, though,
from these types of reports. In the last 10 years, two prospective phase II clinical trials of chemotherapy for children with
desmoid tumor have been conducted in cooperative clinical
trials centered in North America. We review the results of
those clinical trials and suggest future directions for systematically approaching this disease to better define the role of
chemotherapy for children with desmoid tumor.
genetic component is most evident in Gardner syndrome, a

well-described variant of familial adenomatous polyposis
(FAP), an autosomal-dominant disease characterized by the
presence of innumerable adenomatous polyps in the colon
(OMIM #175100). Patients with Gardner syndrome have a
number of extracolonic manifestations, including osteoma
and desmoid tumor. Both FAP and Gardner syndrome are
associated with germline mutation in the adenomatous
polyposis coli (APC) gene.6 The APC protein acts, in part, to
promote degradation of -catenin, an intracellular protein
that aids in the transduction of cell proliferation signals to
the nucleus. A number of studies during the past decade
have identified somatic mutations in APC and CTNNB1, the
gene encoding -catenin, in sporadic desmoid tumor cases7;
both mutations ultimately enhance -catenin activity. Despite an association with Gardner syndome, heritable predisposition is clearly the less common form of desmoid tumor.
Other causative factors include trauma and estrogens.
The former is typically surgical in nature, and it may be
most relevant to FAP patients. The high frequency of abdominal disease in this population seems to correlate with
mesenteric fibromatosis, a putative precursor lesion the
presence of which correlates with increased numbers of
abdominal operations.8 The role of estrogen signaling was
originally suggested by the increased incidence during pregnancy and enhanced tumor growth rate in pregnant women
when compared with that in men or premenopausal or
postmenopausal women.4 This association is supported by
evidence of strong estrogen receptor expression in more than
80% of desmoid tumors.9
Surgery and Radiation as the Historical Standard
of Care
Although we make a case that optimal therapy for desmoid tumor in a child begins with a multidisciplinary
From the University of Texas Southwestern Medical Center and Center for Cancer and
Blood Disorders, Childrens Medical Center, Dallas, TX.
Address reprint requests to Stephen X. Skapek, MD, University of Texas Southwestern
Medical Center, 5323 Harry Hines Blvd, Mail Code 9063, Dallas, TX-9063; email: Stephen.
Skapek@utsouthwestern.edu.
1092-9118/10/1-10
593
POUNDS AND SKAPEK
evaluation, because desmoid tumor is a largely localized

disease, treatment approaches have historically focused on
focal therapies, such as surgery or external beam ionizing
radiation.
Most would agree that surgical resection represents the
best therapy for children in whom the tumor can be completely removed without substantial functional or cosmetic
consequences. The ability to obtain a complete surgical
resection is generally believed to be an important factor that
influences tumor control. However, other factors, such as
whether the disease is primary or recurrent10 and the
presence of certain CTNNB1 mutations,7 also seem to be
important. Numerous issues regarding ongoing physical and
emotional development in children add to the complexity
regarding long-term surgical outcomes in children with
desmoid tumor; and this has not been well studied in this
patient population.
Ionizing radiation therapy has been shown to be effective
in many studies of adult patients. In particular, when
applied in conjunction with surgery, ionizing radiation
therapy improves local disease control.11 The high radiation dose needed makes its use more complicated if the
radiation port includes growing bones or joints. Furthermore, radiation therapy may be less effective in children
than adults. This concern led to a retrospective review,
which concluded that the role of radiation in the initial
treatment of children with desmoid tumors should be reconsidered.12 In this series, five of 13 children received radiation therapy immediately after diagnosis, whereas the
remaining eight patients received radiation therapy after
recurrence. The median dose of radiation used was 50 Gy.
Tumor recurred after radiation therapy in 11 of the 13
patients, and three patients died of their disease. In those
surviving, substantial morbidity associated with the radia-
KEY POINTS
594
Desmoid-type fibromatosis (desmoid tumor) represents a locally aggressive soft tissue neoplasm that
has a propensity for locally aggressive growth and
recurrence after attempted surgical resection.
Desmoid tumor is associated with loss of function
mutations in the adenomatous polyposis coli tumor
suppressor and gain of function mutations in
-catenin.
Surgical resection with or without radiation therapy
has represented the historical standard for treating
this disease.
A variety of cytotoxic and noncytotoxic chemotherapies have demonstrated the capacity to control desmoid tumor when surgery or radiation are not
effective; however, most of this literature is in the
form of case reports or retrospective studies, often
including children and adults.
Two multicenter, prospective clinical trials conducted
through the Pediatric Oncology Group and Childrens
Oncology Group have demonstrated the feasibility of
this approach and the added value stemming from
systematic, prospective study.
Table 1. Event-Free Survival for Children Enrolled in POG 9650

or COG ARST0321 and Treated With Vinblastine/Methotrexate or
Tamoxifen/Sulindac, Respectively
Chemotherapy
Event-Free Survival at 1 Year (Range)
Vinblastine/methotrexate
Tamoxifen/sulindac
0.58 (0.290.57)
0.44
Abbreviations: COG, Childrens Oncology Group; POG, Pediatric Oncology

Group.
tion therapy was noted. A more recent study, though,

supports the role that radiation therapy can play in this
disease in children.13
Like surgery, radiation therapy holds the potential for
consequences that can be particularly troubling in a growing
child. Reported adverse effects include bone fractures, skeletal and soft tissue growth retardation, tissue fibrosis, and
lymphedema.12 Jabbari and colleagues13 also reported substantial morbidity in the form of peripheral neuropathy,
pain, bowel obstruction, and the development of papillary
cancer.
Chemotherapy for Children with Desmoid Tumor
For many children with desmoid tumor, surgery or radiation has proven ineffective or is believed to be associated
with unacceptable consequences, which has led to the use of
chemotherapeutic agents that are believed to act by cytotoxic or noncytotoxic mechanisms. Regrettably, most of the
reports stemming from their use represent retrospective,
single-institution analyses of small groups of patients,
thereby limiting the conclusions that one can draw.
Cytotoxic regimens demonstrated to have some activity in
retrospective analyses include liposomal doxorubicin14;
doxorubicin with dacarbazine15; vincristine, dactinomycin,
and cyclophosphamide16; hydroxyurea17; and vinblastine
and methotrexate.18 Perhaps the most striking report centered on the use of doxorubicin and dacarbazine by 96-hour
infusion and followed by the nonsteroidal anti-inflammatory
drug (NSAID) meloxicam.19 The authors reported complete
and partial responses in three and four, respectively, of 11
adult patients, with a mean progression-free survival of
approximately 6 years. Beyond the biases inherent in retrospective analyses, the complex nature of desmoid tumor,
which has been reported to undergo prolonged stabilization
and even spontaneous regression,20 limits the conclusions
one can draw on the relative efficacy of any of these regimens.
The same is true of reports of noncytotoxic regimens, most
of which include NSAIDs, such as sulindac and meloxicam,
and estrogen antagonists, such as tamoxifen. In some cases,
tamoxifen21 or NSAIDs22 are used alone. In most, though,
tamoxifen is combined with an NSAID, such as sulindac.23
In one series, 10 of 13 adults with FAP-associated disease
had either a partial or complete response.23 Although the
initial rationale for using NSAIDs may have been based on
incomplete understanding of the disease biology, more recent laboratory studies provide a potential mechanism: the
transcription factor PPAR is deregulated in the setting of
APC mutations, but NSAIDs can block PPAR activity.24
Receptor tyrosine kinase inhibitors, particularly imatinib
mesylate, also have shown the capacity to stabilize desmoid
tumor and foster regression in a smaller subset.25 Its activity may be based on expression of platelet-derived growth
factor receptor .25
CHEMOTHERAPY FOR DESMOID TUMOR

Table 2. Complete and Partial Responses for Children Enrolled
in POG 9650 or COG ARST0321 and Treated With Vinblastine/
Methotrexate or Tamoxifen/Sulindac, Respectively
Chemotherapy
Complete and Partial

Responses, No.
Total, No.
CR/PR Rate, %
Vinblastine/methotrexate
Tamoxifen/sulindac
5
5
26
60
19
8
Abbreviations: COG, Childrens Oncology Group; POG, Pediatric Oncology

Group; CR, complete response; PR, partial response.
Chemotherapy Lessons from Prospective, Cooperative

Group Trials
The reports discussed in this article are not fully satisfying, given the long-standing concept that desmoid tumor
may remain stable or undergo spontaneous regression.4
Indeed, a recent retrospective report suggests that
progression-free survival is the same in patients treated
with medical therapy or a wait and see approach; nearly
50% had no disease progression at 5 years.26 The variable
natural history of desmoid tumor spurred members of the
former Pediatric Oncology Group (POG) to prospectively
study desmoid tumor in the cooperative group setting.
Vinblastine and Methotrexate
At the time this study was conceived, Weiss and Lackman18 had published a retrospective study on the use of
vinblastine and methotrexate in adult patients with recur-
rent desmoid tumors. In this report of eight adult patients,

two had complete and four had partial responses. A similar
retrospective study of these two agents in 10 children
showed a 50% response rate and 70% disease stabilization.27
Vinblastine and methotrexate were particularly appealing
for prospective study in children because of a favorable
late effects profile when compared with doxorubin- and
cyclophosphamide-based regimens.
The POG 9650 study represented the first prospective,
multi-institutional trial in children.28 The primary goal was
to estimate the safety and efficacy of vinblastine and
methotrexate in patients younger than 19 years with
desmoid-type fibromatosis that was recurrent or not amenable to surgery or radiation. In this single-arm, phase II
study, both drugs were given weekly for 26 weeks and then
every other week for an additional 26 weeks, at which point
therapy was terminated. Response was assessed by bidimensional measurements on axial imaging using either
computed tomography or magnetic resonance imaging,
and responses were prospectively defined in the study.
Of 26 patients, measurable response was evident in eight
patients (one with complete response, four with partial
response, and three with minor response), whereas an
additional 10 patients had stable disease as the best response (Table 1). The 1-year event-free survival rate was
58%, and the median time to progression was 15.9 months
after therapy was discontinued (Table 2). Seven patients
Fig 1. Potential strategy for risk-based treatment of

desmoid tumor in children (A) and how this type of
approach might be compared with chemotherapy in a
clinical trial (B).
Abbreviations: SD, Stable Disease; R, Response; PD,
Progressive Disease.
595
POUNDS AND SKAPEK
(27% of total) were free of disease progression 50 months

after enrollment.
This regimen was chosen because of the anticipated low
toxicity rate. In fact, 18 patients (67%) reported National
Cancer Institute (NCI) grade 2 or higher toxicity; neutropenia represented the most common grade 3 or 4 toxicity.28 Other toxicities included anemia, nausea, vomiting,
and elevation of hepatic transaminases; all toxicities
were reversed on ceasing chemotherapy. These results suggested that vinblastine and methotrexate administration
arrested tumor progression or promoted frank regression
in most children, but the treatment was associated with
substantial myelosuppression. Perhaps most importantly,
this study set a precedent for prospectively studying a
relatively rare neoplasm in children in the cooperative group
setting.
Sulindac and High-Dose Tamoxifen
While the results of POG 9650 were maturing, the Childrens Oncology Group (COG) supported a follow-up, singlearm, phase II trial of sulindac and high-dose tamoxifen
based on the previously mentioned retrospective studies and
case reports. Like the POG 9650 study, the primary aim of
the ARST0321 was to estimate the safety and efficacy of
sulindac and tamoxifen in the same patient population and
using the same response criteria.29 Secondary objectives
investigated whether magnetic resonance imaging signal
features correlate with response or nonresponse, as previously suggested,27 and whether hormone receptor status
influences the outcome. Patients received sulindac and tamoxifen in combination for 12 4-week cycles or until disease
progression. After a complete response, patients were to
receive an additional 1 month of the study drugs.
A total of 70 patients were enrolled in the study, with an
annual accrual rate of 13.3 patients per year. The 61 eligible
patients included 22 with newly diagnosed disease and 39
with recurrent disease. A response (complete response or
partial response) was observed in five patients (8%), and
progression-free survival at 1 year was 44% (Table 1 and

Table 2). In contrast to vinblastine/methotrexate, myelosuppression was rarely observed; and patients rarely experienced NCI grade 3 or 4 toxicities. One notable exception was
that 11 of the 30 eligible female patients (27%) had large
ovarian cysts documented by ultrasonography; although not
believed to be a medically severe consequence, the ovarian
cysts interfered with treatment in most of the cases.
Conclusion
Although these two successive cooperative group studies

have more accurately defined radiographic response rates
and progression-free survival for children treated with
vinblastine/methotrexate and tamoxifen/sulindac, the trials
are not without their own pitfalls. Because these studies
were single arm, phase II studies, we cannot confidently
state whether one regimen is better than the other, even
though the eligibility criteria and response definitions were
the same. We must also be cautious about attributing a
cause-effect relationship to either treatment regimen. Such
caution is particularly important when one considers reports
that desmoid tumor can remain stable for extended periods
with only a wait and see approach.20,26
These reports have led to the concept that asymptomatic
patients should initially by treated with a period of
observation. However, such an approach, illustrated in
Fig. 1A, is unproven in children. Conceivably, one could
prospectively test this concept while also designing a
study to more definitively prove that chemotherapy plays a
role in disease control. Options to consider might include
studies randomizing two regimens or a study design in
which an observation period is incorporated as a window
for those with asymptomatic and nonlife-threatening disease (Fig. 1B). Building on the POG and COG success
performing prospective trials for this disease, we are optimistic that we can take the needed steps to better define the
natural history and role of chemotherapeutics for children
with desmoid tumor.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Natalie Pounds*
Stephen X. Skapek*
REFERENCES
1. Fletcher CDM, Unni KK, Mertens F (eds). World Health Organization
Classification of Tumours. Pathology and Genetics of Soft Tissue and Bone.
Lyons, France: IARC Press; 2002.
2. Buitendijk S, van de Ven CP, Dumans TG, et al. Pediatric aggressive
fibromatosis: a retrospective analysis of 13 patients and review of literature.
Cancer. 2005;104:1090-1099.
3. Faulkner LB, Hajdu SI, Kher U, et al. Pediatric desmoid tumor:
retrospective analysis of 63 cases. J Clin Oncol. 1995;13:2813-2818.
4. Hayry P, Scheinin TM. The desmoid (Reitamo) syndrome: etiology, manifestations, pathogenesis, and treatment. Curr Probl Surg. 1988;25:233-320.
5. Ayala AG, Ro JY, Goepfert H, et al. Desmoid fibromatosis: a clinicopathologic study of 25 children. Semin Diagn Pathol. 1986;3:138-150.
6. Rustgi AK. The genetics of hereditary colon cancer. Genes Dev. 2007;21:
2525-2538.
7. Lazar AJ, Tuvin D, Hajibashi S, et al. Specific mutations in the
beta-catenin gene (CTNNB1) correlate with local recurrence in sporadic
desmoid tumors. Am J Pathol. 2008;173:1518-1527.
596
8. Clark SK, Smith TG, Katz DE, et al. Identification and progression of a
desmoid precursor lesion in patients with familial adenomatous polyposis.
Br J Surg. 1998;85:970-973.
9. Deyrup AT, Tretiakova M, Montag AG. Estrogen receptor-beta expression in extraabdominal fibromatoses: an analysis of 40 cases. Cancer. 2006;
106:208-213.
10. Gronchi A, Casali PG, Mariani L, et al. Quality of surgery and outcome
in extra-abdominal aggressive fibromatosis: a series of patients surgically
treated at a single institution. J Clin Oncol. 2003;21:1390-1397.
11. Pritchard DJ, Nascimento AG, Petersen IA. Local control of extraabdominal desmoid tumors. J Bone Joint Surg Am. 1996;78:848-854.
12. Merchant TE, Nguyen D, Walter AW, et al. Long-term results with
radiation therapy for pediatric desmoid tumors. Int J Radiation Oncology Biol
Phys. 2000;47:1267-1271.
13. Jabbari S, Andolino D, Weinberg V, et al. Successful treatment of high
risk and recurrent pediatric desmoids using radiation as a component of
multimodality therapy. Int J Radiat Oncol Biol Phys. 2009;75:177-182.
CHEMOTHERAPY FOR DESMOID TUMOR

14. Constantinidou A, Jones RL, Scurr M, et al. Pegylated liposomal
doxorubicin, an effective, well-tolerated treatment for refractory aggressive
fibromatosis. Eur J Cancer. 2009;45:2930-2934.
15. Patel SR, Evans HL, Benjamin RS. Combination chemotherapy in
adult desmoid tumors. Cancer. 1993;72:3244-3247.
16. Raney B, Evans A, Granowetter L, et al. Nonsurgical management of
children with recurrent or unresectable fibromatosis. Pediatrics. 1987;79:394398.
17. Meazza C, Casanova M, Trecate G, et al. Objective response to hydroxyurea in a patient with heavily pre-treated aggressive fibromatosis.
Pediatr Blood Cancer. 2010;55:587-588.
18. Weiss AJ, Lackman RD. Low-dose chemotherapy of desmoid tumors.
Cancer. 1989;64:1192-1194.
19. Gega M, Yanagi H, Yoshikawa R, et al. Successful chemotherapeutic
modality of doxorubicin plus dacarbazine for the treatment of desmoid tumors
in association with familial adenomatous polyposis. J Clin Oncol. 2006;24:
102-105.
20. Lewis JJ, Boland PJ, Leung DH, et al. The enigma of desmoid tumors.
Ann Surg. 1999;229:866-873.
21. Sportiello DJ, Hoogerland DL. A recurrent pelvic desmoid tumor
successfully treated with tamoxifen. Cancer. 1991;67:1443-1446.
22. Belliveau P, Graham AM. Mesenteric desmoid tumor in Gardners
syndrome treated by sulindac. Dis Colon Rectum. 1984;27:53-54.
23. Hansmann A, Adolph C, Vogel T, et al. High-dose tamoxifen and

sulindac as first-line treatment for desmoid tumors. Cancer. 2003;100:612620.
24. He TC, Chan TA, Vogelstein B, et al. PPARdelta is an APC-regulated
target of nonsteroidal anti-inflammatory drugs. Cell. 1999;99:335-345.
25. Heinrich MC, McArthur GA, Demetri GD, et al. Clinical and molecular
studies of the effect of imatinib on advanced aggressive fibromatosis (desmoid
tumor). J Clin Oncol. 2006;24:1195-1203.
26. Fiore M, Rimareix F, Mariani L, et al. Desmoid-type fibromatosis: a
front-line conservative approach to select patients for surgical treatment. Ann
Surg Oncol. 2009;16:2587-2593.
27. Skapek SX, Hawk BJ, Hoffer FA, et al. Combination chemotherapy
using vinblastine and methotrexate for the treatment of progressive desmoid
tumor in children. J Clin Oncol. 1998;16:3021-3027.
28. Skapek SX, Ferguson WS, Granowetter L, et al. Vinblastine and
methotrexate for desmoid fibromatosis in children: results of a Pediatric
Oncology Group phase II trial. J Clin Oncol. 2007;25:501-506.
29. Skapek SX, Anderson JR, Hill DA, et al. The safety and efficacy of
high-dose tamoxifen and sulindac for desmoid-type aggressive fibromatosis in
children: results of a Childrens Oncology Group Phase II study. Paper
presented at: Annual Meeting of the Connective Tissue Oncology Society;
Chicago, IL; October 27, 2011.
597
MOLECULAR PATHWAYS FOR THE PRACTICING

PEDIATRIC ONCOLOGIST
CHAIR
Donald W. Parsons, MD, PhD
Texas Childrens Hospital
Houston, TX
SPEAKERS
Katia Scotlandi, PhD
Istituto Ortopedico Rizzoli
Bologna, Italy
Tobey MacDonald, MD
Emory University
Atlanta, GA
Targeting the Insulin-Like Growth Factor

(IGF) System Is Not as Simple as Just
Targeting the Type 1 IGF Receptor
By Katia Scotlandi, PhD, and Antonino Belfiore, MD
Overview: Increased signaling of the insulin-like growth factor (IGF) system via alterations in expression levels of its
components has been demonstrated in various tumor types.
Numerous experimental studies have supported the involvement of the IGF system signaling axis in tumor initiation and
progression. These studies, combined with data that link
alterations in the levels of circulating IGFs with cancer risk
and prognosis, have focused on the IGF-1 receptor (IGF-1R) as
a therapeutic target for patients with cancer. As a consequence, most therapeutic strategies have been designed to
specifically inhibit IGF-1R but have for the most part ignored
the insulin receptor (IR), based on concerns that targeting IR
would lead to unacceptable toxicity both because of its role in
HE IGF system comprises a phylo-genetically ancient

family of peptides involved in growth, development,
and metabolism, as well as in cellular processes such as
proliferation, survival, cell migration, and differentiation.
The IGF system is composed of three ligands (IGF-1, IGF-2,
and insulin), their receptors (the IGF-1 receptor [IGF-1R],
the mannose 6-phosphate/IGF-2 receptor [M6P/IGF-2R], the
insulin receptor [IR], and the hybrid IR/IGF-1R), at least six
high-affinity binding proteins and binding protein proteases.
IGF binding proteins modulate the activity of IGFs but also
have a life of their own inducing cellular processes in an
IGF-independent way. Molecular details of the IGF system
have been excellently reviewed by Samani and colleagues.1
In this context, it is important to highlight the complexity
of the system and the presence of several critical nodes that
within the signaling networks control various cellular processes. A simple scheme of divergent pathways is usually
sufficient to describe and explain IGF/insulin signaling (Fig.
1). However, when examined in detail, the number of genes
and protein isoforms involved in the activation of mitogenactivated protein kinase (MAPK) or AKT signaling pathways, the two main signaling mediators of the IGF system or
in genes involved in the generation of proliferative, antiapoptotic, differentiating, or metabolic effects, it becomes
clear that hundreds of molecules are involved in the IGF/
insulin-signaling pathway. It is beyond the scope of this
manuscript to describe in detail this molecular level of
complexity and interactions2 but, to make the reader more
aware of the peculiarities of this signaling axis, some examples of at least the best-defined critical nodes are described.
For instance, the IR has two splice isoforms, IR-A that is
highly expressed in fetal tissues and cancer, and IR-B that is
mainly found in adult tissue (details to follow). Both are
usually coexpressed in cells that also express IGF-1R. Insulin and IGFs bind with high affinity to their cognate receptors (e.g., insulin 3 IR; IGFs 3 IGF-1R), whereas IGF-2R
serves mainly as a sink for the regulation of IGF2 levels.
However, IGF-2 also binds IR-A with high affinity,3 although at lower affinity, IGF-1R can also be activated by
insulin, and IR can be activated by IGFs. This implies that
whenever we study the effects of IGF-1R, IR, or both, we
physiologic metabolism and because we frequently try to

oversimplify biologic complexity whenever we are urged to
find practical, friendly solutions for clinical practice. Although
this is an understandable and necessary starting point in the
complex and long-lasting processes that leads to translational
biology, the crude reality of the results obtained from phase I
and II studies suggest a need for researchers to be humble
and go back to the drawing board. Cancer research has
substantially neglected the role of IR, and it remains unclear
whether and to what extent avoiding the inhibition of IR has
compromised the efficacy of antiIGF-1R therapy. Clarifying
its role might also help us take advantage of older drugs that
could offer new perspectives in cancer care.
should also pay attention to the most prevalent types and

expression levels of the ligand(s) in that specific cellular
context.
Similar to IR, IGF-1R consists of two extracellular ligandbinding subunits (the alpha subunits) and of two transmembrane beta subunits, which are linked to alpha subunits by
disulfide bonds and are composed of a transmembrane
domain, an intracellular tyrosine kinase (TK) domain, and a
C-terminal tail. IGF-1R has 70% homology to IR, with which
it shares some signaling pathways. As a consequence of the
close homology of IR and IGF-1R, hybrid receptors can be
formed by an insulin alpha-beta hemireceptor and an IGF-1
alpha-beta hemireceptor in cells expressing both. The biologic response elicited by these hybrid receptors can vary,
depending on the ligands involved and the specific IR isoforms.4 These hybrid receptors appear to bind IGF-1 and
IGF-2 with high affinity similar to IGF-1R. Ligand binding
induces tyrosines within the TK domain to be transphosphorylated by the dimeric subunit partner. Phosphorylated
residues serve as docking sites for other signaling molecules,
such as IR substrates (IRS) and the adaptor protein Shc,
which leads to the activation of the phosphatidylinositol-3
kinase (PI3K)-Akt and MAPK pathways (Fig. 1). However,
both IR and IGF-1R can phosphorylate at least six known
substrate proteins (IRS1-6) that are capable of interacting
with eight known forms of the PI3K regulatory subunit,
which leads to the activation of three known isoforms of AKT
signaling, besides being able to crosstalk with the MAPK
signaling pathway. Moreover, these signaling pathways are
shared with most other TK receptors and it is possible that
other pathways that have yet to be identified are involved or
From the CRS Development of Biomolecular Therapies, Experimental Oncology Lab,

Orthopaedic Rizzoli Institute, Bologna, Italy; Department of Endocrinology, Department of
Health University Magna Graecia of Catanzaro, Catanzaro, Italy.
Address reprint requests to Katia Scotlandi, PhD, Orthopaedic Rizzoli Institute, CRS
Development of Biomolecular Therapies, Experimental Oncology Lab, Via di Barbiano
1/10, 40126 Bologna, Italy; email: katia.scotlandi@ior.it.
1092-9118/10/1-10
599
SCOTLANDI AND BELFIORE
that subtle differences in the recruitment of certain docking

proteins and intracellular mediators exist about which we
still know little regarding their dynamics and biologic effects. Finally, in addition to the well-established signaling
pathway from the cell membrane, recent reports have highlighted how IGF-1R or IR signaling, or both, may also be
dependent on cell localization. Differential endocytosis and
signaling dynamics have been reported for IGF-1R5 as well
as for IR-A and IR-B in relation to the mitogenic or metabolic
activities of the two receptors.6 Nuclear translocation of
IGF-1R has also been reported.7 The modification of IGF-1R
by small ubiquitin-like modifier (SUMO) proteins occurs in a
ligand-dependent manner and is necessary for nuclear
translocation of the receptor. As expected, nuclear IGF-1R
has different biologic functions: it binds to genomic DNA and
may act as a transcriptional enhancer.
This level of complexity has been grossly ignored when
targeted therapies against IGF-1R were developed. However, despite the difficulties of deeply understanding such
complex signaling, steps have been recently taken and
indeed some achievements have already been obtained.
Why Was IGF-1R Chosen as a Therapeutic Target?
The first inkling that IGF-1R played a crucial role in

malignant transformation was provided by Sell and colleagues (1993), who found that R cells could not be transformed by the SV40 large T antigen. R cells are 3T3 cells
originating from mouse embryos with a targeted disruption
of the IGF-1R gene. These cells have a tendency to transform
spontaneously in culture, and the SV40 T antigen is, by
itself, a strong transforming agent in 3T3c cells. The failure
of R cells to become transformed by SV40 T antigen indicated a role of the IGF-1R in transformation of cells in
culture. This finding has since been confirmed with different
viral and cellular oncogenes and in different laboratories.8
Reintroduction of an IGF-1R into R cells promptly renders
KEY POINTS
600
The insulin-like growth factor (IGF) system is an

important mediator of cancer pathogenesis and progression. Drug resistance to conventional or targeted
therapies frequently involves components of the
insulin-like growth factor system.
Researchers and pharmaceutical companies have
focused on IGF-1 receptor (IGF-1R), which is clearly
able to deliver a proliferative, antiapoptotic, and
promigratory signal in cancer cells.
Several approaches inhibiting IGF-1R functions have
shown very encouraging results in preclinical conditions, but only limited evidence of efficacy has been
demonstrated in phase I and II clinical studies.
The IGF system is quite complex, with many players
in the field. Insulin receptor function in cancer cells
has been underestimated, but also little attention has
been paid to the type of ligands that are mainly
involved in each tumor type.
Strategies considering the IGF system in all its complexity are encouraged.
these cells susceptible to transformation. Thus, there should

be a signal originating from the IGF-1R that facilitates and
is quasi-necessary for the transformation by the usual
agents (i.e., physical, chemical, and/or genetic). The level of
expression of IGF-1R does not need to be high. Even low
levels of expression are sufficient to send the permissive
signal that allows oncogenes to transform mammalian cells.
Accordingly, IGF-1R is overexpressed in some malignant
tissues but amplification and overexpression are less common for IGF-1R than for other oncogenetic receptors. Similarly, mutations have not been described as a way to
increase receptor activity. Activation of IGF-1R is indeed
mainly induced by either circulating or locally synthesized
IGFs in an autocrine or paracrine manner.9 Although some
studies report no relationship between IGF-1 levels and
cancer risk, many others report that individuals with IGF-1
levels at the upper end of the normal range have an
increased risk of developing certain cancers (e.g., colon,
breast, and prostate).10 Conversely, individuals with growth
hormone receptor deficiency, also known as Laron syndrome, who have very low IGF-1 levels, appear to be protected from the development of cancer when compared with
their relatives without hormonal deficiency.11,12 Dietary
factors and lifestyle have also been shown to have a substantial effect on the activation of the IGF system. In animal
models, caloric restriction reduced circulating IGF-1 levels
as well as bladder tumor growth, by increasing apoptosis
and decreasing cell proliferation.13
In any case, if IGF-1R is quasi-obligatory for cell transformation, downregulation of IGF-1R in malignant cells ought
to reverse the transformed phenotype. Downregulation or
inhibition of IGF-1R functions, by neutralizing antibodies or
small-molecule TK inhibitors (TKIs), by antisense strategies, or by developing agents to modulate IGF binding
proteins, causes massive apoptosis of tumor cells in vitro
and in vivo and results in the inhibition of tumorigenesis
and metastasis formation. Preclinical data suggest that
agents used to target the IGF system may be more effective
when used in combination with chemotherapy compared
with when used as monotherapy. In addition, the induction
of IGF-1R signaling has been described to be involved in
mediating resistance to both conventional and some targeted drugs.14-16 It is therefore not surprising that targeting
IGF-1R has become popular with pharmaceutical and biotechnology companies. Currently, most therapeutic agents,
monoclonal antibodies (MAbs), or TKIs have been designed to specifically target IGF-1R while sparing IR, on the
basis of the concern that cotargeting IR would lead to
unacceptable toxicity. MAbs targeting IGF-1R were the
furthest in development and had the benefit of inhibiting
hybrid receptors besides IGF-1R. Recently, several phase I
to III clinical trials have been conducted to evaluate the
safety and efficacy of drugs targeting the IGF-1R. From
these studies, we obtained some important indications:
(1) antiIGF-1R drugs have modest toxic effects, with mild
and reversible hyperglycemia as the most common toxicity;
and (2) anti-IGF-1R drugs show limited effectiveness. In
particular, the best tumor responses have been observed
in Ewings sarcoma, in which IGF/IGF-1R functions have
been clearly associated with the pathogenesis of this tumor
and in which few, if any, other TK receptors are fundamentally activated.9 However, despite the presence of the target
TARGETING THE IGF SYSTEM
Fig. 1. The IGF system.5

Abbreviations: ERK, extracellular regulated kinase; IGF, insulin-like growth factor; IRS, insulin receptor substrate; PI3-K, phosphoinositide3-kinase; mTOR, mammalian target of rapamycin.
in all tumors and ample preclinical evidence supporting

the potential value of antiIGF-1R agents, less than 10% of
patients respond to this therapy with extraordinary results.17
On one side, this implies that the presence of the target
is not sufficient to benefit from this targeted therapy and
that other redundant pathways may be present to render
IGF-1Rtargeted cells resistant to antiIGF-1R MAb.
Recent studies in cell lines have demonstrated that knocking out, downregulating, or pharmacologically inhibiting
IGF-1R can lead to a compensatory increase in IR
signaling.18-21 So, the take-home messages of these studies
are: (1) the ratio IGF-1R/IR as well as the type of ligand(s)
that are prevalent in the specific cellular context should be
considered to identify patients that may benefit from anti
IGF-1R therapy; (2) we need to better identify the mechanisms of action of IR in cancer, viewing this receptor in a
new light.
Insulin/IR in Cancer
As mentioned above, IR shares high homology to IGF-1R.

However, unlike IGF-1R, the IR is characterized by the
ability to alternatively splice a small exon (exon 11) encoding a 12amino acid stretch contiguous to the CT
peptide (encoded by the C-terminal sequence). The exclusion of exon 11 generates isoform A (IR-A); its inclusion
generates isoform B (IR-B).4 Although the current thinking is that IR primarily mediates the metabolic effects of
insulin through the activation of the PI3K pathway and
IGF-1R mainly mediates the growth effects of IGFs via the
activation of MAPK, it is now clearly established that in
cancer cells IR, particularly IR-A, is often overexpressed
and its signaling pathway deregulated with substantial
crosstalk with the IGF-1R pathway. Several factors account
for the loss of IR physiologic specificity in cancer. First,
cancer cells predominantly overexpress the IR-A isoform.
601
Although IR-B is a specific receptor for insulin, IR-A also

binds IGF-2 and at lower affinity IGF-1, and may induce
biologic effects in response to both IGFs. Second, overexpressed IR enhances the effects of IGF-1 and IGF-2
through the formation of IR/IGF-1R hybrid receptors,
which bind both IGFs with high affinity. Third, cancers
often produce both IGF-2 and IGF-1 in an autocrine/paracrine manner.1 Finally, in patients with cancer affected by
insulin resistance, the elevated levels of circulating insulin
induce unbalanced IR activation, with predominant activation in the mitogenic pathway rather than the metabolic
pathway.22,23
The importance of IR and insulin in tumor development
and progression has been demonstrated in both animal
models and clinical studies.10 In humans, high levels of
insulin but not blood glucose or obesity per seare associated with increased risk for various malignancies.24
Women with breast cancer that also have insulin resistance
show increased cancer-specific mortality.25,26
Obesity is a very important determinant for inducing or
worsening insulin resistance. Obesity is also a predisposing
factor of type 2 diabetes (T2DM), and several studies have
now firmly established that both obesity and/or T2DM are
associated with an increased risk of cancer.23,27,28 Patients
with T2DM carry an increased risk for almost every cancer
histotype, except prostate cancer. Obese patients are at
increased risk for a variety of malignancies, including most
common cancers and hematologic malignancies. Metabolic
syndrome, a disorder characterized by obesity, hypertension, dyslipidemia, and long-term insulin resistance, is also
associated with worse cancer prognosis.29 Conversely, body
weight reduction decreases cancer risk.22 Because of these
findings and considering the studies suggesting that insulin
analogs may promote tumorigenesis,10 the effects of insulin/IR on tumor growth have recently received greater attention.
New Opportunities for Cancer Prevention and
Therapy Involving the IR pathway
As long-term exposure to hyperinsulinemia is an important risk factor for cancer development and progression in
patients with obesity, T2DM, or both, measures and drugs
aimed at improving insulin resistance and reducing circulating insulin levels should contribute to prevent cancer in
these patients and to ameliorate prognosis in patients with
cancer.25,26 Nonpharmacologic measures, such as lifestyle
changes involving caloric restriction and physical exercise,
may also be useful.
Among drugs aimed at reducing insulin resistance and
circulating insulin levels, biguanides and thiazolidinediones
(TZDs) (collectively classified as insulin sensitizers) have
received attention as potential anticancer agents. Metformin is the only biguanide used in the clinical setting and
is currently recommended as first-line therapy in patients
with T2DM for its excellent long-term safety profile. Metformin impairs the production of adenosine 5-triphosphate
(ATP) by targeting complex I in the mitochondrial electron
transport chain. This event activates AMP-activated protein
kinase (AMPK), a kinase with a key role in the regulation of
cellular energy homeostasis and growth. AMPK causes, on
one hand, downregulation of gluconeogenesis in the liver
with reductions in blood glucose and insulin levels and, on
602
the other hand, direct reduction of cell growth through the

inhibition of the mTOR/AKT pathway. Indeed, in vitro
studies and animal models strongly suggest that metformin
may have anticancer effects.30,31 In humans, observational
clinical studies have actually shown a decrease in cancer
risk in patients with T2DM using metformin compared with
those following other treatment regimens. In a case-control
study, metformin use was associated with a reduced risk for
breast cancer.32 Moreover, the adjunct of metformin to
insulin was reported to offset the increased risk for colorectal or pancreatic cancer observed when insulin was used as
monotherapy, and patients with T2DM treated with metformin were found to have a reduced cancer-specific mortality compared with those using insulin. From a therapeutic
point of view, metformin may improve response rates in
women with breast cancer that have T2DM who are receiving adjuvant chemotherapy,33 as well as progression-free
survival for chemotherapy-treated patients with advanced
cell lung cancer that have diabetes.34 Currently, pilot clinical trials are being conducted with women without diabetes
to evaluate the possible effect of metformin on the outcome
of breast cancer (clinical trials NCT00897884 and
NCT01101438).
TZDs, the second class of insulin sensitizers available for
clinical use, belong to the group of peroxisome proliferatoractivated receptor gamma (PPARgamma) agonists. These
drugs have shown substantial antitumoral effects in vitro
and in some, but not in all, animal models. However to date,
enthusiasm for the anticancer potential of the currently
available TZDs has declined because of toxicity and an
associated increased risk of tumors.35 When new TZDs reach
the market, more studies are warranted to explore the
effects of these drugs for patients with cancer who have
insulin resistance.
As mentioned above, IR-A overexpressed by cancer cells
may be stimulated not only by circulating insulin, but also
by autocrine- or paracrine-produced IGF-2 and IGF-1. In
patients with cancers overexpressing IR-A and IGFs, therefore, lowering insulin levels with the use of insulin sensitizers is not sufficient and direct inhibition of this IGFs/IR-A
pathway should be pursued.
These considerations, together with evidence indicating
that selective IGF-1R inhibitors can favor the emergence of
cell clones with enhanced IGFs/IR-A loops21,36 and worsen
hyperinsulinemia,37 bring about the concept that cotargeting IGF-1R and IR may be a suitable approach for patients
with these malignancies. Small molecules with TK inhibitory activity appear to be the most promising drugs because
of their ability to block the ATP-binding site of the kinase
domain, which shares a high degree of homology between IR
and IGF-1R. These drugs can be given orally and administered in combination with standard chemotherapy. Two
currently available TKIs (BMS-754807 and OSI-906) share
the ability to inhibit both IR and IGF-1R. BMS-754807
inhibits both IGF-1R and IR with very similar activity (the
half maximal inhibitory concentrations [IC50] were 1.8
nmol/L and 1.7 nmol/L, respectively),38 but also elicits substantial inhibition toward other TK receptors (e.g., Met,
recepteur dorigine nantais [RON], TrkA, TrkB) and Aurora
A and B. BMS-754807 is currently being evaluated in
several clinical trials as a single agent and in combination
with other drugs in patients with advanced or metastatic
TARGETING THE IGF SYSTEM
malignancies (clinical trials NCT00898716, NCT00569036,

NCT00908024, NCT00788333, NCT01225172, and NCT00793897). OSI-906 is a selective dual-inhibitor of IR and
IGF-1R (IC50 of 19 nmol/L to 35 nmol/L).39 Preliminary
studies with OSI-906 have yielded encouraging results and
this drug is now being evaluated in phase I escalation
studies as a single agent (clinical trials NCT00514306 and
NCT00514007) and in phase I to III trials (clinical trials
NCT01101906, NCT00924989, NCT01387386, and others)
for various malignancies generally characterized by an activated IGF-2/1R-A loop.
These studies will hopefully provide proof-of-concept that
IR inhibition, in addition to IGF-1R inhibition, may be
clinically relevant for patients with cancer.
Other Controversial Issues
In addition, IGF-1R was shown to mediate differentiation

in some cancers. Specific experimental models indicated
that the balance between mitogenesis and differentiation
is strongly influenced by the relative level of expression
of the two main IGF-1R mediators, Shc and IRS1.40
Unfortunately, this is not a general rule and the exact
mechanism that shifts the message from IGF-1R is still
unknown. In any case, evidence indicates that the IGF
system is an important mediator of mesenchymal or neural
differentiation, an aspect that we need to consider for
sarcomas and brain tumors. We need to be aware that
IGF-1R and its substrates can also send contradictory signals, signals that can actually lead to growth inhibition or to
the inhibition of metastatic spread. These contradictions
ought to become fertile areas of investigation for both basic
and applied research.
There is no doubt that antiIGF-1R therapy should be
combined with conventional or other targeted drugs. Each
tumor requires a unique cocktail of drugs and dedicated
studies. This concept is true for most targeted therapies
and further effort, time, and resources to be translated
into effective treatments are needed. The results achieved
to date are not satisfactory to justify routine clinical use
of IGF-1Rtargeting agents. Nevertheless, for some heavily
pretreated patients with refractory rare tumors, responses
and clinical benefit in combination with chemotherapy
have been observed. Unfortunately, rare tumors seem
to be most sensitive to these targeted therapies, which
does not inspire pharmaceutical companies. However, we
strongly believe that joint efforts between academia and

industry are in the interests of both. We have a good level
of knowledge in the field and several drugs already developed. Just put them together and take another step toward
light.
Conclusion
Recent phase I to II clinical studies with selective anti

IGF-1R MAbs together with epidemiologic data have shifted
attention from IGF-1R to IR, and to a more comprehensive
view of the IGF system. To date, small molecules acting as
dual inhibitors of IGF-1R and IR appear to be the most
promising approaches to deprive cancer cells of this important signaling axis. Unfortunately, hyperglycemia and hyperinsulinemia are important adverse effects of dual IGF-1R
and IR inhibitors, although hyperglycemia seems to be
reversible after the cessation of treatment. It can be hypothesized, therefore, that insulin sensitizers (e.g., metformin)
should be given together with these inhibitors to limit these
adverse effects.
Moreover, we need new biomarkers to select patients
suitable for IR and IGF-1R dual inhibition and to monitor
therapeutic efficacy. Recently, it has been reported that
response to a dual antiIR/IGF-1R inhibitor may be correlated with an IGF expression signature41 or with lack of
epithelial-mesenchymal transition,42 but we clearly need
more extensive studies and validated biomarkers.
Another possible approach involves the use of antibodies
recognizing both IGF-2 and IGF-1. Such antibodies have
been described43 and, in animal models, show promising
results in IGFs-driven malignancies.44 Also in this case,
more studies are needed regarding the applicability of this
approach in humans.
Overall, recent experimental evidence has shed light on to
some new players, like IR, which has been substantially
neglected in the field of cancer research as a mediator of
tumor progression. We are now well aware of the complexity
of the pathway and have some new potentially promising
drugs in our hands. Further efforts are needed to learn how
to maximize their efficacy in patients with cancer.
Acknowledgment
The studies cited in this review were supported by grants
from the Italian Association for Cancer Research (IG-10452
[Katia Scotlandi] and IG-10625 [Antonino Belfiore]).
Author
Katia Scotlandi*
Antonino Belfiore*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Samani AA, Yakar S, LeRoith D, et al. The role of the IGF system in
cancer growth and metastasis: overview and recent insights. Endocr Rev.
2007;28:20-47.
2. Taniguchi CM, Emanuelli B, Kahn CR. Critical nodes in signalling
pathways: insights into insulin action. Nat Rev Mol Cell Biol. 2006;7:85-96.
3. Frasca F, Pandini G, Scalia P, et al. Insulin receptor isoform A, a newly
recognized, high-affinity insulin-like growth factor II receptor in fetal and

cancer cells. Mol Cell Biol. 1999;19:3278-3288.
4. Belfiore A, Frasca F, Pandini G, et al. Insulin receptor isoforms and
insulin receptor/insulin-like growth factor receptor hybrids in physiology and
disease. Endocr Rev. 2009;30:586-623.
5. Martins AS, Ordonez JL, Amaral AT, et al. IGF1R signaling in Ewing
603

sarcoma is shaped by clathrin-/caveolin-dependent endocytosis. PLoS One.
2011;6:e19846.
6. Giudice J, Leskow FC, Arndt-Jovin DJ, et al. Differential endocytosis
and signaling dynamics of insulin receptor variants IR-A and IR-B. J Cell Sci.
2011;124:801-811.
7. Sehat B, Tofigh A, Lin Y, et al. SUMOylation mediates the nuclear
translocation and signaling of the IGF-1 receptor. Sci Signal. 2010;3:ra10
8. Baserga R. The insulin-like growth factor-I receptor as a target for
cancer therapy. Expert Opin Ther Targets. 2005;9:753-768.
9. Scotlandi K, Picci P. Targeting insulin-like growth factor 1 receptor in
sarcomas. Curr Opin Oncol. 2008;20:419-427.
10. Gallagher EJ, LeRoith D. Minireview: IGF, insulin, and cancer. Endocrinology. 2011;152:2546-2451.
11. Steuerman R, Shevah O, Laron Z. Congenital IGF1 deficiency tends to
confer protection against post-natal development of malignancies. Eur J
Endocrinol. 2011;164:485-489.
12. Guevara-Aguirre J, Balasubramanian P, Guevara-Aguirre M, et al.
Growth hormone receptor deficiency is associated with a major reduction in
pro-aging signaling, cancer, and diabetes in humans. Sci Transl Med. 2011;
3:70ra13.
13. Dunn SE, Kari FW, French J, et al. Dietary restriction reduces
insulin-like growth factor I levels, which modulates apoptosis, cell proliferation, and tumor progression in p53-deficient mice. Cancer Res. 1997;57:46674672.
14. Benini S, Manara MC, Baldini N, et al. Inhibition of insulin-like growth
factor I receptor increases the antitumor activity of doxorubicin and vincristine against Ewings sarcoma cells. Clin Cancer Res. 2001;7:1790-1797.
15. Bodzin AS, Wei Z, Hurtt R, et al. Gefitinib resistance in HCC Mahlavu
cells: upregulation of CD133 expression, activation of IGF-1R signaling
pathway, and enhancement of IGF-1R nuclear translocation. J Cell Physiol.
Epub 2011 Sep 29.
16. Rosenzweig SA. Acquired resistance to drugs targeting receptor tyrosine kinases. Biochem Pharmacol. Epub 2011 Dec 26.
17. Olmos D, Postel-Vinay S, Molife LR, et al. Safety, pharmacokinetics,
and preliminary activity of the anti-IGF-1R antibody figitumumab (CP751,871) in patients with sarcoma and Ewings sarcoma: a phase 1 expansion
cohort study. Lancet Oncol. 2010;11:129-135.
18. Buck E, Gokhale PC, Koujak S, et al. Compensatory insulin receptor
(IR) activation on inhibition of insulin-like growth factor-1 receptor (IGF-1R):
rationale for cotargeting IGF-1R and IR in cancer. Mol Cancer Ther. 2010;9:
2652-2664.
19. Ulanet DB, Ludwig DL, Kahn CR, et al. Insulin receptor functionally
enhances multistage tumor progression and conveys intrinsic resistance to
IGF-1R targeted therapy. Proc Natl Acad Sci U S A. 2010;107:10791-10798.
20. Dinchuk JE, Cao C, Huang F, et al. Insulin receptor (IR) pathway
hyperactivity in IGF-IR null cells and suppression of downstream growth
signaling using the dual IGF-IR/IR inhibitor, BMS-754807. Endocrinology.
2010;151:4123-4132.
21. Garofalo C, Manara MC, Nicoletti G, et al. Efficacy of and resistance to
anti-IGF-1R therapies in Ewings sarcoma is dependent on insulin receptor
signaling. Oncogene. 2011;30:2730-2740.
22. Kaaks R, Lukanova A. Effects of weight control and physical activity in
cancer prevention: role of endogenous hormone metabolism. Ann N Y Acad
Sci. 2002;963:268-281.
23. Vigneri P, Frasca F, Sciacca L, et al. Diabetes and cancer. Endocr Relat
Cancer. 2009;16:1103-1123.
604
24. Pisani P. Hyper-insulinaemia and cancer, meta-analyses of epidemiological studies. Arch Physiol Biochem. 2008;114:63-70.
25. Duggan C, Irwin ML, Xiao L, et al. Associations of insulin resistance
and adiponectin with mortality in women with breast cancer. J Clin Oncol.
2011;29:32-39.
26. Irwin ML, Duggan C, Wang CY, et al. Fasting C-peptide levels and
death resulting from all causes and breast cancer: The health, eating, activity,
and lifestyle study. J Clin Oncol. 2011;29:47-53.
27. Calle EE, Kaaks R. Overweight, obesity and cancer: epidemiological
evidence and proposed mechanisms. Nat Rev Cancer. 2004;4:579-591.
28. Rousseau MC, Parent ME, Pollak MN, et al. Diabetes mellitus and
cancer risk in a population-based case-control study among men from Montreal, Canada. Int J Cancer. 2006;118:2105-2109.
29. Stocks T, Borena W, Strohmaier S, et al. Cohort Profile: The Metabolic
syndrome and Cancer project (Me-Can). Int J Epidemiol. 2010;39:660-667.
30. Jalving M, Gietema JA, Lefrandt JD, et al. Metformin: taking away the
candy for cancer? Eur J Cancer. 2010;46:2369-2380.
31. Bost F, Sahra IB, Le Marchand-Brustel Y, et al. Metformin and cancer
therapy. Curr Opin Oncol. 2012;24:103-108.
32. Bosco JL, Antonsen S, Sorensen HT, et al. Metformin and incident
breast cancer among diabetic women: a population-based case-control study
in Denmark. Cancer Epidemiol Biomarkers Prev. 2011;20:101-111.
33. Jiralerspong S, Palla SL, Giordano SH, et al. Metformin and pathologic
complete responses to neoadjuvant chemotherapy in diabetic patients with
34. Tan BX, Yao WX, Ge J, et al. Prognostic influence of metformin as
first-line chemotherapy for advanced nonsmall cell lung cancer in patients
with type 2 diabetes. Cancer. 2011;117:5103-5111.
35. Belfiore A, Genua M, Malaguarnera R. PPAR-gamma agonists and
their effects on IGF-I receptor signaling: implications for cancer. PPAR Res.
2009. Epub Jul 7.
36. Zhang H, Pelzer AM, Kiang DT, et al. Down-regulation of type I
insulin-like growth factor receptor increases sensitivity of breast cancer cells
to insulin. Cancer Res. 2007;67:391-397.
37. Pollak M. Targeting insulin and insulin-like growth factor signalling in
oncology. Curr Opin Pharmacol. 2008;8:384-392.
38. Carboni JM, Wittman M, Yang Z, et al. BMS-754807, a small molecule
inhibitor of insulin-like growth factor-1R/IR. Mol Cancer Ther. 2009;8:33413349.
39. Mulvihill MJ, Cooke A, Rosenfeld-Franklin M, et al. Discovery of
OSI-906: a selective and orally efficacious dual inhibitor of the IGF-1 receptor
and insulin receptor. Future Med Chem. 2009;1:1153-1171.
40. Valentinis B, Baserga R. IGF-I receptor signalling in transformation
and differentiation. Mol Pathol. 2001;54:133-137.
41. Litzenburger BC, Creighton CJ, Tsimelzon A, et al. High IGF-IR
activity in triple-negative breast cancer cell lines and tumorgrafts correlates
with sensitivity to anti-IGF-IR therapy. Clin Cancer Res. 2011;17:2314-2327.
42. Gualberto A, Dolled-Filhart M, Gustavson M, et al. Molecular analysis
of non-small cell lung cancer identifies subsets with different sensitivity to
insulin-like growth factor I receptor inhibition. Clin Cancer Res. 2010;16:
4654-4665.
43. Feng Y, Zhu Z, Xiao X, et al. Novel human monoclonal antibodies to
insulin-like growth factor (IGF)-II that potently inhibit the IGF receptor type
I signal transduction function. Mol Cancer Ther. 2006;5:114-120.
44. Gao J, Chesebrough JW, Cartlidge SA, et al. Dual IGF-I/II-neutralizing
antibody MEDI-573 potently inhibits IGF signaling and tumor growth.
Cancer Res. 2011;71:1029-1040.
Hedgehog Pathway in Pediatric Cancers:

Theyre Not Just for Brain Tumors Anymore
By Tobey J. MacDonald, MD
Overview: The Hedgehog (HH) pathway regulates fundamental processes in embryonic development, including stem cell
maintenance, cell differentiation, tissue polarity, and cell
proliferation. In the vertebrate pathway, Sonic hedgehog
(SHH) binds to Patched1 (PTCH1), which relieves its inhibition
of Smoothened (SMO), allowing the GLI family of transcription
factors to translocate to the nucleus and activate HH target
genes such as GLI1, GLI2, PTCH1, CYCLIN D1, BCL-2, and
MYCN. The HH pathway is also an active participant in
tumorigenesis. In 1996, loss-of-function mutation in PTCH1
was discovered to be the cause of nevoid basal cell carcinoma
syndrome (NBCCS, or Gorlin syndrome), an autosomal dominant disease associated with increased rates of basal cell
carcinoma (BCC), medulloblastoma (MB), and rarely, rhabdomyosarcoma. It is now estimated that 100% of sporadic BCC
and up to 20% to 30% of MB also harbor activating HH
HE HEDGEHOG pathway was first discovered in 1980

by Nobel Laureates Nusslein-Volhard and Wieschaus
after their isolation of mutations in genes that control the
development of anterior-posterior body axis segmentation in
Drosophila melanogaster.1 In mammals, three genes, Desert
Hedgehog (DHH), Indian Hedgehog (IHH), and Sonic
Hedgehog (SHH), function as ligands for the receptor
PTCH1.2 In the unbound state, PTCH1 inhibits the sevenpass transmembrane protein SMO, and on HH ligand binding, PTCH1 undergoes internalization and degradation,
resulting in the activation and release of SMO to enter the
primary cilia where it promotes the dissociation of the
Suppressor of fused (SUFU) glioma-associated oncogene
homolog (GLI) complex.2,3 This results in nuclear translocation and activation of the GLI1 and GLI2 transcription
factors, and degradation of the repressor forms of GLI
(primarily GLI3). GLI proteins stimulate the transcription
of HH pathway target genes, including GLI1, GLI2, PTCH1,
CYCLIN D1, BCL-2, and MYCN, which in turn, function
in a wide range of developmental signaling roles.4 In some
cases, HH ligands function as mitogens, whereas in others
they promote differentiation.2 SHH, the most common vertebrate homolog, is required for the correct patterning of the
neural tube, the somites, and anterior-posterior positioning
of the limb bud.5 The importance of HH signaling in mammalian development is underscored by observations that
mutations in SHH cause holoprosencephaly, a developmental disorder that affects the midline of the face and nervous
system.2-5 The HH pathway can be activated by mutations
in PTCH1 or SUFU (loss of function), or SMO (gain of
function) that lead to ligand-independent, constitutive signaling. A clear genetic contribution of such HH pathway
activation to oncogenesis was established with the discovery
of loss-of-function mutations in germ-line PTCH1 (chromosome 9q22.3) as the cause of familial NBCCS (or Gorlins
syndrome), an autosomal-dominant disease in which patients have an increased tendency to develop basal cell
carcinoma (BCC), medulloblastoma (MB), rhabdomyosarcoma, and ovarian neoplasia.6 This review will further detail
the known relationships of the HH signaling pathway in
cancer development, especially medulloblastoma and other
pathway mutations. Together, these discoveries firmly established the linkage between HH pathway activation and cancer
development. Intense research has since been focused on
further defining the role of the HH pathway in BCC and MB and
potential therapeutic strategies to inhibit HH signaling. Early
clinical trials of SMO inhibitors have shown promising results
in the treatment of adult BCC and SHH-driven MB. More
recently, a number of other pediatric cancers have been
reported to show HH activity, making these tumors potential
candidates for HH inhibitor therapy. To date however, no HH
pathway mutations have been identified in other pediatric
cancers. This review will describe the HH pathway signaling in
development and cancer with a focus on recent evidence for
HH pathway activation in central nervous system (CNS) and
non-CNS pediatric cancers.
pediatric CNS tumors, and will focus on the most recent

evidence implicating HH pathway signaling in non-CNS
pediatric cancers.
HH Pathway Mutations and HH Signaling in Cancer
Following the discovery that germ-line mutations in

PTCH1 are responsible for NBCCS, a number of studies
subsequently confirmed the detection of HH pathway somatic mutations in sporadic BCC and MB.7,8 It is estimated
that almost all BCC tumors show evidence of constitutive
HH pathway activity, with 90% exhibiting loss of PTCH1
and 10% harboring activating mutations in SMO.7 Medulloblastomas appear to be more heterogeneous, with up to 20%
to 30% of tumors displaying a gene expression signature
that is indicative of HH pathway activation. However, only
50% of these tumors have confirmed loss of PTCH1, loss
of SUFU, or gain-of-function SMO mutations.3,9-11 Thus, a
subset of medulloblastoma exhibits HH pathway activation
through an alternative mechanism.
Mice heterozygous for patched (ptc1) mutations, like
heterozygous PTCH1 in humans, have a high rate of MB, as
well as other tumors, and concomitant loss of Tp53 has
been shown to further accelerate MB formation.12 Subsequent studies have confirmed that Ptc1 deletion in lineagerestricted progenitor or stem cells is sufficient to initiate
MB, and that MYCN is an essential downstream effector of
SHH in both normal and neoplastic cerebellar growth.13,14
Not only have these murine studies confirmed the role of
activating HH pathway mutations in the development of a
subset of MB, and provide insight into the function of HH
signaling in tumorigenesis, but also, these transgenic mice
From the Pediatric Neuro-Oncology Program, Aflac Cancer Center and Blood Disorders
Service, Childrens Healthcare of Atlanta, and Emory University School of Medicine, Emory
Childrens Center, Atlanta, GA.
Address reprint requests to Tobey J. MacDonald, MD, Emory University School of
Medicine, Emory Childrens Center, 2015 Uppergate Drive NE, Suite 442, Atlanta, GA;
email: tobey.macdonald@emory.edu.
1092-9118/10/1-10
605
TOBEY J. MACDONALD
MB models have served as an invaluable preclinical tool for

the testing of HH pathway inhibitors for the treatment of
SHH-driven MB.
It has long been known that a constellation of midline
developmental anomalies that are observed in embryos with
primary congenital defects in SHH activation has similarly
been observed in developing embryos exposed to naturally
occurring teratogenic alkaloids such as cyclopamine, found
in extracts from Veratrum Album, commonly known as
White Hellebore, a plant within the lily family.3 It has been
demonstrated that these teratogens function as specific
inhibitors of HH signaling by binding to SMO. As all HH
signaling through the canonical pathway requires SMO,
small molecules such as cyclopamine, which inhibit SMO
function, completely block all HH pathway signaling regardless of the ligand.3 HH pathway antagonists also provide
valuable tools for dissecting the biochemistry and biology of
HH signaling and have enabled the current development of
unique molecularly targeted therapies for HH-driven cancer. For example, treatment with SMO inhibitors have been
shown to completely eradicate HH-driven murine MB as
well as markedly inhibit SHH in human BCC, and transiently induce full remission in adult human metastatic MB,
indicating great promise for the clinical use of these agents
for HH-driven disease.15-17
Although the HH pathway is activated by autocrine signaling by HH ligands, it can also initiate paracrine signaling
with cells in the microenvironment. This creates a network
of HH pathway signaling that determines the malignant
behavior of the tumor cells. As a result of paracrine signal
transmission, the effects of HH signaling most profoundly
influence the stromal cells that constitute the tumor microenvironment. The stromal cells in turn produce factors
that nurture the tumor. Thus, such cellular cross-talk can
greatly amplify HH signaling, resulting in the promotion of

tumor progression and metastasis. Ultimately, the linkage
of aberrant HH signaling to tumorigenesis is thought to
be mediated through cell-cycle dysregulation, protection of
cancer cells against apoptosis, and modulation of angiogenesis.2,3 Several lines of evidence support three mechanistic roles for HH signaling in cancer: (1) a cancer cellautonomous role, in which tumor growth is driven by
activating mutations in the pathway; (2) a paracrine signaling role involving tumor and stromal interactions that
promote tumor growth and invasion; and (3) an autocrine
signaling role via cancer stem cells that promotes selfrenewal and proliferation.2,3 The downstream effectors of
the HH signaling pathway are the GLI transcription factors,
which promote cell proliferation, differentiation, and survival through induction of relevant target genes.2,3 To date,
different molecular lesions in PTCH1, SMO, and SUFU of
the HH pathway have been described in tumors. In each
case, these alterations have resulted in increased transcriptional activity of the GLI1 and GLI2 transcription factors.
Indeed, the first indication that genes in the HH pathway
were associated with human cancer was the observation
that GLI1 was amplified in glioblastoma, although it is now
believed that this is not a common primary mechanism
underlying glioblastoma formation.3 Together with GLI and
PTCH1, these effector targets are representative of the gene
signature indicating SHH active tumors. Therefore, GLI1
mRNA levels either in tumor tissue or relevant surrogate
tissue is considered a reliable indicator of HH pathway
activity.2,3 Most recently, a number of common pediatric
cancers have been shown to have expression and activation
of the HH pathway through the measurement of these target
effectors.
HH Pathway in Pediatric Cancer
KEY POINTS
606
Hedgehog (HH) signaling has been investigated for

its role in tumorigenesis because of its known function in embryonic stem cell maintenance, cell differentiation, tissue polarity, and cell proliferation.
Key constituents of the HH pathway include Sonic
hedgehog (SHH) ligand and its receptor Patched1
(PTCH1), which in the absence of SHH represses
Smoothened (SMO) and prevents GLI transcription
factor activation.
PTCH1 germ-line mutation results in nevoid basal
cell carcinoma (BCC) syndrome, while somatic HH
pathway mutations are found in sporadic BCC and
medulloblastoma (MB), thereby establishing a linkage between HH activation and cancer development.
Encouraging results have been observed using SMO
inhibitors to treat adult BCC and MB; however, the
potential risks of SMO inhibitors in developing children remain a concern.
HH pathway activity, but not mutations, has recently
been shown in other pediatric cancers, yet it remains
to be seen whether SMO inhibition will be effective
for this group of cancers.
The HH signaling pathway is believed to be active in

early-onset pediatric tumors, both of CNS and non-CNS
origin, because of the important role that HH plays in
embryonic development. Indeed, GLI1 amplification has also
been described in childhood sarcoma. In a more comprehensive study of a series of pediatric surgical tumor specimens,
Oue and colleagues used expression of GLI1 as a marker of
HH pathway activation to demonstrate that almost 70% of
the pediatric tumors examined, including neuroblastoma,
hepatoblastoma, high-grade glioma, and osteosarcoma, were
positive for HH activity.18 However, to date no HH mutations have been identified in these other pediatric cancers.
Only descriptive studies have been performed, and thus it
remains to be determined whether a valid functional relationship exists between HH activity and tumor progression
in these cancers. The findings of these pediatric studies
and others similarly reporting HH pathway expression and
activation in both CNS and non-CNS pediatric cancers are
further detailed below.
CNS tumors
The role of HH pathway in pediatric MB is well established. However, more recent reports now suggest that the
HH pathway may be implicated in other non-MB pediatric
CNS tumors.
HEDGEHOG PATHWAY IN PEDIATRIC CANCER
Rush and colleagues20 demonstrated that mRNA expression levels of members of the HH pathway were elevated in
45% of juvenile pilocytic astrocytoma specimens analyzed
and that the expression of the HH pathway correlated
inversely with patient age. Immunohistochemical (IHC)
staining for PTCH1, GLI1, and the proliferation marker
Ki67 demonstrated that patients diagnosed before the age of
10 years had an increased frequency and level of marker
immunopositivity compared with those diagnosed after 10
years of age.20 A significant correlation was also observed
between Ki67, PTCH1, and GLI1 positive staining, with 86%
of Ki67-positive cells also expressing PTCH1.
subsets of ERMS, but only rarely in ARMS tumors.24 Importantly, neither PTCH1 mutations nor activating SMO mutations were detected in ERMS tumors with high GLI1
expression, and in contrast to other reports, HH pathway
activity in ERMS tumors did not correlate with a unique
clinical phenotype. Furthermore, the gene expression patterns in ERMS indicated that approximately 29% exhibited
evidence of HH pathway activity, yet this pattern was
always coassociated with either p53 or RB pathway signatures.25 Finally, Oue and colleagues examined 18 RMS by
IHC and showed that the majority of the tumors were
immunopositive for SHH (78%), PTCH1 (100%), and GLI1
(78%).18 In contrast to the study by Pressey and colleagues,
marker expression was higher in ARMS than in ERMS.
Caution should be taken in interpreting the clinical significance of these results, and attempts to clinically translate
these findings to therapeutic interventions would be premature as the presence of the HH pathway signature does
not necessarily mean that the tumor cells are dependent on
SMO activity. Cyclopamine studies in the Ptc1 mice are
an example of this, whereby loss of ptc1 may contribute
to tumor initiation, but is not required for tumor maintenance.26
Non-CNS Pediatric Cancer
Neuroblastoma
Rhabdomyosarcoma
Neuroblastoma (NB) is a heterogeneous pediatric malignancy, with variable differentiation and growth potential,
that shares a common origin arising from neural crest cells
in the sympathetic nervous system. Using IHC, Souzaki and
colleagues examined 82 NB and 10 ganglioneuroblastoma
(GNB) and demonstrated tumor immunopositivity for SHH,
GLI1, and PTCH1 in 67 (73%), 62 (67%), and 73 (79%),
respectively.27 Most NBs without MYCN amplification were
positive for all three HH pathway markers. Only two (10%)
of 20 cases with MYCN amplification were also positive for
SHH and GLI1, and four (20%) were positive for PTCH1.
The percentage of GLI1-positive cells without MYCN amplification was significantly higher than those with MYCN
amplification, and the prognosis of the GLI1-positive cases
was significantly better than that of the GLI1-negative
cases. In tumors without MYCN amplification, high expression of GLI1 was significantly associated with early clinical
stage, more differentiated tumors, and a good prognosis.
Oue and colleagues examined 25 NB by IHC and similarly
showed that 24 (96%), 17 (68%), and 25 (100%) stained
positive for SHH, PTCH, and GLI1, respectively.18 Likewise,
all of the Gli1-negative tumors were poorly differentiated
and exhibited advanced-stage disease. In this study, there
was no significant relationship between GLI1 expression
and MYCN amplification or prognosis. This study also
showed that GLI1 transduction of NB cells inhibited proliferation in vitro and induced a gene expression pattern that
resembled benign differentiated ganglioneuroma.28 Notably,
GLI1 transduction did not induce MYCN expression in NB
cells.
Diffuse Intrinsic Pontine Glioma
In a report by Monje and colleagues, early postmortem

tissue from patients with diffuse intrinsic pontine glioma
(DIPG) was used to establish in vivo xenograft models of
DIPG.19 Subsequent analysis of the established tumors
showed that the HH signaling pathway is active in DIPG
tumor cells and that DIPG self-renewal capacity in neurosphere culture is significantly reduced with inhibition of the
HH signaling pathway.
Juvenile Pilocytic Astrocytoma
Rhabdomyosarcoma (RMS) is the most common soft tissue

sarcoma in children and comprises two major histologic
subtypes: alveolar rhabdomyosarcoma (ARMS) and embryonal rhabdomyosarcoma (ERMS). Although RMS is frequently said to be associated with activating mutations in
the HH pathway, the role of the HH pathway in sporadic
human RMS is not clear, and linkage to specific HH pathway
mutations has not been confirmed.3 Although one study
reported mutations in PTCH1 and SUFU in RMS, most of
the analysis was restricted to loss of heterozygosity (LOH)
analysis of PTCH1 and SMO in rhabdomyoma, and in only
one case of RMS.21 Since this region contains other potential
tumor suppressor candidate genes, definitive conclusion
from LOH analysis alone cannot be drawn. Soft tissue
sarcomas that are very similar to human RMS are seen in
Ptc1 mice, but the incidence of these tumors is strongly
influenced by the murine genetic background.3
The initial descriptions of Gorlin syndrome identified an
association of the disease with benign fetal rhabdomyoma,
as well as rare instances of RMS. Tostar and colleagues
reported overexpression of PTCH1 and GLI1 mRNA, as
determined by in situ hybridization, in sporadic RMS.21
Using array comparative genomic hybridization to examine
a specific subset of clinically defined intermediate-risk
ERMS tumors, Paulson and colleagues similarly demonstrated that over 50% of tumors had low-level gains of a
region containing GLI1 along with a gene expression signature consistent with HH-pathway activation.22 Zibat and
colleagues reported that PTCH1, GLI1, GLI3, and MYF5 are
expressed at significantly higher levels in ERMS and that
GLI1 expression consistently correlates with PTCH1 expression in ERMS, as well as a specific subset of ARMS.23 This
study also showed that high PTCH1 expression significantly
correlated with reduced survival in a subset of RMS. Likewise, Pressey and colleagues demonstrated that expression
of GLI1, with or without PTCH1, is detectable in substantial
Renal Tumors
IHC analysis of seven Wilms tumors showed that five

(71%), seven (100%), and three (43%) stained positive for
SHH, PTCH1, and GLI1, respectively.18 Two renal clear cell
sarcoma and three rhabdoid tumors of the kidney showed
very high expression of HH pathway markers, suggesting
607
TOBEY J. MACDONALD
that the HH pathway may contribute to the more unfavorable biologic behaviors of these renal tumors.18
Hepatic Tumors
Eichenmuller and colleagues reported that HH signaling

is active in pediatric hepatoblastoma and showed that blocking HH signaling with cyclopamine in hepatoblastoma
cells leads to a significant decrease in cell viability and
increased apoptosis.29 In another study, IHC of 11 hepatoblastoma cases demonstrated that 100% of tumors stained
positive for SHH and PTCH1, while eight (73%) were positive for GLI1.18 One case of embryonal carcinoma and two
hepatic tumor cell lines (Heh6 and Heh7) showed strong
expression of all three HH pathway markers. No relationship was observed between GLI1 expression and histologic
subtype, clinical stage, or prognosis.
Osteosarcoma
Recently published data also suggest HH signaling may

play a role in the pathogenesis of osteosarcoma. A study of
osteosarcoma cell lines and biopsy specimens showed overexpression of HH target genes via real-time PCR.30 Inhibition of the HH pathway with cyclopamine in vitro promoted
G1 arrest and reduced the expression of positive cell-cycle
regulators, including cyclins D1 and E1. In addition, SMO
knockdown with SMO shRNA prevented the growth of
osteosarcoma, both in vitro and in vivo.30
HH Pathway Inhibitors in Pediatric Clinical Trials
Inhibitors of HH signaling have recently been the focus of

intense research. It is noteworthy that many tumors that
appear to lack specific HH pathway mutations have been
reported to be sensitive to SMO inhibitors in vitro and in
vivo.3 This has led to estimates that as many as 25% of
human tumors may depend on HH pathway activity for
growth. As a consequence, a wide range of tumors have been
included in the early-adult clinical trials investigating SMO
inhibitors.3 Xenograft tumor models have shown that hu-
Table 1. Hedgehog Pathway Inhibitors in Current Clinical Trials

Agent
Target
Mechanism
of Action
Manufacturer
GDC-0449*
LDE225*
LEQ506
PF-04449913
IPI-926
BMS-833923
SMO
SMO
SMO
SMO
SMO
SMO
Antagonist
Antagonist
Antagonist
Antagonist
Antagonist
Antagonist
GDC-0449 (Genentech)
LDE225 (Novartis)
LEQ506 (Novartis)
PF-04449913 (Pfizer)
IPI-926 (Ifinity)
BMS-833923 (Bristol-Myers Squibb)
* Currently being investigated in pediatric cancer. As of February 8, 2012

(Clinicaltrials.gov).
man tumors without HH pathway mutations that express

HH ligands actually induce upregulation of HH pathway
target genes in the stromal cells of mouse origin, rather than
in the tumor cells.3 Under these conditions, treatment with
SMO inhibitors inhibited tumor growth to some degree, but
did not eliminate the tumors. Some leukemias have also
been reported to depend on SMO signaling for growth, but
more recent studies indicate that survival was not dependent on HH ligands or SMO activity.3 The proof of concept
for the clinical utility of SMO inhibitors has been established in patients with metastatic or locally advanced BCC
or MB.15-17 The more recent data suggest that this class of
agents may have broader utility in other adult and pediatric
cancers. The dilemma is to identify tumors that would
benefit most from HH inhibitor treatment, as the mere
presence of HH ligands or a pathway signature does not
guarantee a response. Because of the importance of the HH
signaling pathway in the normal growth and development,
the use of SMO inhibitors in children is a specific concern
that will need to be carefully assessed in order to manage
the potential adverse effects on bone growth plates, developing teeth, and the immature reproductive system.3 At
present, early phase I and II clinical trials with the SMO
inhibitors LDE225 (Novartis) and GDC-0449 (Genentech)
are being evaluated in pediatric medulloblastoma and other
pediatric tumors that are potentially dependent on HH
pathway signaling (Table 1).
Author
Tobey J. MacDonald
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Novartis
REFERENCES
1. Nusslein-Volhard C, Wieschaus E. Mutations affecting segment number
and polarity in Drosophila. Nature. 1980;287:795-801.
2. Ingham PW, Nakano Y, Seger C. Mechanisms and functions of Hedgehog
signalling across the metazoa. Nat Rev Genet. 2011;12:393-406.
3. Ng JM, Curran T. The Hedgehogs tale: Developing strategies for
targeting cancer. Nat Rev Cancer. 2011;11:493-501.
4. McMahon AP, Ingham PW, Tabin CJ. Developmental roles and clinical
significance of hedgehog signaling. Curr Top Dev Biol. 2003;53:1-114.
5. Lee J, Platt KA, Censullo P, et al. Gli1 is a target of Sonic hedgehog that
induces ventral neural tube development. Development. 1997;124:2537-2552.
6. Johnson RL, Rothman AL, Xie J, et al. Human homolog of patched, a
candidate gene for the basal cell nevus syndrome. Science. 1996;272:1668-1671.
7. Epstein EH. Basal cell carcinomas: Attack of the hedgehog. Nat Rev
Cancer. 2008;8:743-754.
8. Raffel C, Jenkins RB, Frederick L, et al. Sporadic medulloblastomas
contain PTCH mutations. Cancer Res. 1997;57:842-845.
9. Thompson MC, Fuller C, Hogg TL, et al. Genomics identifies medullo-
608
blastoma subgroups that are enriched for specific genetic alterations. J Clin
Oncol. 2006;24:1924-1931.
10. Taylor MD, Liu L, Raffel C, et al. Mutations in SUFU predispose to
medulloblastoma. Nature Genet. 2002;31:306-310.
11. Pastorino, L. Ghiorzo P, Nasti S, et al. Identification of a SUFU
germline mutation in a family with Gorlin syndrome. Am J Med Genet A.
2009;149A:1539-1443.
12. Wetmore C, Eberhart DE, Curran T. Loss of p53 but not ARF accelerates medulloblastoma in mice heterozygous for patched. Cancer Res. 2001;61:
513-516.
13. Yang ZJ, Ellis T, Markant SL, et al. Medulloblastoma can be initiated
by deletion of Patched in lineage-restricted progenitors or stem cells. Cancer
Cell. 2008;14:135-145.
14. Hatton BA, Knoepfler PS, Kenney AM, et al. N-myc is an essential
downstream effector of Shh signaling during both normal and neoplastic
cerebellar growth. Cancer Res. 2006;66:8655-8661.
15. Von Hoff DD, LoRusso PM, Rudin CM, et al. Inhibition of the hedgehog
HEDGEHOG PATHWAY IN PEDIATRIC CANCER

pathway in advanced basal-cell carcinoma. N Engl J Med. 2009;361:11641172.
16. Romer JT, Kimura H, Magdaleno S, et al. Suppression of the Shh
pathway using a small molecule inhibitor eliminates medulloblastoma in
Ptc1(/)p53(/) mice. Cancer Cell. 2004;6:229-240.
17. Rudin CM, Hann CL, Laterra J, et al. Treatment of medulloblastoma
with hedgehog pathway inhibitor GDC-0449. N Engl J Med. 2009;361:11731178.
18. Oue T, Yoneda A, Uehara S, et al. Increased expression of the hedgehog
signaling pathway in pediatric solid malignancies. J Pediatr Surg. 2010;45:
387-392.
19. Monje M, Mitra SS, Freret ME, et al. Hedgehog-responsive candidate
cell of origin for diffuse intrinsic pontine glioma. Proc Natl Acad Sci U S A.
2011;108:4453-4458.
20. Rush SZ, Abel TW, Valadez, et al. Activation of the Hedgehog pathway
in pilocytic astrocytomas. Neuro Oncol. 2010;12:790-798.
21. Tostar U, Malm CJ, Meis-Kindblom JM, et al. Deregulation of the
hedgehog signalling pathway: a possible role for the PTCH and SUFU genes
in human rhabdomyoma and rhabdomyosarcoma development. J Pathol.
2006;208:17-25.
22. Paulson V, Chandler G, Rakheja D, et al. High-resolution array CGH
identifies common mechanisms that drive embryonal rhabdomyosarcoma
pathogenesis. Genes Chromosomes Cancer. 2011;50:397-408.
23. Zibat A, Missiaglia E, Rosenberger A, et al. Activation of the hedgehog

pathway confers a poor prognosis in embryonal and fusion gene-negative
alveolar rhabdomyosarcoma. Oncogene. 2010;29:6323-6330.
24. Pressey JG, Anderson JR, Crossman DK, et al. Hedgehog pathway
activity in pediatric embryonal rhabdomyosarcoma and undifferentiated
sarcoma: A report from the Childrens Oncology Group. Pediatr Blood Cancer.
2011;57:930-938.
25. Rubin BP, Nishijo K, Chen HI, et al. Evidence for an unanticipated
relationship between undifferentiated pleomorphic sarcoma and embryonal
rhabdomyosarcoma. Cancer Cell. 2011;19:177-191.
26. Ecke I, Rosenberger A, Obenauer S, et al. Cyclopamine treatment of
full-blown Hh/Ptch-associated RMS partially inhibits Hh/Ptch signaling, but
not tumor growth. Mol Carcinog. 2008;47:361-372.
27. Souzaki R, Tajiri T, Souzaki M, et al. Hedgehog signaling pathway in
neuroblastoma differentiation. J Pediatr Surg. 2010;45:2299-2304.
28. Gershon TR, Shirazi A, Qin LX, et al. Enteric neural crest differentiation in ganglioneuromas implicates Hedgehog signaling in peripheral neuroblastic tumor pathogenesis. PLoS One. 2009;4(10):e7491.
29. Eichenmuller M, Gruner I, Hagl B, et al. Blocking the hedgehog
pathway inhibits hepatoblastoma growth. Hepatology. 2009;49:482-490.
30. Hirotsu M, Setoguchi T, Sasaki H, et al. Smoothened as a new
therapeutic target for human osteosarcoma. Molecular Cancer. 2010;9:5.
609
PEDIATRIC ONCOLOGY EDUCATIONAL SESSION IN

HONOR OF DR. JAMES NACHMAN
CHAIR
Stephen P. Hunger, MD
University of Colorado Denver School of Medicine
Aurora, CO
SPEAKERS
Melissa M. Hudson, MD
Memphis, TN
Carola A. S. Arndt, MD
Mayo Clinic
Rochester, MN
D e v e l o p m e n t a n d R e fi n e m e n t o f A u g m e n t e d
Treatment Regimens for Pediatric High-Risk
Acute Lymphoblastic Leukemia
By Stephen P. Hunger, MD
Overview: The 5-year survival rate for children and adolescents with acute lymphoblastic leukemia (ALL) is now at least
90%. However, clinical features (age and initial white blood
cell count [WBC]), early treatment response, and the presence/absence of specific sentinel genomic lesions can identify
subsets of high-risk (HR) ALL patients with a much higher risk
of treatment failure. Chemotherapy regimens used to treat HR
ALL have been refined over the past 3 decades through
randomized clinical trials conducted by the Childrens Oncology Group (COG) in North America and the Berlin-FrankfurtMuenster (BFM) group in Western Europe. Contemporary COG
HR ALL treatment regimens were developed from the BFM-76
regimen, with subsequent changes that led to development
and refinement of a so-called augmented BFM (ABFM) regimen
used today. Although contemporary COG and BFM treatment
regimens are not identical, there are many more similarities

than differences. With improvements in survival, it has become clear that although the outcome of some patients with
HR ALL can be improved by optimizing use of standard
cytotoxic chemotherapy agents, this approach has had only
limited success for other patient subsets. In contrast, introduction of the tyrosine kinase inhibitor imatinib has led to
dramatic outcome improvements for children and adolescents
with Philadelphia chromosomepositive ALL. Genomic studies
are identifying new sentinel genomic lesions that can serve as
potential therapeutic targets, which will likely lead to the
testing of novel and/or targeted therapies in more children
with HR ALL. Such studies will require increased collaboration
between Western European and North American cooperative
groups.
treatment regimen7 that bears significant similarity to, but

has fundamental differences from, the BFM regimen developed in the early 1970s by Riehm and colleagues in Berlin.8
Table 1 provides a comparison of different contemporary
BFM and COG treatment regimens. Different therapies are
used for certain HR subtypes of ALL, most notably Phpositive ALL that is treated with intensive chemotherapy
and tyrosine kinase inhibitors (TKI) such as imatinib or
dasatinib.9
LL IS the most common cancer that occurs in children

and adolescents younger than age 20, comprising
approximately 25% of malignancies occurring before age 15
and 12% of those occurring in adolescents aged 15 to 20.1
ALL was virtually incurable until the early- to mid-1960s,
with less than a 10% survival rate as recently as the late
1960s. A recent large review of results of COG clinical trials
showed 5-year survival rates of 90% for more than 7,000
children diagnosed with ALL between 2000 and 2005, and it
is anticipated that the long-term survival rate for those
diagnosed between 2006 and 2010 will approach or exceed
90%.2 Despite these dramatic improvements in survival,
clinical features, early-treatment response characteristics,
and the presence/absence of specific favorable- or poor-risk
sentinel genetic lesions in the leukemia cells can be used to
identify patient subsets at higher risk of relapse.3 Powerful
predictive characteristics include the combination of age and
WBC at initial diagnosis, which together are used to define
the so-called NCI/Rome standard risk (SR; age 1 to 9.99
years and WBC less than 50,000/microliter) and HR (age
1 or 10 years and/or WBC 50,000 microliter) groups.4
The second powerful predictor of outcome is initial treatment response, measured either by response to a 1-week
prednisone (PRED) (plus a single dose of intrathecal methotrexate [IT MTX]) prephase (also termed prophase), bone
marrow morphology after 1 to 2 weeks of multiagent chemotherapy, or measurements of minimal residual disease
(MRD) at end of induction and/or consolidation therapy. The
MRD response is the most powerful single prognostic factor
and is now used by most groups in North America and
Western Europe to modulate the intensity of postinduction
therapy.5,6 Leukemia genetics also provides critical prognostic information, with ETV6-RUNX1 (TEL-AML1) fusion and
hyperdiploidy and/or favorable chromosome trisomies recognized as favorable features, and BCR-ABL1 fusion (Phpositive ALL) or MLL translocations generally considered to
be adverse features that merit specific therapies and/or
intensified treatment.3
HR ALL is generally treated with more intensive therapies than SR ALL. The COG uses an ABFM baseline
Development and Refinement of the BFM Treatment

Regimen for ALL
In 1977, Riehm and colleagues reported outstanding early

results of treatment outcome with an intensive 8-week,
eight-drug induction regimen, cranial irradiation, and maintenance therapy.8 The 8-week induction was later termed
protocol I and included two distinct phases. Protocol Ia is
what the COG calls a four-drug induction and included
PRED, asparaginase, vincristine, and daunorubicin. This
was immediately followed by protocol Ib (the COG consolidation), which included cyclophosphamide, repeated low
doses of ara-C, 6-mercaptopurine (6-MP), four weekly doses
of IT MTX, and cranial irradiation (18 Gy).
Subsequently, the BFM 76/79 study showed that outcome,
particularly for patients with a high WBC, could be improved when protocol I was repeated with some modifications designed to minimize drug resistance (replacement of
PRED with dexamethasone [DEX], daunorubicin with doxorubicin, and 6-MP with 6-thioguanine). This was termed
protocol II (delayed intensification [DI] in COG terminology). The 10-year disease-free survival rate was 67% for
patients treated in the BFM 76/79 trial with protocols I and
From the Childrens Hospital Colorado, Aurora, CO; Department of Pediatrics, University of Colorado School of Medicine, Aurora, CO.
Address reprint requests to Stephen P. Hunger, MD, Center for Cancer and Blood
Disorders, Childrens Hospital Colorado, 13123 East 16th Ave., Box B115, Aurora, CO
80045; email: stephen.hunger@childrenscolorado.org.
1092-9118/10/1-10
611
STEPHEN P. HUNGER
Table 1. Comparison of Components of BFM and COG ALL Therapies
Treatment Phase
BFM
BFM HR
COG SR ALL
Induction (protocol Ia)
4 drugs/4 wks
PRED, Vcr, ASNase, Dauno
4 drugs/4 wks
PRED, Vcr, ASNase, Dauno
3 drugs/4 wks; DEX, Vcr,

ASNase
Consolidation (protocol Ib)
6 wks
CPM, Ara-C, 6-MP
HD MTX
6 wks
CPM, Ara-C
Three intensive 3-wk HR blocks
IM #2
8 wks
DEX, Vcr, ASNase, Doxo, CPM,
Ara-C, 6-TG
Not given
8 wks
Ara-C, 6-TG
4 wks
6-MP, MTX
4 wks
6-MP, MTX
8 wks
Capizzi MTX, Dex, Vcr
8 wks
Ara-C, 6-TG
8 wks
Capizzi MTX, DEX, Vcr
DI #2 (protocol II)
Not given
Maintenance
6-MP, MTX
8 wks
Ara-C, 6-TG
6-MP, MTX
Treatment duration
30 mo
30 mo
IM #1
DI #1 (protocol II)
COG HR ALL (hABFM)
Not given
4 drugs/4 wks; PRED (age 10) or

DEX (age 19.99), Vcr, ASNase,
Dauno
8 wks (augmented)
CPM, Ara-C, 6-MP, Vcr, ASNase
8 wks
HD MTX, Vcr
8 wks (augmented)
Ara-C, 6-TG
8 wks
Capizzi MTX ASNase, DEX, Vcr,
(some not all patients)
Not given
6-MP, MTX plus monthly

DEX/Vcr pulses
27 mo, females; 39 mo, males
6-MP, MTX plus monthly PRED/Vcr

pulses
27 mo, females; 39 mo, males
Abbreviations: ALL, acute lymphoblastic leukemia; ASNase, asparaginase; BFM, Berlin-Frankfurt-Muenster; COG, Childrens Oncology Group; CPM, cyclophosphamide; Dauno, daunorubicin; DEX, dexamethasone; DI, delayed intensification; Doxo, doxorubicin; hABFM, hemi-augmented BFM; HR, high risk; IM, interim
maintenance; mo, months; MTX, methotrexate; PRED, prednisone; SR, standard risk; Vcr, vincristine; wks, weeks; 6-MP, 6-mercaptopurine; 6-TG, 6-thioguanine.
II separated by an 8-week interim maintenance (IM) phase

followed by maintenance therapy.10
Subsequently, the BFM has refined therapy to include the
PRED prephase and replaced the 8-week IM phase with
protocol M that consists of four courses of high-dose methotrexate (HD MTX; 5 g/m2 over 24 hours followed by leucovorin rescue) given at 2-week intervals, with outstanding
contemporary outcomes.6 A fundamental difference between
BFM and COG regimens is that the BFM regimens do not
include steroid/vincristine pulses during maintenance.11 Patients identified to be at HR of relapse based on a poor
response to the PRED prephase (more than 1000 blasts/
microliter at day 8), adverse cytogenetics, or more recently
high levels of MRD at end induction/consolidation receive
KEY POINTS
612
Acute lymphoblastic leukemia (ALL) is the most

common pediatric cancer.
Five-year survival rates for pediatric ALL now exceed
90%, as compared to less than 10% in the 1960s.
Contemporary Childrens Oncology Group ALL treatment regimens are derived from regimens developed
by the Berlin-Frankfurt-Muenster group in the 1970s
but have been refined through a series of randomized
clinical trials.
Optimizing the use of standard cytotoxic chemotherapy agents has led to significant outcome improvements for some subsets of high-risk ALL patients,
whereas other subsets have benefitted little from this
approach.
Novel and/or targeted therapies have benefitted children and adolescents with Philadelphia chromosomepositive ALL and will likely be tested in other
high-risk ALL subsets.
more intensive therapy in which protocol M is replaced by

three intensive multiagent chemotherapy blocks (HR blocks)
and protocol II is given twice.12
CCG/COG Adoption of BFM-76-Based Therapy
The Childrens Cancer Group (CCG) recognized in the

1980s that outcomes reported by the BFM were superior to
those obtained in CCG trials with less intensive therapy.
This led to development of two pivotal trials that compared
the CCG regimens of that time to the BFM-76 regimen,
which did not include a PRED prephase or protocol M.
Because of this, CCG and COG ALL treatment regimens
resemble but do not recapitulate BFM therapies, although
they are now becoming more similar, with COG adoption of
HD MTX (see below).
The CCG 105 trial (1983 to 1989) was designed for
children with intermediate-risk ALL, which is similar but
not identical to NCI SR ALL.13 Children enrolled in CCG
105 were randomly selected to receive the standard CCG
regimen of that time, which included a three-drug induction
(PRED, vincristine, and asparaginase but no daunorubicin)
followed by central nervous system (CNS) control and maintenance, or the BFM regimen that included protocols I and
II. Two other randomly selected groups received protocol I
followed by CCG postinduction therapy or the CCG induction/consolidation followed by protocol II. The results of this
study (updated in 2005 with 16-year follow-up) showed that
all three of the BFM-based regimens were superior to the
CCG regimen.13,14 There was little outcome difference between the three BFM-based regimens; thus, the CCG selected the CCG induction/consolidation followed by protocol
II to bring forward in subsequent trials, as that regimen had
less short-term toxicity than the other two BFM-based
regimens. For this reason, all subsequent CCG and COG
regimens for SR ALL have included a three-drug induction
(without the intensive consolidation phase) and a DI phase.
Based on the results of the CCG 1922 study, children with
SR ALL enrolled in COG ALL trials now receive DEX rather
ABFM THERAPY FOR PEDIATRIC ALL
than PRED during induction and in the steroid/vincristine

pulses given every 4 weeks during maintenance therapy.15
The CCG 1991 trial showed that Capizzi escalating intravenous (IV) MTX without rescue (no asparaginase) was superior to oral 6-MP and MTX during the IM phase(s),16 and
this approach is now standard in contemporary COG SR
ALL trials.
In parallel to CCG 105, the CCG 106 trial (1983 to 1987)
compared BFM therapy to the contemporary CCG HR regimen and to the more intensive New York regimen.17
Better outcomes were seen with both the BFM and New
York regimens as compared to the CCG regimen, but the
BFM regimen was associated with less toxicity and lower
cumulative doses of chemotherapy than the New York regimen. Based on the results of CCG 106, the BFM-76-based
regimen became the baseline CCG/COG regimen used for
HR ALL, and subsequent trials included a four-drug induction, the intensive consolidation (protocol Ib), and one or
more protocol II (DI) phases.
Development of Refinement of ABFM Therapy
by the CCG/COG
The CCG recognized that a poor early response to chemotherapy was a strong adverse prognostic factor.18 In CCG
1882 (1991 to 1995), children with HR ALL and a slow early
response to therapy (SER; more than 25% marrow blasts at
day 8 of induction) were randomly selected to received the
CCG-modified BFM regimen or an ABFM regimen that
contained a number of changes to baseline treatment.7
These changes included intensifying consolidation (Ib) therapy by extending it to 8 weeks and including doses of
vincristine and asparaginase during the neutropenic phases
that followed cyclophosphamide and ara-C, using Capizzi I
escalating IV MTX without leucovorin rescue plus asparaginase during the 8-week IM #1 phase, giving second IM and
DI phases, and giving prophylactic cranial irradiation to all
patients. The ABFM regimen produced results that were
significantly better than the standard CCG regimen, with
5-year event-free survival (EFS) rates of 75% versus 55%
(p 0.001) and overall survival (OS) rates of 78% versus
67% (p 0.02).7
Based on the results of CCG 1882, the subsequent CCG
1961 HR-ALL trial tested the ABFM regimen in patients
with a rapid early response to therapy (day 8 marrow blasts
less than or equal to 25%) and attempted to determine the
most important components of ABFM therapy by randomly
selecting patients in a 2 x 2 manner to receive the baseline
regimen, the full ABFM regimen, the baseline regimen with
2 IM and DI phases, or the ABFM regimen with only single
IM and DI phases (termed hemi-ABFM, or hABFM).19 The
results of CCG 1961 showed that the augmented parts of
ABFM therapy were critical and improved 5-year EFS (81%
vs. 72%; p 0.001) and OS (89% vs. 83%; p 0.003)
significantly. Equally important, repeating the IM and DI
phases did not improve outcome (5-year EFS: 76% vs. 76.8%
for single compared with double IM/DI; p 0.94). Based on
these results, the hABFM regimen with single IM/DI phases
became the standard regimen for children with HR ALL and
a good response to induction therapy in COG trials, although
a second IM and/or DI phase has been retained in some
trials for those with a poor early response.
COG AALL0232 Shows Superiority of HD MTX
The COG AALL0232 (2003 to 2011) trial for children and

adolescents with HR B-cell precursor (BCP) ALL used the
hABFM regimen and had a 2 x 2 randomization to 28 days
of PRED 60 mg/m2/day versus 14 days of DEX 10 mg/m2/day
during induction and to HD MTX versus Capizzi MTX plus
asparaginase during IM #1. Accrual to this trial was stopped
in early 2011 when results for the MTX randomization
crossed predefined monitoring boundaries. Eric Larsen, the
COG AALL0232 study chair, presented results of the MTX
randomization at the plenary session of the 2011 ASCO
annual meeting.20 Planned interim monitoring showed that
the 5-year EFS for patients randomly selected to receive
HD-MTX (1,209 patients) was 82% 3.4% vs. 75.4% 3.6%
for the Capizzi MTX plus asparaginase (1,217 patients)
regimen, p 0.006. Based on these practice-changing results, the COG now considers HD MTX to be the standard of
care for BCP HR ALL, further increasing the symmetry
between COG and BFM regimens.
The results of the induction steroid randomization were
more complex. In 2008, the COG stopped the steroid randomization in children age 10 or older because of an increased incidence of osteonecrosis (ON) among patients of
this age treated with DEX. Because the rate of ON was quite
low in children younger than age 10 and there was no
difference in ON rates between the two steroid arms, the
randomization continued for younger patients. There was an
interaction between the steroid and MTX regimens among
patients younger than age 10, so when the outcome for
children randomly selected to receive DEX versus PRED
was compared, the analysis was limited to those randomly
selected to receive HD MTX.21 The younger children who
received DEX/HD MTX had a superior outcome to those who
received PRED/HD MTX (5-year EFS: 93.7% 5.4% vs.
81.2% 7.7%; p 0.03). Retrospective review of the
patients age 10 or older who were randomly selected to
receive DEX versus PRED during the initial 4 years of the
study showed no differences in outcome (5-year EFS: 74.7%
4.6% and 76.5% 4.6%, respectively; p 0.80) and
confirmed the higher rate of ON seen in the patients age 10
or older who were treated with DEX (24.3% vs. 15.1%; p
0.0007). Based on these results, 14 days of DEX 10 mg/m2/
day on days 1 through 14 is now considered the standard
regimen for children younger than age 10 enrolled in COG
HR BCP ALL trials, whereas PRED 60 mg/m2/day on days 1
through 28 is considered the standard for those age 10 or
older.
COG T-ALL Trials
Because of differences between T-cell ALL (T-ALL) and

BCP ALL in biology, treatment response, prognostic factors,
and outcome, the COG is conducting separate trials for BCP
and T-ALL, whereas the BFM group has a single trial that
includes both T-ALL and BCP ALL. The backbone therapy
for the COG AALL0434 T-ALL study (opened to patient
accrual in 2007 and currently projected to complete accrual
in 2014) is the PRED/Capizzi MTX plus asparaginase arm of
AALL0232 (see above). The study is a 2 x 2 randomized trial.
The first comparison is the identical HD MTX versus Capizzi
MTX plus asparaginase randomization conducted in
AALL0232. Careful consideration was given to the results of
AALL0232 when they became available, with a decision to
613
STEPHEN P. HUNGER
continue this randomization based on the different biology of

BCP and T-ALL (the reason that the question was asked
separately in two parallel trials in the first place). Interestingly, the best reported contemporary results in pediatric
T-ALL come from the UKALL 2003 trial, which used essentially a COG ABFM backbone regimen with Capizzi MTX
plus asparaginase during IM.22
The second randomization in AALL0434 is to receive
versus not receive six 5-day courses of nelarabine during the
first 16 months of therapy. Nelarabine, a prodrug of ara-G,
showed substantial clinical activity in phase I/II T-ALL
trials, but was associated with significant, often severe, and
sometimes fatal CNS toxicity.23,24 The COG then conducted
a pilot study of nelarabine plus BFM-86 type chemotherapy
in children with newly diagnosed T-ALL and found encouraging activity but much lower rates of overall and severe
toxicity than had been seen in relapsed patients treated in
the phase I/II trials.25 AALL0434 included a pilot safety
phase during which only an HR subset of T-ALL patients
with a poor response was randomly selected to be treated
with/without nelarabine. This safety phase was concluded
with no significant differences in toxicity among the patients
treated with/without nelarabine, and AALL0434 is now in
the efficacy phase in which 90% of patients are randomly
selected to be treated with/without nelarabine (all except for
a tightly defined low-risk patient subset).26
VHR ALL
A small subset of pediatric ALL patients can be defined as

being at very high risk (VHR) of relapse. The definition of
this subgroup is constantly evolving, but there is a general
consensus that alternative treatment strategies may be
indicated for these patients in comparison with treatment
given to other patients with HR ALL. As discussed above,
the BFM group has termed this group HR and uses intensive
multiagent chemotherapy HR blocks, an additional protocol
II for these patients, with hematopoietic stem cell transplant (HSCT) in CR1 for the subset with a poor treatment
response at end of protocol Ib. The COG AALL0031 pilot
study (2002 to 2006) tested an intensive multiagent chemotherapy regimen in children with VHR ALL, which included
Ph-positive ALL, hypodiploidy (chromosome number less
than 44), poor response to induction therapy (more than 25%
marrow blasts at day 29 or 5% to 25% marrow blasts/more
than 1% MRD at day 43 after 2 weeks of extended induction
therapy), or MLL translocation and an SER (more than 5%
blasts at day 15 of induction).9,27 The results were encouraging for patients with Ph-negative VHR ALL, with a
nonsignificant trend toward improved outcome for those
treated with HSCT.27 The study design with all patients
receiving identical chemotherapy and those with Ph-positive
ALL also receiving imatinib facilitated assessment of imatinib toxicity. Notably, the addition of imatinib was quite
safe and did not lead to increased toxicity.9 More importantly, adding imatinib led to major improvements in early
614
outcome for patients with Ph-positive ALL, with a 3-year

EFS rate of 80% 11% (95% CI, 64% to 90%), which was
more than twice that of historic controls treated in the
preimatinib era (35% 4%; p 0.0001). There was no
apparent advantage to HSCT as compared with intensive
chemotherapy plus imatinib. These results have been stable
with longer follow-up28 and have led to changes in clinical
practice in treatment of children with Ph-positive ALL.
The COG has now completed enrollment (2008 to 2012) on
a successor Ph-positive ALL trial (AALL0622) that utilized
the AALL0031 chemotherapy backbone with imatinib replaced with dasatinib, a more potent second-generation Abl
TKI. The study showed that it was safe to add continuous
dasatinib treatment (60 mg/m2/day) to this intensive chemotherapy regimen; it is still much too early to assess treatment efficacy. In 2012, the COG and European EsPhALL
groups will commence enrollment on a joint pilot study (developed in conjunction with Bristol Myers Squibb) that will test
the safety and efficacy of dasatinib added to the BFM HR
treatment backbone, which has lower cumulative doses of
many chemotherapy agents than the AALL0031 regimen.
Future Perspectives
The substantial improvements that have occurred in outcome for pediatric HR ALL over the past few decades have
been accompanied by recognition that there are a number of
different ALL subsets that come under the umbrella term
HR ALL. For some subsets, outcomes have been improved
by optimizing delivery of chemotherapy agents that have
been in widespread clinical use for the past 25 years, as
exemplified by the results of COG AALL0232. For other
subsets, such as Ph-positive ALL, optimizing traditional
cytotoxic chemotherapy agents led to limited improvements
in outcome,29 but introduction of new and/or targeted therapies had a major effect on outcome.9 It is expected that
ongoing genomic studies will lead to recognition of additional VHR ALL subsets defined by the presence of specific
sentinel genomic lesions that can potentially be targeted by
novel therapies.30 The rarity of these and other VHR ALL
subsets will require increased collaborations between North
American and Western European investigators in order to
test the efficacy of novel/targeted therapies in these patient
subsets.
Acknowledgments
The author dedicates this manuscript to the memory of Jim
Nachman, who passed away unexpectedly, and far too early, in
2011. Jim developed the ABFM regimen; played a major role in
developing productive, fruitful interactions between the CCG/
COG and BFM groups; and shaped the design of most COG ALL
trials of the past 25 years. He was widely recognized as an
international leader in pediatric oncology clinical research and
as a mentor to many investigators. The author and the rest of
the pediatric oncology community miss his advice, his infectious
enthusiasm, his laugh and politically incorrect sense of humor,
and most of all his friendship.
ABFM THERAPY FOR PEDIATRIC ALL
Author
Stephen P. Hunger
Employment or
Leadership
Positions
Consultant or
Advisory Role
Bristol-Myers
Squibb (U);
Genzyme (U)
Stock
Ownership
Amgen (B);
Bristol-Myers
Squibb (I);
Merck (B);
Pfizer (B)
Honoraria
Research
Funding
Expert
Testimony
Bristol-Myers
Squibb; Genzyme
(U)
Other
Remuneration
REFERENCES
1. Ries LA, Smith MA, Gurney JG, et al (eds). Cancer Incidence and
Survival among Children and Adolescents: United States SEER Program
1975-1995. Bethesda, MD: National Cancer Institute; 1999.
2. Hunger SP, Lu X, Devidas M, et al. Improved survival for children and
adolescents from 1990-2005: a report from the Childrens Oncology Group.
J Clin Oncol. In press.
3. Schultz KR, Pullen DJ, Sather HN, et al. Risk- and response-based
classification of childhood B-precursor acute lymphoblastic leukemia: a combined analysis of prognostic markers from the Pediatric Oncology Group
(POG) and Childrens Cancer Group (CCG). Blood. 2007;109:926-935.
4. Smith M, Arthur D, Camitta B, et al. Uniform approach to risk
classification and treatment assignment for children with acute lymphoblastic leukemia. J Clin Oncol. 1996;14:18-24.
5. Borowitz MJ, Devidas M, Hunger SP, et al. Clinical significance of
minimal residual disease in childhood acute lymphoblastic leukemia and its
relationship to other prognostic factors: a Childrens Oncology Group study.
Blood. 2008;111:5477-5485.
6. Conter V, Bartram CR, Valsecchi MG, et al. Molecular response to
treatment redefines all prognostic factors in children and adolescents with
B-cell precursor acute lymphoblastic leukemia: results in 3184 patients of the
AIEOP-BFM ALL 2000 study. Blood. 2010;115:3206-3214.
7. Nachman JB, Sather HN, Sensel MG, et al. Augmented post-induction
therapy for children with high-risk acute lymphoblastic leukemia and a slow
response to initial therapy. N Engl J Med. 1998;338:1663-1671.
8. Riehm H, Gadner H, Welte K. The West-Berlin therapy study of acute
lymphoblastic leukemia in childhoodreport after 6 years (authors transl).
Klin Padiatr. 1977;189:89-102.
9. Schultz KR, Bowman WP, Aledo A, et al. Improved early event-free
survival with imatinib in Philadelphia chromosome-positive acute lymphoblastic leukemia: a Childrens Oncology Group study. J Clin Oncol. 2009;27:
5175-5181.
10. Henze G, Langermann HJ, Bramswig J, et al. The BFM 76/79 acute
lymphoblastic leukemia therapy study (authors transl). Klin Padiatr. 1981;
193:145-154.
11. Conter V, Valsecchi MG, Silvestri D, et al. Pulses of vincristine and
dexamethasone in addition to intensive chemotherapy for children with
intermediate-risk acute lymphoblastic leukaemia: a multicentre randomised
trial. Lancet. 2007;369:123-131.
12. Stanulla M, Schrappe M. Treatment of childhood acute lymphoblastic
leukemia. Semin Hematol. 2009;46:52-63.
13. Tubergen DG, Gilchrist GS, OBrien RT, et al. Improved outcome with
delayed intensification for children with acute lymphoblastic leukemia and
intermediate presenting features: a Childrens Cancer Group phase III trial.
J Clin Oncol. 1993;11:527-537.
14. Hunger SP, Winick NJ, Sather HN, et al. Therapy of low-risk subsets of
childhood acute lymphoblastic leukemia: when do we say enough? Pediatr
Blood Cancer. 2005;45:876-880.
15. Bostrom BC, Sensel MR, Sather HN, et al. Dexamethasone versus
prednisone and daily oral versus weekly intravenous mercaptopurine for
patients with standard-risk acute lymphoblastic leukemia: a report from the
Childrens Cancer Group. Blood. 2003;101:3809-3817.
16. Matloub Y, Bostrom BC, Hunger SP, et al. Escalating intravenous
methotrexate improves event-free survival in children with standard-risk
acute lymphoblastic leukemia: a report for the Childrens Oncology Group.
Blood. 2011;118:243-251.
17. Gaynon PS, Steinherz PG, Bleyer WA, et al. Improved therapy for
children with acute lymphoblastic leukemia and unfavorable presenting
features: a follow-up report of the Childrens Cancer Group Study CCG-106.
J Clin Oncol. 1993;11:2234-2242.
18. Gaynon PS, Desai AA, Bostrom BC, et al. Early response to therapy and
outcome in childhood acute lymphoblastic leukemia: a review. Cancer. 1997;
80:1717-1726.
19. Seibel NL, Steinherz PG, Sather HN, et al. Early postinduction intensification therapy improves survival for children and adolescents with highrisk acute lymphoblastic leukemia: a report from the Childrens Oncology
Group. Blood. 2008;111:2548-2555.
20. Larsen EC, Salzer WL, Devidas M, et al. Comparison of high-dose
methotrexate (HD-MTX) with Capizzi methotrexate plus asparaginase (CMTX/ASNase) in children and young adults with high-risk acute lymphoblastic leukemia (HR-ALL): a report from the Childrens Oncology Group study
AALL0232. J Clin Oncol. 2011;29:6s (suppl; abstr 3).
21. Winick NJ, Salzer WL, Devidas M, et al. Dexamethasone (DEX) versus
prednisone (PRED) during induction for children with high risk acute
lymphoblastic leukemia (HR-ALL): a report from the Childrens Oncology
Group study AALL0232. J Clin Oncol 2011;29:586s (suppl; abstr 9504).
22. Vora A, Wade R, Mitchell CD, et al. Improved outcome for children and
young adults with T-cell acute lymphoblastic leukaemia (ALL): results of the
United Kingdome Medical Research Council (MRC) Trial UKALL 2003. Blood
(ASH Annual Meeting Abstracts). 2008;112:908.
23. Kurtzberg J, Ernst TJ, Keating MJ, et al. Phase I study of 506U78
administered on a consecutive 5-day schedule in children and adults with
refractory hematologic malignancies. J Clin Oncol. 2005;23:3396-3403.
24. Berg SL, Blaney SM, Devidas M, et al. Phase II study of nelarabine
(compound 506U78) in children and young adults with refractory T-cell
malignancies: a report from the Childrens Oncology Group. J Clin Oncol.
2005;23:3376-3382.
25. Dunsmore K, Devidas M, Borowitz MJ, et al. Nelarabine in combination with intensive modified BFM AALL00P2: a pilot study for the treatment
of high-risk T-ALL a report from the Childrens Oncology Group. J Clin Oncol.
2008;26:539s (suppl; abstr 10002).
26. Winter SS, Devidas M, Wood BL, et al. Nelerabine may be safely
incorporated into a phase III study for newly diagnosed T-lineage acute
lymphoblastic leukemia: a report from the Childrens Oncology Group. Blood
(ASH Annual Meeting Abstracts). 2010;116:865a.
27. Schultz KR, Bowman WP, Slayton WB, et al. Philadelphia chromosome
negative (Ph-) very high risk (VHR) acute lymphoblastic leukemia (ALL) in
children and adolescents: the impact of intensified chemotherapy on early
event free survival (EFS) in Childrens Oncology Group study AALL0031.
Blood (ASH Annual Meeting Abstracts). 2008;112:911.
28. Schultz KR, Bowman WP, Aledo A, et al. Continuous dosing imatinib
with intensive chemotherapy gives equivalent outcomes to allogeneic BMT for
Philadelphia chromosome-positive (Ph) acute lymphoblastic leukemia
(ALL) with longer term follow up: Updated results of Childrens Oncology
Group (COG) AALL0031. Pediatr Blood Cancer. 2010;54:788.
29. Arico` M, Schrappe M, Hunger SP, et al. Clinical outcome of children
with newly diagnosed Philadelphia chromosome-positive acute lymphoblastic
leukemia treated between 1995 and 2005. J Clin Oncol. 2010;28:4755-4761.
30. Hunger SP, Raetz EA, Loh ML, et al. Improving outcomes for high-risk
ALL: translating new discoveries into clinical care. Pediatr Blood Cancer.
2001;56:984-993.
615
R e fi n i n g t h e R o l e o f R a d i a t i o n T h e r a p y i n
Pediatric Hodgkin Lymphoma
By Melissa M. Hudson, MD, and Louis S. Constine, MD
Overview: The role of radiation therapy in the treatment of

pediatric Hodgkin lymphoma has continued to be refined,
motivated by the desire to avoid disruption to normal tissue
development and function and secondary carcinogenesis.
Such progress has occurred in tandem with modifications of
the multiagent chemotherapy regimens that have been used in
place of or in combination with low-dose involved-field radiation that are also associated with dose-related risks of
cardiopulmonary and gonadal dysfunction and leukemogenesis. Consequently, treatment strategies for young patients,
who have an excellent prognosis of long-term survival, utilizes
ADIATION THERAPY (RT) has played a seminal role

in improving outcomes for children and adolescents
with Hodgkin lymphoma (HL). Following delineation of a
tumoricidal dose, RT became the first modality to produce
prolonged disease-free survival when delivered to consistent
treatment fields of contiguous lymph nodes. Although curative for a large number of patients with localized disease, its
combination with multiagent chemotherapy was needed to
improve long-term survival in individuals with advancedstage and/or bulky node disease. Early RT approaches for
children and adults prescribed high doses (35 44 Gy) to
treatment volumes routinely extended to encompass adjacent uninvolved nodal regions. Recognition of the adverse
effects of high-dose RT on musculoskeletal development in
children motivated investigations of multiagent chemotherapy alone or in combination with lower radiation doses
(1525.5 Gy) to reduce treatment volumes (involvedfields).1,2 Increasing numbers of aging pediatric HL survivors consequently permitted recognition of the excess risk of
cardiovascular disease and secondary carcinogenesis associated with RT.3,4 This knowledge led to the abandonment of
the use of RT as a single modality and its restricted use in
contemporary trials.
Research elucidating unique profiles of chemotherapyrelated toxicity in children subsequently guided the development of multiagent chemotherapy regimens that balanced
the dose-related toxicity of anthracyclines, alkylating agents,
and bleomycin, often by adding the modality of low-dose
involved- field radiation therapy (IFRT). Results from these
studies provide strong support for the responsiveness of HL
to a variety of multiagent chemotherapy combinations that
are largely derived from the original MOPP (mechlorethamine, vincristine, procarbazine, prednisone)5 and ABVD
(doxorubicin, bleomycin, vinblastine, dacarbazine)6 combinations. The desire to maximize treatment efficacy and
minimize its related long-term morbidity has increasingly
focused attention on the role of RT in the treatment of
pediatric HL. This manuscript aims to review published
data to define the optimal treatment strategies for children
and adolescents with HL in regards to the use of RT.
Prognostic Factors Used in Risk Designation of
Pediatric HL
As therapy for pediatric HL becomes more effective, factors associated with outcome have become more difficult to
616
a risk-adapted approach that provides optimal efficacy for

disease control whereas limiting toxicity associated with both
radiation and chemotherapy. Because of the differences in
age-related developmental status and gender-related sensitivity to chemotherapy and radiation toxicity, no single treatment approach is ideal for all pediatric patients. This
manuscript summarizes results from published clinical trials
with the goal of defining optimal treatment strategies for
children and adolescents with Hodgkin lymphoma in regards
to the use of radiation therapy.
identify. Regardless, contemporary treatment for pediatric

HL uses a risk-adapted and response-based paradigm that
assigns the length and intensity of therapy based on diseaserelated factors such as stage, number of involved nodal
regions, tumor bulk, the presence of B symptoms, and early
response to chemotherapy by functional imaging. In addition to consideration of cancer-related factors that may
permit therapy reduction or require dose intensification, the
treatment approach may also consider unique host factors,
such as age and gender, that may enhance the risk for
specific treatment toxicities. Most protocols stratify groups
according to low-, intermediate-, and high-risk designations.
Low-risk clinical features typically include localized nodal
involvement in the absence of B symptoms and bulky disease. Risk factors considered in other studies include the
number of involved nodal regions, the presence of hilar
adenopathy, the size of peripheral lymphadenopathy, and
extranodal extension. High-risk clinical features include the
presence of B symptoms, bulky mediastinal or peripheral
lymphadenopathy, extranodal extension of disease, and advanced (stage IIIB- IV) disease. Bulky mediastinal lymphadenopathy is designated when the ratio of the maximum
measurement of mediastinal lymphadenopathy to intrathoracic cavity on an upright chest radiograph equals or exceeds 33%. Intermediate-risk features include localized
disease (stages I, II, and IIIA) with unfavorable features;
these cases may be treated similarly to advanced-stage
disease in some treatment protocols or treated with therapy
of intermediate intensity. Unfortunately, inconsistency in
risk categorization across cooperative group studies often
makes comparison of study outcomes challenging.
Radiation Therapy for Low-Risk HL
Considering the excellent survival rates achieved in children and adolescents with low-risk presentations of HL with
From the Department of Oncology, Division of Cancer Survivorship, St. Judes Childrens
Research Hospital, Memphis, TN; Departments of Radiation Oncology and Pediatrics,
Philip Rubin Center for Cancer Survivorship, James P. Wilmot Cancer Center at University
of Rochester Medical Center, Rochester, NY.
Address reprint requests to Melissa M. Hudson, MD, St. Jude Childrens Research
Hospital, Department of Oncology, Division of Cancer Survivorship, 262 Danny Thomas
Place, Mailstop 735, Memphis, TN 38105; email: melissa.hudson@stjude.org.
1092-9118/10/1-10
RADIOTHERAPY FOR PEDIATRIC HODGKIN LYMPHOMA

Table 1. Selected Risk-Adapted Treatment Approaches for Low-Risk Pediatric Hodgkin Lymphoma
Chemotherapy (No. Cycles)
25
VAMP (4)
COPP/ABV (4)11
OEPA/OPPA (2)8
DBVE (4)26
Radiation (Gy)
Stage
No. Patients
Event-Free Survival (years)
Survival (years)
1525.5, IF
21, IF/None
2035, IF/None
25.5, IF
CS I/II
CS IA/B, IIA*
I, IIA
IA, IIA, IIIA1
110
294
281/113
51
89 (10)
97/91 (3)
94/97 (5)
91 (6)
96 (10)
100/100 (3)
NA
98 (6)
Abbreviations: VAMP, vinblastine, doxorubicin, methotrexate, prednisone; ABV, doxorubicin, bleomycin, vinblastine; COPP, cyclophosphamide, vincristine,
procarbazine, prednisone; OEPA, vincristine etoposide, prednisone, doxorubicin; OPPA, vincristine, procarbazine, prednisone, doxorubicin; DBVE, doxorubicin,
bleomycin, vincristine, etoposide; E, extralymphatic; IF, involved-field radiation therapy.
Without bulky mediastinal (defined as one-third or more of intrathoracic ratio measured on an upright posteroanterior chest radiograph) or peripheral
lymphadenopathy (defined 6 cm or more) or B symptoms.
* Without adverse features defined as one or more of the following: hilar adenopathy, involvement of more than four nodal regions; mediastinal tumor with diameter
greater than or equal to one-third of the chest diameter, and node or nodal aggregate with a diameter greater than 10 cm.
chemotherapy, RT, and combined-modality therapy, clinical

trials have focused on reducing or eliminating agents and
modalities associated with a high risk for morbidity. Numerous studies have established the effectiveness of non-crossresistant chemotherapy alone in these patients, which offers
advantages for children managed in centers lacking diagnostic and therapeutic radiation facilities and avoids the
potential long-term growth inhibition, organ dysfunction,
and carcinogenesis associated with RT. However,
chemotherapy-alone treatment protocols typically prescribe
KEY POINTS
Pediatric Hodgkin lymphoma trials focus on maximizing treatment efficacy and minimizing risks for
late toxicity associated with both radiation therapy
and chemotherapy.
The use of low-dose involved-field radiation in pediatric Hodgkin lymphoma permits reduction in duration or intensity of chemotherapy and thus doserelated toxicity of anthracyclines, alkylating agents,
and bleomycin that may preserve cardiopulmonary
and gonadal function and reduce the risk of secondary leukemia.
Radiation has been used as an adjunct to multiagent
chemotherapy in clinical trials for intermediate/highrisk pediatric Hodgkin lymphoma with the goal of
reducing risk of relapse in initially involved sites and
preventing toxicity associated with retrieval therapy.
Compared with treatment with chemotherapy alone,
adjuvant radiation produces a superior event-free
survival for intermediate/high-risk children with
Hodgkin lymphoma who achieve a complete response
to multiagent chemotherapy, but does not affect overall survival because of the success of salvage therapy.
Radiation consolidation may facilitate local disease
control in individuals with refractory/recurrent disease, especially in those who have limited or bulky
sites of disease progression/recurrence, or persistent
disease that does not completely respond to chemotherapy.
Future directions in the use of radiation therapy
include reducing the targeted volume to include only
the initially involved nodes rather than the lymph
node regions that harbored those nodes; this may
further reduce radiation-associated toxicities.
higher cumulative doses of alkylating agents, anthracyclines, and bleomycin, which may produce late treatment
morbidity from cardiopulmonary and gonadal injury and
secondary leukemia. In early trials, high-dose RT provided
an alternative therapeutic option for skeletally mature patients that avoided MOPP chemotherapy-related infertility
and leukemogenesis.7 Later trials combined low-dose (15
25.5 Gy) IFRT with multiagent chemotherapy and sequentially reduced the number of chemotherapy cycles, especially
those including alkylating agents (Table 1).
Once the effectiveness of combined modality regimens
with fewer cycles of multiagent chemotherapy was established, contemporary studies employed a response-based
paradigm to guide further reduction of chemotherapy and
RT exposure. Regimens utilizing low-dose IFRT with chemotherapy combinations delivered at maximal dose intensity
such as OPPA (vincristine, procarbazine, prednisone, doxorubicin)8 and DBVE (doxorubicin, bleomycin, vincristine,
etoposide)9 produced excellent outcomes after treatment
with only two cycles of chemotherapy. A combined modality
approach using nonalkylator-based chemotherapy and
response-based low-dose IFRT also proved to be successful.
For example, consortium investigators from St. Jude, Stanford, and Dana Farber demonstrated that local control was
not compromised by reducing IFRT dose to 15 Gy in low-risk
patients who achieved an early complete response to VAMP
(vinblastine, doxorubicin, methotrexate, prednisone) chemotherapy.10
Other groups undertook clinical trials aiming to eliminate
RT for low-risk patients who achieved a complete response
to chemotherapy.8,11 A randomized controlled trial implemented by the Childrens Cancer Group (CCG) reported a
significantly (stratified log-rank test; p 0.057) higher
3-year event-free survival (EFS) in patients who received 21
Gy IFRT consolidation (97%; standard error [SE] 1.7%),
compared with those treated with four cycles of COPP/ABV
(cyclophosphamide, vincristine, procarbazine, prednisone/
doxorubicin, bleomycin, vinblastine) chemotherapy alone
(91%; SE 2.8%).11 German investigators subsequently observed no difference in disease-free survival among nonirradiated (97%; SE 2%) and irradiated (94%; SE 2%)
low-risk girls treated with OPPA (vincristine, prednisone,
procarbazine, doxorubicin) and boys treated with OEPA
(substitution of etoposide for procarbazine in the OPPA
combination).8 Likewise, the EFS for children and adolescents treated with four cycles of VAMP chemotherapy after
achieving an early complete response (89%; SE 5.7%) did
not differ from those treated with four cycles of VAMP and
response-based IFRT (87%; SE 6.4%) for those who did not
achieve an early complete remission.12 Common to all these
617
HUDSON AND CONSTINE

Table 2. Selected Risk-Adapted Treatment Approaches for Intermediate-Risk Pediatric Hodgkin Lymphoma
Chemotherapy (Number of Cycles)
11
COPP/ABV (6)
OEPA/OPPA (2) COPP (2)8
OEPA/OPPA (2) COPDAC (2)15
ABVE-PC (35)18
Radiation (Gy)
21, IF
2035, IF
2035, IF
21, IF
Stage
No. Patients
CS I/II*, CS IIB, CS III

IIEA, IIB, IIIA
IIEB, IIIEA/B, IIIB, IVA/B
IB, IIA, IIIA
394
212
139
53
84 (3)
92 (5)
88.3 (5)
84 (5)
Survival (years)
100 (3)
N/A
98.5 (5)
95 (5)
Abbreviations: COPP, cyclophosphamide, vincristine, procarbazine, prednisone; ABV, doxorubicin, bleomycin, vinblastine; OEPA, vincristine etoposide, prednisone,
doxorubicin; OPPA, vincristine, procarbazine, prednisone, doxorubicin; COPDAC, cyclophosphamide, vincristine, prednisone, dacarbazine; ABVE-PC, doxorubicin,
bleomycin, vincristine, etoposide-prednisone, cyclophosphamide; E, extralymphatic; IF, involved-field radiation therapy.
* With adverse disease features defined as one or more of the following: hilar adenopathy, involvement of more than four nodal regions; mediastinal tumor with
diameter greater than or equal to one-third of the chest diameter, and node or nodal aggregate with a diameter greater than 10 cm.
studies is that because of effective retrieval therapy among

relapsed patients, overall survival did not differ among the
irradiated and nonirradiated groups.
Radiation Therapy for Intermediate/High-Risk HL
Concerns regarding long-term chemotherapy and RTrelated toxicities have led to the development of divergent
therapeutic approaches for children and adolescents compared with that of adults. In contrast to adult trials, clinical
trials for intermediate/high-risk pediatric HL typically use
RT as an adjunct to multiagent chemotherapy, with the goal
of reducing risk of relapse and preventing toxicity associated
with retrieval therapy. Combination chemotherapy including vinca alkaloids, alkylating agents, anthracyclines, and
often etoposide provides the cornerstone of therapy (Tables 2
and 3). Gender-based regimens consider that male patients
are more vulnerable to gonadal toxicity from alkylating
agent chemotherapy and that female patients have a substantial risk of breast cancer after chest RT. In this regard,
German Multi-Center investigators have undertaken a series of gender-based risk-adapted trials aiming to reduce
gonadal toxicity in male patients while maintaining the
excellent disease-free survival accomplished with the OPPA/
COPP regimen. The DAL-HD-90 study established that
substitution of etoposide for procarbazine in the OPPA
combination (OEPA) in boys produces comparable EFS to
that of girls treated with OPPA and is associated with
hormonal parameters suggesting a lower risk of gonadal
toxicity.13,14 In the GPOH-HD 2002 trial, substitution of
dacarbazine for procarbazine (OEPA-COPDAC) in boys produced comparable results to standard OPPA-COPP in girls
when used in combination with IFRT.15 Long-term follow-up
is needed to determine if restriction of alkylating agent
cumulative dose translates into improved rates of fertility
preservation.
Risk-adapted multiagent chemotherapy for intermediate/
high-risk HL is typically followed by consolidative RT to
involved sites of disease. Effective systemic therapy coupled
with advancements in diagnostic imaging has led to increasingly restricted treatment fields that generally encompass
lymph node regions initially involved at the time of diagnosis; field refinement is then routinely made to account for
tumor regression with chemotherapy.16 IFRT is the most
common approach used in pediatric HL trials, although
some groups are now evaluating involved-nodal and limited
volume conformal (tailored field), intensity modulated, and
proton RT as approaches that further reduce potential
injury to normal tissues.
In the past decade, several trials have investigated the
benefit that adjuvant RT contributes to survival outcomes
among intermediate/high-risk children with HL who achieve
a complete response to multiagent chemotherapy. In the
CCGs randomized controlled trial using COPP/ABV hybrid
chemotherapy, the projected 3-year EFS among patients
who achieved a complete response to initial therapy, was
92% (SE 1.9%) for those randomized to receive low-dose
IFRT and 87% (SE 2.2%) for those randomized to receive
no further therapy.11 The difference in 3-year EFS was most
marked for stage IV patients randomized to receive
combined-modality therapy with IFRT (90%, SE 5.5%)
compared with those randomized to receive chemotherapy
alone (81%, SE 6.9%). Likewise, omission of RT for
patients completely responding to risk- and gender-based
OEPA/COPP or OPPA/COPP chemotherapy resulted in significantly lower EFS in intermediate/high-risk patients compared with irradiated patients (79% vs. 91%, p 0.0008).8
Notably, this study also demonstrated a survival benefit of
providing a 510 Gy RT boost to lymph node regions with
an insufficient remission following chemotherapy, thereby
overcoming the adverse prognostic implications of bulky
mediastinal lymphadenopathy.17 For both studies, estimates for overall survival did not differ between the irradiated and nonirradiated groups because of successful salvage
therapy after relapse.8,11
Contemporary trials have investigated if chemotherapy
and RT can be limited in patients who achieve a rapid early
Table 3. Selected Risk-Adapted Treatment Approaches for High-Risk Pediatric Hodgkin Lymphoma
Chemotherapy (Number of Cycles)
8
OEPA/OPPA (2) COPP (4)

OEPA/OPPA (2) COPDAC (4)15
ABVE-PC (35)18
BEACOPP (4); COPP/ABV (4) (RER; girls)19
BEACOPP (4); ABVD (2) (RER; boys)19
BEACOPP (8) (SER)19
Radiation (Gy)
2035, IF
2035, IF
21, IF
None
21, IF
21, IF
Stage

IB, IIA, IIIA
IIB, IIIB, IV
IIB, IIIB, IV
IIB, IIIB, IV
No. Patients
265
239
163
38
34
25
91 (5)
86.9 (5)
85 (5)
94 (5)
Survival (years)
N/A
94.9 (5)
95 (5)
97 (5)
Abbreviations: OEPA, vincristine etoposide, prednisone, doxorubicin; OPPA, vincristine, procarbazine, prednisone, doxorubicin; COPP, cyclophosphamide,
vincristine, procarbazine, prednisone; COPDAC, cyclophosphamide, vincristine, prednisone, dacarbazine; ABVE-PC, doxorubicin, bleomycin, vincristine, etoposideprednisone, cyclophosphamide; BEACOPP, bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, prednisone; ABV, doxorubicin, bleomycin,
vinblastine; ABVD, doxorubicin, bleomycin, vinblastine, dacarbazine; E, extralymphatic; IF, involved-field radiation therapy; RER, rapid early response; SER, slow early
response.
618
RADIOTHERAPY FOR PEDIATRIC HODGKIN LYMPHOMA
response to dose-intensive chemotherapy regimens. These

studies have been facilitated by assessment of interim or end
of chemotherapy response based on anatomic or functional
changes on computed tomography or functional imaging like
positron emission tomography. The Pediatric Oncology
Group utilized a response-based therapy utilizing dosedense ABVE-PC (doxorubicin, bleomycin, vincristine,
etoposide-prednisone, cyclophosphamide) for patients with
unfavorable advanced-stage disease in combination with 21
Gy IFRT.18 The dose-dense approach permitted reduction in
chemotherapy exposure in 63% of patients who achieved a
rapid early response to three ABVE-PC cycles. Five-year
EFS was comparable for rapid early responders (86%) and
slow early responders (83%) treated with three and five
cycles of ABVE-PC, respectively, followed by 21 Gy IFRT.
The CCG (CCG-59704) evaluated response-adapted therapy featuring four cycles of the dose-intensive BEACOPP
(bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, prednisone) regimen followed by a
gender-tailored consolidation for pediatric patients with
stage IIB, IIIB with bulk disease, and IV HL.19 For rapid
early responding girls, an additional four courses of COPP/
ABV (without IFRT) were given in an effort to reduce breast
cancer risk. Rapid early responding boys received two cycles
of ABVD followed by IFRT. Slow early responders received
four additional courses of BEACOPP and IFRT. Rapid early
response (defined by resolution of B symptoms and greater
than 70% reduction in tumor volume) was achieved by 74%
of patients after four BEACOPP cycles and 5-year EFS
among the cohort is 94% (median follow-up 6.3 years).19
Results from this study support that early intensification
followed by less intense response-based therapy results in
high EFS. However, the potential for late treatment effects
on cardiopulmonary and gonadal function have not been
evaluated.
Radiation Therapy in Refractory/Relapsed HL
The generally excellent outcome of children and adolescents with HL limits opportunities to evaluate retrieval
therapy. This is particularly true in regard to the benefits of
using RT in the setting of refractory/relapsed disease. Uniformly, chemotherapy is the recommended retrieval therapy, with the choice of specific agents, dose-intensity, and
number of cycles determined by the initial therapy, disease
characteristics at progression/relapse, and response to retrieval therapy. In children with localized favorable (relapse
after 12 months after completing therapy) disease recurrences whose original therapy involved reduced cycles of
risk-adapted therapy, IRFT consolidation may be offered
following treatment with more intensive conventional chemotherapy. Autologous hematopoietic cell transplantation
(autoHCT) is the recommended approach for patients who
develop refractory/relapsed disease during or within 1 year
after completing therapy.20 Results of investigations
(largely comprised of adult patients) evaluating the use of
IFRT immediately before or after transplant have been
conflicting in their support of a potential benefit conferred by
RT.21-23 However, most studies are limited by their retrospective nature, nonrandom treatment assignments, and
small patient numbers. Nevertheless, IFRT is often used to
consolidate local control in individuals with refractory/recurrent disease, especially in those who have limited or bulky
sites of disease recurrence or persistent disease that does
not completely respond to salvage chemotherapy. In the

latter, local radiotherapy can be considered to promote local
disease resolution before autoHCT rather than as consolidation following autoHCT. However, concerns regarding toxicity have prevented the use of IFRT as a standard adjuvant to
high-dose chemotherapy and autoHCT in HL. Prospective
trials are needed to definitely establish the long-term risks
and survival benefits provided by inclusion of RT in salvage
therapy approaches.
Advances in the Delivery of Radiation
As previously stated, contemporary protocols are testing

field reductions from involved lymph node regions to involved nodal sites. In addition, the standard technique for
radiation delivery in pediatric HL is a two-dimensional
CT-based approach. However, three-dimensional conformal
radiation therapy (3DCRT), intensity-modulated RT
(IMRT), or proton therapy may be offer the ability to reduce
exposure to critical normal tissues such as heart, lungs, and
developing breast tissue. Uncertainty exists about the potential for increased late effects from IMRT, particularly
secondary malignancy, bacause IMRT results in a lower
dose to a larger volume compared with conventional techniques. Unfortunately, any benefits of lower doses to critical
organs, or changes (either an increase or decrease) in the
potential for a subsequent malignancy, will not be testable
for many years.
Conclusion
The role of RT in the treatment of pediatric HL has

evolved considerably over the last 50 years, primarily motivated by the desire to avoid its adverse long-term effects.
Although elimination of RT has been a key focus of contemporary trials, this modality continues to play an important
role in optimizing survival outcomes for many children and
adolescents with the disease. For those with low-risk HL,
the addition of low-dose IFRT permits reduction in chemotherapy duration or intensity and thus the potential doserelated toxicity of anthracyclines, alkylating agents, and
bleomycin that may preserve cardiopulmonary and gonadal
function. For those with intermediate/high-risk HL, the
superior disease-free survival demonstrated by some trials
may ultimately translate into a survival benefit by avoiding
the need for more toxic salvage therapy. In the setting of
disease with suboptimal chemotherapy response, RT consolidation provides an effective alternative modality to enhance
tumor control. Until the availability of validated biologically
based prognostic factors, the inclusion of RT consolidation in
pediatric HL treatment regimens will continued to be guided
by prognostic factors at diagnosis correlated with disease
burden and chemotherapy responsiveness as assessed by
functional imaging. For patients who require RT to optimize
disease control, advances in radiation technology and reduced volume radiation strategies are anticipated to greatly
reduce exposure to normal tissues and the subsequent risk
of late RT-associated toxicity.24
Acknowledgements
Research grant support: Dr. Hudson is supported in part by
the Cancer Center Support (CORE) grant CA 21765 from the
National Cancer Institute and by the American Lebanese Syrian Associated Charities (ALSAC).
619
HUDSON AND CONSTINE
Author
Melissa M. Hudson*
Louis S. Constine*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Donaldson SS, Kaplan HS. Complications of treatment of Hodgkins
disease in children. Cancer Treat Rep. 1982;66:977-989.
2. Donaldson SS, Link MP. Combined modality treatment with low-dose
radiation and MOPP chemotherapy for children with Hodgkins disease.
J Clin Oncol. 1987;5:742-749.
3. Bhatia S, Robison LL, Oberlin O, et al. Breast cancer and other second
neoplasms after childhood Hodgkins disease. N Engl J Med. 1996;334:745751.
4. Hancock SL, Tucker MA, Hoppe RT. Factors affecting late mortality
from heart disease after treatment of Hodgkins disease. JAMA. 1993;270:
1949-1955.
5. Devita VT, Jr., Serpick AA, Carbone PP. Combination chemotherapy in
the treatment of advanced Hodgkins disease. Ann Intern Med. 1970;73:881895.
6. Bonadonna G, Santoro A. ABVD chemotherapy in the treatment of
Hodgkins disease. Cancer Treat Rev. 1982;9:21-35.
7. Donaldson SS, Whitaker SJ, Plowman PN, et al. Stage I-II pediatric
Hodgkins disease: Long-term follow-up demonstrates equivalent survival
rates following different management schemes. J Clin Oncol. 1990;8:11281137.
8. Dorffel W, Luders H, Ruhl U, et al. Preliminary results of the multicenter trial GPOH-HD 95 for the treatment of Hodgkins disease in children
and adolescents: Analysis and outlook. Klin Padiatr. 2003;215:139-145.
9. Schwartz CL. Special issues in pediatric Hodgkins disease. Eur J
Haematol. Suppl 2005;55-62.
10. Donaldson SS, Hudson MM, Lamborn KR, et al. VAMP and low-dose,
involved-field radiation for children and adolescents with favorable, earlystage Hodgkins disease: Results of a prospective clinical trial. J Clin Oncol.
2002;20:3081-3087.
11. Nachman JB, Sposto R, Herzog P, et al. Randomized comparison of
low-dose involved-field radiotherapy and no radiotherapy for children with
Hodgkins disease who achieve a complete response to chemotherapy. J Clin
Oncol. 2002;20:3765-3771.
12. Metzger ML, Weinstein HJ, Hudson MM, et al. Results of a prospective
clinical trial of VAMP alone without irradiation for pediatric favorable,
early-stage Hodgkin lymphoma patients who achieve an early complete
response. J Clin Oncol Proc Am Soc Clin Oncol. 2011;29:585s.
13. Gerres L, Bramswig JH, Schlegel W, et al. The effects of etoposide on
testicular function in boys treated for Hodgkins disease. Cancer. 1998;83:
2217-2222.
14. Schellong G, Potter R, Bramswig J, et al. High cure rates and reduced
long-term toxicity in pediatric Hodgkins disease: The German-Austrian
multicenter trial DAL-HD-90. The German-Austrian Pediatric Hodgkins
Disease Study Group. J Clin Oncol. 1999;17:3736-3744.
620
15. Mauz-Korholz C, Hasenclever D, Dorffel W, et al. Procarbazine-free

OEPA-COPDAC chemotherapy in boys and standard OPPA-COPP in girls
have comparable effectiveness in pediatric Hodgkins lymphoma: The GPOHHD-2002 study. J Clin Oncol. 2010;28:3680-3686.
16. Yahalom J, Mauch P. The involved field is back: Issues in delineating
the radiation field in Hodgkins disease. Ann Oncol. 2002;13 Suppl 1:79-83.
17. Ruhl U, Albrecht M, Dieckmann K, et al. Response-adapted radiotherapy in the treatment of pediatric Hodgkins disease: An interim report at 5
years of the German GPOH-HD 95 trial. Int J Radiat Oncol Biol Phys.
2001;51:1209-1218.
18. Schwartz CL, Constine LS, Villaluna D, et al. A risk-adapted, responsebased approach using ABVE-PC for children and adolescents with
intermediate- and high-risk Hodgkin lymphoma: The results of P9425. Blood.
2009;114:2051-2059.
19. Kelly KM, Sposto R, Hutchinson R, et al. BEACOPP chemotherapy is a
highly effective regimen in children and adolescents with high-risk Hodgkin
lymphoma: A report from the Childrens Oncology Group. Blood. 2011;117:
2596-2603.
20. Colpo A, Hochberg E, Chen YB. Current Status of Autologous Stem Cell
Transplantation in Relapsed and Refractory Hodgkins Lymphoma. Oncologist. 2012;17:80-90.
21. Majhail NS, Bajorunaite R, Lazarus HM, et al. Long-term survival and
late relapse in 2-year survivors of autologous haematopoietic cell transplantation for Hodgkin and non-Hodgkin lymphoma. Br J Haematol. 2009;147:
129-139.
22. Moskowitz AJ, Perales MA, Kewalramani T, et al. Outcomes for
patients who fail high dose chemoradiotherapy and autologous stem cell
rescue for relapsed and primary refractory Hodgkin lymphoma. Br J Haematol. 2009;146:158-163.
23. Wendland MM, Asch JD, Pulsipher MA, et al. The impact of involved
field radiation therapy for patients receiving high-dose chemotherapy followed by hematopoietic progenitor cell transplant for the treatment of
relapsed or refractory Hodgkin disease. Am J Clin Oncol. 2006;29:189-195.
24. Koh ES, Tran TH, Heydarian M, et al. A comparison of mantle versus
involved-field radiotherapy for Hodgkins lymphoma: Reduction in normal
tissue dose and second cancer risk. Radiat Oncol. 2007;2:13.
25. Donaldson SS, Link MP, Weinstein HJ, et al. Final results of a
prospective clinical trial with VAMP and low-dose involved-field radiation for
children with low-risk Hodgkins disease. J Clin Oncol. 2007;25:332-337.
26. Tebbi CK, Mendenhall N, London WB, et al. Treatment of stage I, IIA,
IIIA1 pediatric Hodgkin disease with doxorubicin, bleomycin, vincristine and
etoposide (DBVE) and radiation: A Pediatric Oncology Group (POG) study.
Pediatr Blood Cancer. 2006;46:198-202.
Role of Doxorubicin in Rhabdomyosarcoma:

Is the Answer Knowable?
By Carola A. S. Arndt, MD
Overview: The role of doxorubicin in treatment of rhabdomyosarcoma (RMS) has been controversial for 30 years. Despite its known activity in RMS, because of its risk of
cardiotoxicity, its use is not justified in low-risk patients who
have an excellent chance of cure with vincristine, actinomycin
with or without cyclophosphamide, and primary tumor treatment. For patients with intermediate and high risks, the
risk/benefit ratio must be carefully considered. In addition, the
peak incidence of RMS is in toddlers, with whom the risk of
HE ROLE of doxorubicin in the treatment of rhabdomyosarcoma has been controversial for over three decades. The first report of its activity in rhabdomyosarcoma
was by Massimo and colleagues in 1969,1 followed by an
early phase II study in children by Tan, which showed that
doxorubicin was active against both newly diagnosed and
previously treated rhabdomyosarcoma.2
However, one of the major disadvantages of doxorubicin
is its cardiotoxicity, which was recognized early in its use.3
Risk factors for cardiotoxicity include young age at
administration, higher cumulative dose, and radiation fields
including the heart. With the peak incidence of rhabdomyosarcoma being in toddlers, and the majority of patients being
under age 10 at diagnosis, cardiotoxicity is a factor in the
risk/benefit calculation. Low-risk patients (low stage and
clinical group, favorable histology) have excellent outcomes
with survival rates of 80% to 90% with VAC (vincristine,
actinomycin, cyclophosphamide) or VA (vincristine, actinomycin) and primary tumor treatment alone; in them, the use
of doxorubicin with its potential for long-term cardiotoxicity
cannot be justified.
The goal of this discussion is to review the historic data
on use of doxorubicin in rhabdomyosarcoma, as well as to
discuss ongoing clinical studies utilizing it.
Historical Perspectives
North American Experience. Patients with group III and

group IV rhabdomyosarcoma were randomly assigned in Intergroup Rhabdomyosarcoma Study (IRS) to receive therapy
with VAC and radiation, with or without doxorubicin. The
dose of doxorubicin was 60 mg/m2, given at 10 to 12 week
intervals (alternating with actinomycin) either alone or with
continuous daily oral cyclophosphamide. There was no difference in outcome between regimens and the conclusion was
that doxorubicin did not improve the outcome of patients
with group III or IV disease.4 It should be noted that the
intervals between treatment cycles were much longer in the
early RMS studies than treatment intervals are in the current
era, and much of the cyclophosphamide was given orally.
IRS II, conducted between 1978 and 1984, again evaluated
the role of doxorubicin in patients with group III and IV
rhabdomyosarcoma, but this time given along with vincristine and intravenous cyclophosphamide as pulsed therapy.
Patients were randomly assigned between repetitive pulse
VAC or repetitive pulse VAC/VDC (D doxorubicin).5 Once
again, there was no difference in outcomes between the two
regimens.
cardiotoxicity of anthracyclines is higher. A number of trials

both in North America and Europe, which are reviewed in this
article, have investigated the role of doxorubicin in RMS, with
no conclusive outcomes. In addition, differences in riskgroup assignment on two sides of the Atlantic further complicate comparisons and analyses. The current European EpSSG
2005 study for high-risk RMS (by the European definition) may
come closest to giving an answer to the role of doxorubicin in
RMS.
IRS III, conducted between 1984 and 1991, had a very

complex study design.5 For patients with group II tumors,
the major objective was to see whether the addition of
doxorubicin to VA and radiation therapy (RT) improved the
outcome of favorable histology tumors. Patients were randomly assigned to treatment with VA or VA plus doxorubicin
(with RT in both cases). The outcome was better for the
patients treated with VA plus doxorubicin (89% vs. 54%
5 years survival, p 0.03). However, when historic controls
from IRS II treated with VA and RT were included, the
statistical significance disappeared. Moreover, the VA regimen had a worse outcome on IRS III than the identical
therapy on IRS II, further confounding the results. Nevertheless, the paper stated, In this randomized comparison,
there was suggestive statistical evidence that the addition of
doxorubicin (30 mg/m2/d IV X 2 during weeks 3, 6, 12, 15, 21,
and 24) to a basic VA regimen improved clinical outcome.
Patients with group III and group IV tumors were treated
with doxorubicin containing regimens that also included
other agents in addition to VAC, making the individual
contribution of doxorubicin impossible to determine. Patients with clinical group I and II unfavorable histology
tumors received pulsed VDC and VAC plus cisplatin, and
showed significant improvement in 5-year progression free
survival and survival rates compared with similar patients
treated less intensively on IRS II (71% 6% and 80% 6%
versus 59% 5% and 71% 5%, respectively; p 0.002 and
0.01, respectively). However, the regimens were quite complex, making the individual contribution of doxorubicin
impossible to determine.
In a series of phase II window studies for high-risk
patients with metastatic disease conducted by the Childrens Oncology Group (COG) between 1988 and 2000,
the combination of doxorubicin/ifosfamide and etoposide/
ifosfamide had the highest response rates, although there
was no survival difference.6 So once again, the role of
doxorubicin was not able to be determined conclusively.
A pilot study utilizing VDC alternating with ifosfamide/
From the Department of Pediatric and Adolescent Medicine, Mayo Clinic Childrens
Center, Rochester, MN.
Address reprint requests to Carola A. S. Arndt, MD, Mayo Clinic, Mayo Clinic Childrens
Center, Department of Pediatric and Adolescent Medicine, 200 First Street SW, Rochester,
MN; email: carndt@mayo.edu.
1092-9118/10/1-10
621
CAROLA A. S. ARNDT
etoposide for intermediate risk rhabdomyosarcoma reported

excellent results,7 but when compared to similar patients
treated on IRS IV, a difference in outcome could not be
demonstrated.8 The study was a small pilot study, which
included other agents (ifosfamide and etoposide in addition
to doxorubicin).
European Experience
European studies have incorporated anthracyclines for

rhabdomyosarcoma for many years, using doxorubicin or
epirubicin. The German studies CWS-81, 86, and 91 utilized
doxorubicin in combination with other drugs (ifosfamide,
cyclophosphamide, etoposide in regimens VACA, VAIA, and
EVAIA) but again there were no randomizations in the use
of doxorubicin.9-11 The Malignant Mesenchymal Tumor
(MMT) 84 study used doxorubicin only for patients with poor
response to upfront chemotherapy, again without randomization.12 In the MMT 95 study, high-risk nonmetastatic
patients were randomly assigned to receive IVA (ifosfamide,
vincristine, actinomycin) or IVA plus carboplatin, epirubicin, and etoposideand found no difference in outcome.13
KEY POINTS
The role of doxorubicin in treatment of rhabdomyosarcoma (RMS) remains controversial.

The use of doxorubicin in patients with low-risk RMS
is not justified.
Several studies have failed to show benefit in the use
of anthracycline, but the studies were complex, confounding the issue.
Although differences in risk group assignment between Europe and North America make comparisons
difficult, the current EpSSG 2005 study may shed
light on the role of anthracycline in high-risk (by
European definition) RMS.
EpSSG will study the addition of doxorubicin to
ifosfamide, actinomycin, vincristine, not a substitution study of doxorubicin for actinomycin.
More recently, a phase II window study of single-agent

doxorubicin in patients with high risk metastatic disease
showed a 65% overall response rate after two courses of
60 mg/m2 at 21 day intervals.14 The current high-risk
European pediatric soft tissue sarcoma study (EpSSG 2005)
is randomly assigning patients with high-risk RMS to IVA
versus IVA with doxorubicin. High-risk patients will also be
randomly assigned to discontinue therapy or continue with
maintenance chemotherapy, which may complicate the analysis. This study will be the purest in evaluating the addition
of doxorubicin (as opposed to the substitution of doxorubicin
for actinomycin) and may come the closest to addressing the
value of doxorubicin in patients with high-risk RMS. Adding
doxorubicin to IVA will likely make that arm of the study
more toxic. Significant differences in patient classification
may make trans-Atlantic comparisons difficult, however.15
Conclusion
Despite over 30 years of clinical trials in rhabdomyosarcoma on both sides of the Atlantic, we have yet to define the
appropriate use of doxorubicin in RMS. Clearly, it is not
appropriate to investigate further the use of this cardiotoxic
drug in patients with low-risk disease who have an outstanding prognosis. What about the intermediate-risk patients (by North American definition)? Currently, COG has
chosen other new agents to investigate, such as topoisomerase inhibitors, and in the future, multityrosine kinase
inhibitors. Is it justifiable to expose toddlers with
intermediate-risk RMS to a cardiotoxic agent that has never
been proven to improve outcome when there are newer,
perhaps more interesting drugs available? At this time, most
(but not all) investigators would say no. What about patients
with node-positive alveolar histology or other patients at
high risk of recurrence? They are considered high risk by
the European definition. The outcome is poor for these
patients, who are the subjects of the IVA versus IVADo
study in EpSSg 2005. We hope that this study will provide
some clarity for us in a subset of patients, as long as it is not
confounded by the randomization between maintenance
chemotherapy and stopping chemotherapy, or differences in
toxicity by the addition of doxorubicin to IVA, which may
compromise dose intensity. For patients with metastatic
disease, adding doxorubicin to regimens that incorporate
other agents has also not improved outcome.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Carola A. S. Arndt*
REFERENCES
1. Massimo L, Cottafava F, Mori PG, et al. [Preliminary clinical trial of
adriamycin in solid malignant tumors in infants]. Minerva Pediatr. 1969;21:
2182-2186.
2. Tan C, Etcubanas E, Wollner N, et al. Adriamycinan antitumor
antibiotic in the treatment of neoplastic diseases. Cancer. 1973;32:9-17.
3. Gilladoga AC, Manuel C, Tan CT, et al. The cardiotoxicity
of adriamycin and daunomycin in children. Cancer. 1976;37:10701078.
4. Maurer HM, Beltangady M, Gehan EA, et al. The Intergroup Rhabdomyosarcoma Study-I. A final report. Cancer. 1988;61:209-220.
622
5. Maurer HM, Gehan EA, Beltangady M, et al. The Intergroup Rhabdomyosarcoma Study-II. Cancer. 1993;71:1904-1922.
6. Lager JJ, Lyden ER, Anderson JR, et al. Pooled analysis of phase II
window studies in children with contemporary high-risk metastatic rhabdomyosarcoma: a report from the Soft Tissue Sarcoma Committee of the
Childrens Oncology Group. J Clin Oncol. 2006;24:3415-3422.
7. Arndt CA, Nascimento AG, Schroeder G, et al. Treatment of intermediate risk rhabdomyosarcoma and undifferentiated sarcoma with alternating cycles of vincristine/doxorubicin/cyclophosphamide and etoposide/
ifosfamide. Eur J Cancer. 1998;34:1224-1229.
DOXORUBICIN IN RHABDOMYOSARCOMA
8. Arndt CA, Hawkins DS, Meyer WH, et al. Comparison of results of a
pilot study of alternating vincristine/doxorubicin/cyclophosphamide and etoposide/ifosfamide with IRS-IV in intermediate risk rhabdomyosarcoma: A report
from the Childrens Oncology Group. Pediatr Blood Cancer. 2008;50:33-36.
9. Dantonello TM, Int-Veen C, Harms D, et al. Cooperative trial CWS-91 for
localized soft tissue sarcoma in children, adolescents, and young adults. J Clin
Oncol. 2009;27:1446-1455.
10. Koscielniak E, Jurgens H, Winkler K, et al. Treatment of soft tissue
sarcoma in childhood and adolescence. A report of the German Cooperative Soft
Tissue Sarcoma Study. Cancer. 1992;70:2557-2567.
11. Koscielniak E, Harms D, Henze G, et al. Results of treatment for soft
tissue sarcoma in childhood and adolescence: a final report of the German Cooperative Soft Tissue Sarcoma Study CWS-86. J Clin Oncol. 1999;17:3706-3719.
12. Flamant F, Rodary C, Rey A, et al. Treatment of non-metastatic rhabdomyosarcomas in childhood and adolescence. Results of the second study of the
International Society of Paediatric Oncology: MMT84. Eur J Cancer. 1998;34:
1050-1062.
13. Stevens M, Rey A, et al. SIOP MMT 95: Intensified (6 drug) versus
standard (IVA) chemotherapy for high risk non metastatic rhabdomyosarcoma
(RMS). J Clin Oncol. 2004;22:14s (suppl; abstr 8515).
14. Bergeron C, Thiesse P, Rey A, et al. Revisiting the role of doxorubicin in
the treatment of rhabdomyosarcoma: an up-front window study in newly
diagnosed children with high-risk metastatic disease. Eur J Cancer. 2008;44:
427-431.
15. Sultan I, Ferrari A. Selecting multimodal therapy for rhabdomyosarcoma.
Expert Rev Anticancer Ther. 2010;10:1285-1301.
623
PONTINE GLIOMAS IN CHILDREN:

TO BIOPSY OR NOT TO BIOPSY
CHAIR
Mark W. Kieran, MD, PhD
Boston, MA
SPEAKERS
Stephanie Puget, MD, PhD
Necker Hospital, Universite Paris-Descartes
Paris, France
Nicholas K. Foreman, MD
The Childrens Hospital
Aurora, CO
I d e n t i fi c a t i o n o f N o v e l B i o l o g i c T a r g e t s i n t h e
Treatment of Newly Diagnosed Diffuse
Intrinsic Pontine Glioma
By Nathan J. Robison, MD, and Mark W. Kieran, MD, PhD
Overview: Diffuse intrinsic pontine gliomas (DIPGs) carry an

extremely poor prognosis. Standard practice has been to base
the diagnosis on classic imaging and clinical characteristics
and to treat with focal radiation therapy, usually accompanied
with experimental therapy. As a result of the desire to avoid
upfront biopsy, little has been learned regarding the molecular
features of this disease. Findings from several autopsy series
have included loss of p53 and PTEN, and amplification of
PDGFR. Based on these and other findings, murine models

have been generated and provide a new tool for preclinical
testing. DIPG biopsy at diagnosis has increasingly become
incorporated into national protocols at several centers, bringing the prospect of a better understanding of DIPG biology in
the future. Initial analyses of pretreatment tumors cast valuable new light and establish the importance of p53 inactivation
and the RTK-PI3K pathway in this disease.
chromosome 10 in four cases.8 Similarly, p53 mutations

occurring in 8 of 13 (61%) high-grade pontine glioma specimens obtained at autopsy have also been reported.9 An
examination of 28 malignant brainstem gliomas, including
18 pretreatment and 10 postmortem specimens identified
TP53 mutation in six tumors (21%) while EGFR amplification was seen in three (11%). EGFRvIII mutations were not
seen, and no correlation was seen between TP53 and EGFR
abnormalities.10
IFFUSE INTRINSIC pontine glioma is a disease of

childhood that carries an abysmal prognosis. Marked
advances in oncology over the last five decades have not been
paralleled in this disease. DIPG is now the main cause of
brain tumor death in children.1 The standard of care for
DIPG remains focal radiation therapy alone, which alleviates symptoms in over 75% of patients; however rapid
disease progression is almost universal.2 Median overall
survival is less than 1 year, and the 2-year survival rate
is less than 10%.1,3 Concerted experimental clinical trials
over several decades have yielded no improvement in
progression-free or overall survival.4
DIPG biology is not well understood. In the magnetic
resonance imaging era, DIPG has become a radiologic diagnosis with no clear indication to biopsy, except in rare cases
with atypical features. Paucity of upfront tissue available for
analysis has been one of the main reasons that knowledge
of this disease has failed to keep pace with that of other
cancers. In recent years, an exponential increase in the
technology for molecular profiling has begun to fundamentally change approaches to cancer treatment. There is a
growing awareness that traditional diagnostic classifications based on microscopic appearance of the tumor must be
reconsidered in light of our greater understanding of cancer
biology and the relevant signal transduction pathways that
lead to new therapeutic options. Tumors previously thought
to represent a single diagnosis have been found, by molecular characterization, to in fact represent multiple distinct
diseases. Conversely, recurring mutations such as BRAF
and EGFR have been identified across seemingly unrelated
cancers. With the advent of molecularly targeted therapy,
these features take on new therapeutic importance. Identification of targetable oncogenic molecules has become a
major focus of translational oncology research. However,
identification of drugable targets in DIPG has lagged significantly behind other cancers. Initial attempts to use targeted modalities in DIPG treatment, relying by necessity on
targets identified in adult glioblastoma rather than those
specifically identified in DIPG, have shown no efficacy.5-7
Clearly, a better understanding of DIPG biology is needed.
Postmortem Findings
Several more recent studies have evaluated series of DIPG

samples collected at autopsy.11-13 Tissue samples collected
at autopsy have been shown to yield DNA and RNA of
quality sufficient for genomic analysis in the majority of
cases.16 However, autopsy findings must be interpreted with
caution. Significant alterations may occur in tumor biology
between the time of initial presentation and post-treatment
recurrence or progression.13 Radiation and chemotherapy
can select for new genomic aberrations at the time of
recurrence and loss of targets present at diagnosis.17 Discovered targets from autopsy specimens, though arguably
relevant to the treatment of progressive advanced-stage
disease, may be less so to new-onset disease. Some of the
studies described here do include a small number of upfront
biopsy specimens,13 although these were cases in which
atypical features were present and were the reason these
tumors underwent biopsy. The extent to which these molecular findings are representative of the biology of classic
newly diagnosed DIPG is therefore uncertain.
A genome-wide analysis of 11 DIPGs, including 9 postmortem and 2 pretreatment samples, were analyzed by hybridized single-nucleotide polymorphism (SNP) array, with PCR
confirmation.11 PDGFR amplification was seen in four tumors (36%). Chromosomes commonly showing large areas of
gain included 1q, 9q, 17q, and 10p. Loss of chromosomes
11p, 17p, 14q, 18p, and 22q was also common. TP53, on
chromosome 17p, demonstrated hemizygous deletion in
seven (64%) of 11 DIPGs. PTEN deletion was seen in three
Initial Molecular Studies of DIPG
In one of the earliest molecular genetic studies of DIPG,

gene sequencing showed p53 mutation in five of seven DIPG
(71%) samples analyzed. Portions of 17p that included the
p53 gene were lost in four; none showed EGFR amplification. The same small series noted loss of the long arm of
From the Dana-Farber Childrens Hospital Cancer Center, Boston, MA.

Address reprint requests to Mark W. Kieran, MD, PhD, Pediatric Medical NeuroOncology, Dana-Farber Childrens Hospital Cancer Center, 450 Brookline Avenue, Rm
SW331, Boston, MA 02215; email: mark_kieran@dfci.harvard.edu.
1092-9118/10/1-10
625
ROBISON AND KIERAN
(27%), including one of two pretreatment specimens. Loss of

heterozygosity was seen of genes involved in DNA repair
pathways, including RPA1 and MYH1 on 17p, MNAT1 and
RAD51L1 on 14q, and MSH4 on 1p, among others. Low-level
amplification of PARP-1, involved in DNA repair, was seen
in three (27%) cases.
DIPG specimens obtained at autopsy from 11 patients
showed gain of MDM4, an inhibitor of p53 activity, in seven
cases (64%). Immunohistochemistry for p53 was positive in
four (36%). Abnormalities of genes involved in the RTKPI3K pathway were common, including amplification of
EGFR in three (27%), PDGFRA in two (18%), and IRS2 in
one, and loss of PTEN (associated with loss of chromosome
10q) in six (54%). Gains of 1q, 7q, and 7p were common.12
A larger recent study similarly revealed focal amplifications in the PI3K pathway, most commonly involving PDGFRA (30%) and MET (26%), in 20 of 43 (47%) DIPGs.13
Thirty percent of tumors had focal amplifications of genes
regulating retinoblastoma (RB) protein phosphorylation,
most commonly CDK4, CDK6, and CCND2. Notably, 21%
had concurrent amplification of both PI3K and RB signaling
pathways. In one case, for which autopsy specimen both
from primary tumor and from an area of cerebellar extension were analyzed, significant heterogeneity was observed
in copy number alterations within the same tumor. It is
unclear whether this same heterogeneity would be seen in a
tumor prior to treatment. Intratumoral biologic heterogeneity is one of the limitations of a personalized medicine
approach. On the other hand, this may simply highlight the
limitations of molecular data from post-treatment autopsy
specimens.
Very recently, whole-genome sequencing of DNA from
seven DIPGs with matched normal tissue showed mutations
in histone H3, isoforms H3.3 or H3.1, resulting in substitution to methionine at lysine 27, in five of seven patients.14 In
a validation cohort, similar mutations were found in 78% of
DIPGs and 22% of non-brainstem pediatric glioblastomas.
Similar mutations have not been seen in adult glioblastoma.
It is hypothesized that the mutation results in a gain-offunction phenotype, affecting epigenetic gene expression
regulation. Of note, H3F3A, the gene encoding the more
KEY POINTS
626
DIPG diagnosis is based on radiographic and clinical

criteria rather than biopsy because of concerns related to the morbidity of biopsy in the pons.
Current treatment approaches include focal radiation
therapy and experimental agents based on pathways
identified in adult glioblastoma multiforme.
No progress has been made in the event-free or
overall survival of DIPG over the last four decades.
Advances in neurosurgical techniques and molecular
profiling of small samples have resulted in the development of national protocols that include pretherapy
biopsy and molecular profiling of DIPGs.
Specific targets in pretherapy DIPGs have been identified, and in conjunction with new preclinical models
of DIPG, can now be tested with direct translation to
the clinic.
commonly affected H3.3 isoform, is located on chromosome

1q, a region of large-scale chromosomal gain in more than
20% of DIPGs, as well as of other pediatric high-grade
gliomas.
An important area of target discovery is the identification
of tumor-specific cell surface markers, which may not drive
malignant behavior, but which can be used to localize
delivery of cytotoxic agents or induce an immune response.
In one such study, increased staining of interleukin-13
receptor alpha2 chain (IL-13Ralpha2), which is not expressed at significant levels in normal brain, was seen in 17
of 28 (61%) brainstem gliomas.15 Increased expression of
IL-13Ralpha2 is also seen in adult high-grade gliomas.
Pretreatment Biopsy Analysis Series
Developments in neurosurgical technique, as well as developments in tissue processing and extraction techniques,
which have decreased the amount of tissue necessary for
analysis, have significantly decreased the potential risks of
DIPG biopsy while increasing the information obtained.
Experience from several European countries where biopsies
are performed as part of a formal clinical trial has shown
biopsy of DIPG to be safe and feasible.18 This has important
implications for our prospects of gaining a better understanding of this disease.
In a high-grade glioma study that included seven pretreatment DIPGs, high-resolution analysis of genomic imbalances showed PDGR amplification in two tumors (29%).19 In
another small study of formalin-fixed paraffin-embedded
tissue from 13 DIPGs, 10 of which were pretreatment
samples, high-resolution 244 K oligo array comparative
genomic hybridization identified PDGFRA amplification
(4q1113), confirmed by qPCR, in 2 DIPGs (15%). One tumor
had amplification of the cell cycle regulator cyclin D1
(11q13). CDKN2A/CDKN2B deletions were notably lacking,
as was EGFR amplification. One of 12 scorable tumors
showed MYCN amplification. 1q gain, which is seen a
diverse array of both the central nervous system and other
pediatric malignancies, was seen in 3 tumors (23%); in 2,
this was associated with 1p loss. Loss of 17p, the site of the
well-characterized tumor suppressor gene p53, was seen in
four tumors (31%); loss of 14q was likewise seen in four.17
Included on each of these chromosomal arms (1p, 17p, 14q)
are a number of genes involved in DNA repair pathways,
previously noted to be lost in DIPG.11 Loss of 10q, which is
the most common aberration in adult GBM, and relatively
common in pediatric high-grade glioma (HGG), was only
seen in one DIPG. One tumor had a balanced genome, with
no aberrations identified by aCGH.17
Very recently, we have reported on a mutational analysis
of 20 classic DIPG tumors all biopsied at diagnosis.20 Using
OncoMap, a mass spectrometric method of allele detection
that analyzes for the presence of 983 different mutations in
115 oncogenes, nine of 20 (45%) tumors were found to have
no identifiable mutations. In keeping with findings of earlier
studies, eight (40%) tumors were found to harbor TP53
mutations. As OncoMap assays for only the seven most
frequent TP53 mutations, this may underestimate the TP53
mutation prevalence. Interestingly, five of eight identified
TP53 mutations were at the 273 arginine locus. PI3KCA
mutations were seen in three tumors, which is the first
report of a mutated oncogene in newly diagnosed DIPG
samples. Additional PI3K pathway-related abnormalities
NOVEL TARGETS IN THE TREATMENT OF DIPG
included ATM and MPL mutations, both in the same tumor,

and PDGFR amplification in an additional three (15%)
TP53-mutated tumors. The presence of detectable PI3Krelated abnormalities in seven (35%) tumors suggests that
the PI3K pathway may play an important role in the genesis
of this disease. PTEN deletions (on chromosome 10q) were
seen in two tumors. Notably absent were mutations in many
genes commonly implicated in malignant gliomas and other
pediatric tumors, including RB1, EGFR, MET, CTNNB1, N-,
H-, or K-RAS, MLH1, EPHA genes and the tyrosine-kinase
domain of KIT or PDGFRA.20 Expression analysis by immunohistochemistry in an overlapping cohort showed EGFR
positivity without EGFR amplification in eight of 20 patients
(40%) and absent PTEN expression in the majority. This
reinforces the emerging importance of the RTK-PI3K pathway in this disease.21
The potential of identifying targetable mutations with
high-throughput mutational profiling suggests a new clinical utility to biopsy at diagnosis. In the emerging era of
personalized medicine, tissue-less diagnosis may rapidly
cease to be a viable option. In the United States, a recently
opened multi-institutional trial for patients with DIPG mandates tumor biopsy in all cases at the time of diagnosis, and
involves a treatment program determined by specific molecular findings. The purpose of this study is fundamentally
two-fold: to enable detailed molecular analysis of a larger
number of primary specimens than has ever previously been
possible, and to evaluate a specific molecularly guided treatment approach (ClinicalTrial.gov; NCT01182350).
Comparative Genomics
Molecular analysis has shown significant differences between pediatric and adult HGG,19,22,23 as well as between
DIPG and other pediatric HGG.11,13,17 EGFR amplification,
for instance, the most common focal abnormality in adult
high-grade glioma, appears to be relatively rare in DIPG, as
in other pediatric high-grade gliomas. However, underlying
similarities are also seen.17 Proneural, proliferative, and
mesenchymal expression subgroups described in adult GBM
have also been described in pediatric GBM regardless of
site.19 Interestingly, PDFRA amplification, which is consistently seen in a minority of DIPGs, is associated with
secondary GBM in both adults and children.17,19
Preclinical Models
Until very recently, DIPG research has been limited by

the lack of faithful animal models. The development of a
genetically engineered mouse model of brainstem glioma
using the RCAS/tv-a system has recently been reported.24
Upregulation of PDGF signaling in the Nestin cells of the
subventricular zone in the fourth ventricle gives rise to
dorsal pontine glioblastoma. Similarly, a mouse model of
DIPG generated by retroviral vector-induced upregulation of
PDGF signal nonspecifically in cells of the dorsolateral pons
has also been reported.25 These models provide a potential
tool for preclinical testing in DIPG, which had previously not
been possible.
A recent preclinical study highlights the potential role of a
unique pontine postnatal neural progenitor cell population
in the genesis of DIPG. A human Nestin and Olig2
neural progenitor cell population that is unique to the
ventral pons and immunophenotypically similar to DIPG
has generated significant interest in the cell type that gives

rise to DIPG.26 This cell population wanes after infancy,
but then peaks again at 6 years, the age at which DIPG
incidence is highest. A similar Olig2 postnatal cell progenitor cell population was noted in the mouse pons, and was
characterized by Hedgehog (Hh) pathway activation. Selective further upregulation of the Hh pathway in these Olig-2
positive cells resulted in postnatal DIPG-like pontine hypertrophy, with proliferation of PDGFR-alpha oligodendrocyte precursor cells, although without dysplastic
transformation. Upregulation of Hh pathway in neurospheres from a DIPG autopsy-derived cell line cells resulted
in increased self-renewal. These findings suggest that DIPG
may represent a deregulation of Hh-activated neuronal
progenitor cells, with additional molecular events causing
malignant transformation.
Conclusion
Any conclusion regarding the biology of DIPG must, at

this point, be considered preliminary. Existing series are
small; with the one exception, all rely at least partially if
not exclusively on autopsy specimens. Nonetheless, there
are certain consistent findings that provide an important
window into the biology of this disease. P53 loss or mutation
is clearly present in a substantial portion of tumors. The
RTK-Ras-PI3K-Akt pathway is affected in a substantial
portion of tumors as well. The prevalence of PTEN loss
suggests that mTOR may be a valid target for further
study.27 The identification of a possible Hh-dependent cancer stem cell provides an intriguing additional potential
target, and raises the possibility of rational multitarget
combinations. Significant differences between DIPG and
both pediatric and adult HGG confirm that DIPG treatment
cannot simply be based on the biology or treatment of
nonbrainstem HGG.28 Many important questions remain
unanswered. Ongoing collection and analysis of pretreatment tumors using advanced molecular techniques such as
next-generation sequence and epigenetic profiling will be
vital.20
The question may rightly be asked whether identification of drugable targets will be sufficient to alter outcome of
DIPG. Adult high-grade glioma remains a poor-prognosis
disease despite the abundance of biologic data already
available. Indeed, for most cancers, with some notable exceptions, the initial use of targeted strategies has not resulted in the immediate dramatic changes in outcome that
some might have hoped. However, molecular targeting as a
therapeutic strategy is still in its infancy. Identification of
genomic aberrations and overexpressed genes is necessary
but not sufficient; a deeper understanding of tumor biology
is required for truly informed target selection. Other important problems, such as ensuring adequate drug delivery and
overcoming drug resistance, must be dealt with. Nonetheless, historic experience in such malignancies as pediatric
acute lymphoblastic leukemia, which over the span of a few
decades was transformed from an incurable to a largely
curable disease, teaches us that innovative approaches and
persistent concerted investigational efforts can ultimately
result in real and lasting change. We are optimistic that new
molecular information in DIPG will prove the crucial first
step toward a better understanding of the biology of this
disease, which, in turn, will be the crucial first step toward
better treatments and a better outcome.
627
ROBISON AND KIERAN
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Nathan J. Robison*
Mark W. Kieran
AstraZeneca;
Celgene; Infinity;
Merck; Novartis
Advantagene;
Amersham;
AstraZeneca;
Celgene; Merck
KGaA; Novartis;
Schering-Plough;
Transmolecular
REFERENCES
1. Hargrave D, Bartels U, Bouffet E. Diffuse brainstem glioma in children:
critical review of clinical trials. Lancet Oncol. 2006;7:241-248.
2. Fangusaro J. Pediatric high-grade gliomas and diffuse intrinsic pontine
gliomas. J Child Neurol. 2009;24:1409-1417.
3. Khatua S, Moore KR, Vats TS, et al. Diffuse intrinsic pontine gliomacurrent status and future strategies. Childs Nerv Syst. 2011;27:1391-1397.
4. Cohen KJ, Heideman RL, Zhou T, et al. Temozolomide in the treatment
of children with newly diagnosed diffuse intrinsic pontine gliomas: a report
from the Childrens Oncology Group. Neuro-oncology. 2011;13:410-416.
5. Haas-Kogan DA, Banerjee A, Poussaint TY, et al. Phase II trial of
tipifarnib and radiation in children with newly diagnosed diffuse intrinsic
pontine gliomas. Neuro-oncology. 2011;13:298-306.
6. Pollack IF, Stewart CF, Kocak M, et al. A phase II study of gefitinib and
irradiation in children with newly diagnosed brainstem gliomas: a report from
the Pediatric Brain Tumor Consortium. Neuro-oncology. 2011;13:290-297.
7. Pollack IF, Jakacki RI, Blaney SM, et al. Phase I trial of imatinib in
children with newly diagnosed brainstem and recurrent malignant gliomas: a
Pediatric Brain Tumor Consortium report. Neuro-oncology. 2007;9:145-160.
8. Louis DN, Rubio MP, Correa KM, et al. Molecular genetics of pediatric
brain stem gliomas. Application of PCR techniques to small and archival
brain tumor specimens. J Neuropathol Exp Neurol. 1993;52:507-515.
9. Zhang S, Feng X, Koga H, et al. p53 gene mutations in pontine gliomas
of juvenile onset. Biochem Biophys Res Commun. 1993;196:851-857.
10. Gilbertson RJ, Hill DA, Hernan R, et al. ERBB1 is amplified and
overexpressed in high-grade diffusely infiltrative pediatric brain stem glioma.
Clin Cancer Res. 2003;9:3620-3624.
11. Zarghooni M, Bartels U, Lee E, et al. Whole-genome profiling of
pediatric diffuse intrinsic pontine gliomas highlights platelet-derived growth
factor receptor alpha and poly (ADP-ribose) polymerase as potential therapeutic targets. J Clin Oncol. 2010;28:1337-1344.
12. Warren KE, Killian K, Suuriniemi M, et al. Genomic aberrations in
pediatric diffuse intrinsic pontine gliomas. Neuro-oncology. 2011. doi:10.1093/
neuonc/nor190.
13. Paugh BS, Broniscer A, Qu C, et al. Genome-wide analyses identify
recurrent amplifications of receptor tyrosine kinases and cell-cycle regulatory
genes in diffuse intrinsic pontine glioma. J Clin Oncol. 2011;29:3999-4006.
14. Wu G, Broniscer A, McEachron TA, et al. Somatic histone H3 alterations in pediatric diffuse intrinsic pontine gliomas and non-brainstem glioblastomas. Nat Genet. 2012. doi:10.1038/ng. 1102.
15. Joshi BH, Puri RA, Leland P, et al. Identification of interleukin-13
628
receptor alpha2 chain overexpression in situ in high-grade diffusely infiltrative pediatric brainstem glioma. Neuro-oncology. 2008;10:265-274.
16. Broniscer A, Baker JN, Baker SJ, et al. Prospective collection of tissue
samples at autopsy in children with diffuse intrinsic pontine glioma. Cancer.
2010;116:4632-4637.
17. Barrow J, Adamowicz-Brice M, Cartmill M, et al. Homozygous loss of
ADAM3A revealed by genome-wide analysis of pediatric high-grade glioma
and diffuse intrinsic pontine gliomas. Neuro-oncology. 2011;13:212-222.
18. Roujeau T, Machado G, Garnett MR, et al. Stereotactic biopsy of diffuse
pontine lesions in children. J Neurosurg. 2007;107:1-4.
19. Paugh BS, Qu C, Jones C, et al. Integrated molecular genetic profiling
of pediatric high-grade gliomas reveals key differences with the adult disease.
J Clin Oncol. 2010;28:3061-3068.
20. Grill J, Puget S, Andreiuolo F, et al. Critical oncogenic mutations in
newly diagnosed pediatric diffuse intrinsic pontine glioma. Pediatr Blood
Cancer. 2011. doi:10.1002/pbc. 24060.
21. Geoerger B, Hargrave D, Thomas F, et al. Innovative therapies for
children with cancer pediatric phase I study of erlotinib in brainstem glioma
and relapsing/refractory brain tumors. Neuro-oncology. 2011;13:109-118.
22. Faury D, Nantel A, Dunn SE, et al. Molecular profiling identifies
prognostic subgroups of pediatric glioblastoma and shows increased YB-1
expression in tumors. J Clin Oncol. 2007;25:1196-1208.
23. Bax DA, Mackay A, Little SE, et al. A distinct spectrum of copy number
aberrations in pediatric high-grade gliomas. Clin Cancer Res. 2010;16:33683377.
24. Becher OJ, Hambardzumyan D, Walker TR, et al. Preclinical evaluation of radiation and perifosine in a genetically and histologically accurate
model of brainstem glioma. Cancer Res. 2010;70:2548-2557.
25. Masui K, Suzuki SO, Torisu R, et al. Glial progenitors in the brainstem
give rise to malignant gliomas by platelet-derived growth factor stimulation.
Glia. 2010;58:1050-1065.
26. Monje M, Mitra SS, Freret ME, et al. Hedgehog-responsive candidate
cell of origin for diffuse intrinsic pontine glioma. Proc Natl Acad Sci U S A.
2011;108:4453-4458.
27. Geoerger B, Kieran MW, Grupp S, et al. Phase II trial of temsirolimus
in children with high-grade glioma, neuroblastoma and rhabdomyosarcoma.
Eur J Cancer. 2012;48:253-262.
28. Gururangan S, Chi SN, Young Poussaint T, et al. Lack of efficacy of
bevacizumab plus irinotecan in children with recurrent malignant glioma and
diffuse brainstem glioma: a Pediatric Brain Tumor Consortium study. J Clin
Oncol. 2010;28:3069-3075.
Is Biopsy Safe in Children with Newly

Diagnosed Diffuse Intrinsic Pontine Glioma?
By Stephanie Puget, MD, PhD, Thomas Blauwblomme, MD, and Jacques Grill, MD, PhD
Overview: Diffuse intrinsic pontine gliomas (DIPGs), with a

median survival of 9 months, represent the biggest therapeutic
challenge in pediatric neuro-oncology. Despite many clinical
trials, no major improvements in treatment have been made
over the past 30 years. In most cases, biopsy is not needed for
diagnosis because DIPG diagnosis is based on a typical
clinical picture with radiologic evidence on magnetic resonance imaging. Therefore, little data on newly diagnosed DIPG
have been published and are confounded by including autopsy
(i.e., postradiation therapy) cases. In most cancers, advancing
to cure has been linked to the discovery of relevant biomarkers, only found by access to tissue. Therefore, to further
understand the biology of DIPG, fresh tissue samples must be
IFFUSE INTRINSIC pontine gliomas (DIPGs), which

represent 15% of all childhood brain tumors, are inoperable neoplasms. Response to radiation therapy is only
transient and chemotherapy has not improved long-term
survival. The development of targeted therapies for these
tumors has been hampered by the lack of knowledge of their
biology and many trials to date have been carried out based
on the misconception that DIPG biology is similar either to
its adult counterparts1-3 or to other pediatric supratentorial
malignant gliomas.1,4 Diagnosis is usually based on the
association of a short history of less than 2 months, cranial
nerves palsies, long-tract signs, and ataxia with typical
imaging findings. DIPG is usually described as an infiltrating tumor mass of the pons, hypointense on T1 and hyperintense on T2 and fluid-attenuated inversion recovery; by
definition, at least 50% of the pons should be involved.
Contrast enhancement, if any, is usually limited and annular. Grading of these lesions has been difficult based on
small biopsies and could therefore not be linked to outcome.
Biopsy of these tumors has been controversial, and most
neurosurgical teams limit the use biopsy to patients with
lesions that have unusual presentation. However, the urgent need to improve the prognosis for patients with these
devastating tumors has led to the reconsideration of the role
of stereotactic biopsy for patients with DIPG. This article
will address the feasibility and safety of stereotactic biopsy
for patients with DIPG, its diagnostic yield, and its role in
redefining this tumor by its molecular signature and profiling targeted therapy.
Is It Safe to Perform a Biopsy for Patients Newly

Diagnosed with DIPG?
Stereotactic biopsies are now completely integrated in the

diagnosis and management of several intracranial lesions.
The role of stereotactic biopsy for patients with newly
diagnosed DIPG remains controversial, and currently the
general attitude is not to biopsy the tumors of these patients.
Stereotactic biopsy of brainstem tumors is an old procedure; it became popular after the first report of this procedure in 1978.5 Ten years later, arguments against brainstem
biopsy were strong because it was believed to have no use, to
be dangerous, and to offer poor yield.6-8 The 1993 manuscript published by Albright and colleagues changed the
course of pediatric DIPG management by claiming that
obtained at diagnosis. However, most neurosurgical teams are

reluctant to perform biopsy in pediatric patients, citing potential risks and lack of direct benefit. Yet, in reviewing 90
patients with and the published data on brainstem biopsy,
these procedures have a diagnostic yield and morbidity and
mortality rates similar to those reported for other brain
locations. In addition, the quality and quantity of the material
obtained confirm the diagnosis and inform an extended molecular screen, including biomarker studyinformation important to designing next-generation trials with targeted agents.
Stereotactic biopsies can be considered a safe procedure in
well-trained neurosurgical teams and could be incorporated in
well-defined protocols for patients with DIPG.
magnetic resonance imaging (MRI) . . . scans provide images that are virtually diagnostic of brainstem gliomas and
yield prognostic information equivalent to that obtainable
from biopsies . . ..8 Since then, the neurosurgical world was
divided into for and against the brainstem biopsy. Despite the reluctance of some neurosurgical teams, others
chose to perform biopsies of brainstem lesions in children
and adults for both unusual lesions and typical ones as part
of a clinical trial.9-16 Papers on stereotactic biopsies in the
brainstem published in the last 20 years represent a substantial amount of knowledge, yet unfortunately often report mixed series of adults and children with a wide range of
diagnoses. These mixed series cite morbidity rates between
0% and 10% and mortality rates between 0% to 3% (Table 1).
However, when biopsy data on pediatric patients with DIPG
are extracted, the diagnostic yield ranges from 96% to 100%,
with no mortality and morbidity less than 5% for the largest
series.
Samadani and Judy performed a meta-analysis of 13
studies of stereotactic biopsy of brainstem lesions in 381
children and adults.17 With a diagnostic yield of 96%, this
study reported one death attributable to a biopsy of a
vascular lesion in an adult, with rates of permanent and
transient neurologic deficits of 4% and 1%, respectively. A
few years later, a second meta-analysis on brainstem lesions
in pediatric patients was published by Pincus and colleagues.13 This review of 192 children revealed a diagnostic
yield of 94.9%, and mortality and morbidity rates of 0.7%
and 4.9%, respectively. Recently, Rajshekhar and Moorthy
reported a series of stereotactic biopsies in 106 children with
brainstem masses. With no mortality or permanent morbidity reported, the authors highlighted that . . . this procedure is safe in children and the benefits outweigh the risks
in patients who are appropriately selected to undergo this
procedure. . . 11 A few years ago, our group started to use
From the Necker Enfants Malades Hospital, Universite Paris Descartes, Sorbonne Paris
Cite, France; Gustave Roussy Cancer Institute, Universite Paris Sud, Villejuif, France.
Address reprint requests to Stephanie Puget, MD, PhD, Department of Pediatric Neurosurgery, Necker Enfants Malades Hospital, 149 rue de Se`vres, 75015 Paris, Universite Paris
Descartes, Sorbonne Paris Cite, France; email: stephanie.puget@gmail.com
1092-9118/10/1-10
629
PUGET, BLAUWBLOMME, AND GRILL

Table 1. Brainstem Stereotactic Biopsies in Children and Mixed Patient Series
Year
No. of
Patients (Type)
Puget
Dellaretti
2012
2011
90 (C)
44 (C)
Rajshekhar
2010
106 (C)
Pirotte
2007
20 (C)
Pincus (meta-analysis)
2006
192 (C)
CRW, TF and TC
94.9a
Samadani (meta-analysis)
2003
381 (A
and C)
CRW BRW, Todd-Weillis,

Leksell, Richert, 292 TF
96
Lead Author
Technique (Frame and Route)
Diagnostic
Yield, %
Leksell, all TC
Talairach, 42 TF, 2 TC
100
93
BRW or CRW, 77 TF, 29 TC
100
PET (18FDG), 3 TF, 17 TC
100
Histopathologic Diagnosis, %, Type
100 DIPG
36 HGG
34.6 LGG, miscellaneous
90 glioma
10 inflammatory
75 glioma
25 others (PNET, teratoma, germinoma)
PNET, neurocytoma, ependymoma,
vasculitis, germinoma. . .
31 HGG
23 LGG
10 metatasis
16 hematomas and miscellaneous
Morbidity,
n (%)
Mortality, %
4 T (4.4)
13 (9.8)
0
0
3 T (2.8)
1 T (5)
1 P (5)
(4.9)b
T (4)
P (1)
0.7c
0.3d
Abbreviations: , not reported; 18 FDG, 18F-fluorodeoxyglucose; A, adults; BRW, Brown-Roberts-Wells; C, children; CRW, Cosman-Robert-Wells; HGG, high-grade
glioma; LGG, low-grade glioma; P, permanent; PET, positron emission tomography; T, transient; TC, transcerebellar; TF, transfrontal.
a
Range, 75% to 100%
b
Range, 0% to 16%
c
0% to 3.3%
d
One adult patient.
stereotactic biopsies for patients with DIPG to obtain both

pathologic confirmation and immunohistochemical assessment of specific biomarkers before including patients in
trials of targeted agents.10,18,19 During the 10-year period,
90 children with pontine lesions resembling DIPG received
this technique at Necker Enfants Malades Hospital in Paris.
All patients received corticosteroids at least 3 days before
the procedure. Using the Leksell stereotactic system, a
transcerebellar approach was used for all patients. The
procedure was carried out with the patient in a prone
position under general anesthesia and the stereotactic coordinates were determined by computed tomography or MRI.
The contrast enhancement was targeted when possible (i.e.,
when it was not too anteriorly close to the pyramidal tract or
KEY POINTS
630
Every attempt to treat children with diffuse intrinsic

pontine glioma has failed in the last 30 years.
In most cases, biopsy is not needed to diagnose a
diffuse intrinsic pontine glioma because diagnosis is
based on the association of a typical clinical picture
and radiologic evidence on magnetic resonance imaging.
Tissue samples are required to further understand
the biology of these lesions and research new therapeutic targets.
Stereotactic biopsy for patients with diffuse intrinsic
pontine glioma is as safe as biopsy for patients with
supratentorial lesions and have a diagnostic yield
above 90%.
The quality and quantity of the material obtained
allow for the confirmation of the diagnosis and for the
performance of an extended molecular screen, including biomarker study, permitting children with newly
diagnosed diffuse intrinsic pontine glioma to enroll
in next-generation clinical trials with targeted
therapies.
near the nuclei of the cranial nerves). In patients with no

contrast enhancement, we targeted the infiltrative part,
close to the middle cerebellar peduncle (Fig. 1). Using a
single trajectory, we used to take two samples at the beginning of our experience. This number was increased up to
eight samples in the past 5 years without additional risk,
allowing for DNA and RNA extraction. The diagnostic yield
was 100% in our series, with no mortality or permanent
deficit, and we observed transient worsening of neurologic
deficit in four patients.
From the literature and our own data, stereotactic biopsy
for children with DIPG can be considered a safe procedure
with high diagnostic yield (Table 1). In a large series of
stereotactic brain biopsy in 270 adults, it has been shown
that increasing numbers of specimens obtained per trajectory and brainstem lesions were not significant risk factors
for morbidity.20
Two routes have been described for brainstem biopsies:
the transcerebellar approach and the transfrontal approach
(Table 1, Figs. 1 and 2). The transfrontal route is longer and
allows sampling of masses located in all the segments of the
brainstem. The positioning of the entry site has to be chosen
carefully to avoid the ventricles, the vascular structures,
and the tentorium. The use of software is recommended for
planning this trajectory. The transcerebellar approach is
shorter, through the middle cerebellar peduncle, and has
less eloquent structures in its trajectory.11,21 It can be used
only for pontine and upper medullary masses, and because
of this, the transcerebellar route for DIPG is preferred by the
authors. In a series comparing both routes, no significant
differences were reported regarding the rates of complication and diagnostic yield.22 The use of functional MRI and
nuclear medicine techniques to define biopsy targets might
increase the diagnostic and prognostic value of the tissue
samples obtained, but the studies comparing the two routes
are mainly concentrated in adults. However, Pirotte and
colleagues published a series about 20 children with infiltrative brainstem tumors in which it was suggested that the
integration of metabolic information from positron emission
tomography might improve the diagnostic yield of the biopsy
sampling.23
SAFETY OF DIPG BIOPSY IN CHILDREN
Fig. 1. Plan neuroimaging that shows the trajectory

from a trancerebellar approach. E Entry; T Target.
(A) Three-dimensional neuroimaging plan. (B) Axial
gadolinium-enhanced T1-weighted magnetic resonance
image. (C) Sagittal gadolinium-enhanced T1-weighted
magnetic resonance image. (D) Coronal gadoliniumenhanced T1-weighted magnetic resonance image.
In addition, to increase direct application of chemotherapeutic agents to the DIPG, attempts have been pursued to
circumvent the blood-brain barrier using the convectionenhanced delivery technique. Theoretically, the safety of
this procedure is similar to that of biopsy. Preclinical models
and clinical trials have demonstrated the feasibility, efficacy, and safety of this technique.24 Moreover, two children
with brainstem lesions demonstrated no neurologic deficits
after infusion.25 Therapeutic protocols are currently under
way, but the choice of the appropriate agent to deliver with
this technique remains in question.
Why Perform a Biopsy for Patients with Newly

Diagnosed DIPG?
The lack of samples from patients with newly diagnosed

DIPG has limited our understanding of tumor biology, which
hinders the development of newer therapies for patients
with these devastating tumors. As targeted therapy development undoubtedly requires tissue, it could be argued that
such advances will only be optimized with the knowledge
that biopsies provide in terms of tumor biology and the
identification of new targets. In recent papers, authors
defend the idea that a biopsy of patients with newly diag-
Fig. 2. Plan neuroimaging showing the trajectory

with a transfrontal approach. E Entry; T Target.
(A) Three-dimensional neuroimaging plan. (B) Axial
gadolinium-enhanced T1-weighted magnetic resonance
image. (C) Sagittal gadolinium-enhanced T1-weighted
magnetic resonance image. (D) Coronal Gadoliniumenhanced T1-weighted magnetic resonance image.
631
PUGET, BLAUWBLOMME, AND GRILL
nosed DIPG should be performed to increase our knowledge

of tumor biology to provide clues to improve their prognosis.9,24,26 However, for patients with typical DIPG evidence
on MRI, biopsies should not be performed only to confirm the
diagnosis, because the risk of the procedure, even though
minimal, is not null. Biopsy should be part of a wellconducted clinical trial or a research program approved by
an ethics committee.9,26 The authors agree with the conclusions of Wilkinson and Harris, which stated that . . . Once
emotional and social interests are taken into account there
seems little doubt that brainstem biopsy could be lawful
even if there was no benefit to the childs medical interests . . ..27
Even if the amount of tumor tissue obtained by stereotactic biopsy is limited by the size of the needle, it has been
shown that it could provide enough tissue for histopathologic
diagnosis and immunohistochemical staining. Moreover, the
authors have recently shown that this surgical technique
could allow multiple biopsies samples (up to eight) to provide
enough tissue for further genomic analyses and stem-cell
culture.26,28,29 In the author series, one or two biopsies were
used for histologic diagnosis and immunohistochemistry.
The remaining biopsies were snap-frozen with cytologic
control smears within minutes of surgical removal, and
nucleic acids were extracted later. A median of 3.325 microgram (microg) of DNA (range, 0.805 microg to 21.5 microg)
and 2.332 microg of RNA (range, 0.048 microg to 15.84
microg) could be extracted from the biopsies. One to three
samples were needed to obtain enough nucleic acids, depending on the infiltrative rate of the tumor cells. An
integrated molecular profile was created to identify two
distinct subgroups of DIPG with specific abnormalities: one
showing a mesenchymal gene expression profile and the
other showing a more proliferative gene expression signature. The former group of tumors showed a better survival
than the latter. The poor-prognosis group defined by gene
expression profiling showed an oligodendroglial differentiation that could be correlated with an adverse prognosis in
another validation cohort. In addition, this group of tumors
showed amplification in platelet-derived growth factor
receptor-alpha with or without mutation in the external

domain and MET gain that could represent relevant therapeutic targets. This study is the first to comprehensively
define the biologic alterations of DIPG at diagnosis, allowing
the discovery of novel therapeutic targets.29 The material
obtained also afforded screening for oncogenic mutations
with currently available targeted drugs. Our group was able
to identify PI3KCA mutations in 15% of DIPG, which offers
another relevant therapeutic target.26 Finally, cultures of
DIPG could be derived as cell lines or as neurospheres.28
They represent irreplaceable tools both for preclinical studies of new therapeutic agents and for understanding the
oncogenesis of DIPG.
Recently, there has been a worldwide resurgence of interest in pediatric brainstem biopsy in the hopes that molecular
profiling could help to find new therapeutic targets. To this
end, several international consensus conferences on DIPG
have been organized in North America and Europe in the
past 3 years.30,31 One can feel that there is a growing body
of evidence that biopsy for most patients with newly diagnosed DIPG is now considered rational because the procedure might alter treatment with targeted therapy and might
help in correlative biology with appropriate biomarker response. Biopsy might also help to guide therapies for patients with relapsed disease, to look for active treatments,
and to develop relevant biologic models.20
Conclusion
DIPG remains a leading cause of death for children with

brain tumors. The role of diagnostic biopsy for patients with
these tumors has been controversial because of the high
eloquence of the brainstem and the lack of direct benefit for
the patient. Based on the literature and our own data,
stereotactic biopsy for patients with DIPG is approximately
as safe and diagnostic as supratentorial biopsy, and the
amount of tissue obtained allows for molecular analysis.
This technique should be offered to these patients and opens
new perspectives for the characterization of biomarkers that
permit children with newly DIPG to enroll in nextgeneration clinical trials with targeted therapies.26
Author
Stephanie Puget*
Thomas Blauwblomme*
Jacques Grill*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Bax DA, Mackay A, Little SE, et al. A distinct spectrum of copy number
aberrations in pediatric high-grade gliomas. Clin Cancer Res. 16:3368-3377.
2. Qu HQ, Jacob K, Fatet S, et al. Genome-wide profiling using singlenucleotide polymorphism arrays identifies novel chromosomal imbalances in
pediatric glioblastomas. Neuro Oncol. 12:153-163.
3. Paugh BS, Qu C, Jones C, et al. Integrated molecular genetic profiling of
pediatric high-grade gliomas reveals key differences with the adult disease.
J Clin Oncol. 28:3061-3068.
4. Zarghooni M, Bartels U, Lee E, et al. Whole-genome profiling of pediatric
diffuse intrinsic pontine gliomas highlights platelet-derived growth factor
receptor alpha and poly (ADP-ribose) polymerase as potential therapeutic
targets. J Clin Oncol. 28:1337-1344.
5. Gleason CA, Wise BL, Feinstein B. Stereotactic localization (with com-
632
puterized tomographic scanning), biopsy, and radiofrequency treatment of

deep brain lesions. Neurosurgery. 1978;2:217-222.
6. Stroink AR, Hoffman HJ, Hendrick EB, et al. Diagnosis and management of pediatric brain-stem gliomas. J Neurosurg. 1986;65:745-750.
7. Epstein F, McCleary EL. Intrinsic brain-stem tumors of childhood:
surgical indications. J Neurosurg. 1986;64:11-15.
8. Albright AL, Packer RJ, Zimmerman R, et al. Magnetic resonance scans
should replace biopsies for the diagnosis of diffuse brain stem gliomas: a
report from the Childrens Cancer Group. Neurosurgery. 1993;33:1026-1029.
9. Leach PA, Estlin EJ, Coope DJ, et al. Diffuse brainstem gliomas in
children: should we or shouldnt we biopsy? Br J Neurosurg. 2008;22:619-624.
pontine lesions in children. J Neurosurg. 2007;107:1-4 (suppl).
SAFETY OF DIPG BIOPSY IN CHILDREN

11. Rajshekhar V, Moorthy RK: Status of stereotactic biopsy in children
with brain stem masses: insights from a series of 106 patients. Stereotact
Funct Neurosurg. 2010;88:360-366.
12. Chico-Ponce de Leon F, Perezpena-Diazconti M, Castro-Sierra E, et al.
Stereotactically-guided biopsies of brainstem tumors. Childs Nerv Syst.
2003;19:305-310.
13. Pincus DW, Richter EO, Yachnis AT, et al. Brainstem stereotactic
biopsy sampling in children. J Neurosurg. 2006;104:108-114.
14. Valdes-Gorcia J, Espinoza-Diaz DM, Paredes-Diaz E. Stereotactic biopsy of brain stem and posterior fossa lesions in children. Acta Neurochir
(Wien). 1998;140:899-903.
15. St George EJ, Walsh AR, Sgouros S. Stereotactic biopsy of brain
tumours in the paediatric population. Childs Nerv Syst. 2004;20:163-167.
16. Dellaretti M, Touzet G, Reyns N, et al. Correlation among magnetic
resonance imaging findings, prognostic factors for survival, and histological
diagnosis of intrinsic brainstem lesions in children. J Neurosurg Pediatr.
8:539-543.
17. Samadani U, Judy KD. Stereotactic brainstem biopsy is indicated for
the diagnosis of a vast array of brainstem pathology. Stereotact Funct
Neurosurg. 2003;81:5-9.
18. Geoerger B, Hargrave D, Thomas F, et al. Innovative therapies for
children with cancer pediatric phase I study of erlotinib in brainstem glioma
and relapsing/refractory brain tumors. Neuro Oncol. 2011;13:109-118.
19. Geoerger B, Morland B, Ndiaye A, et al. Target-driven exploratory
study of imatinib mesylate in children with solid malignancies by the
Innovative Therapies for Children with Cancer (ITCC) European Consortium.
Eur J Cancer. 2009;45:2342-2351.
20. McGirt MJ, Woodworth GF, Coon AL, et al. Independent predictors of
morbidity after image-guided stereotactic brain biopsy: a risk assessment of
270 cases. J Neurosurg. 2005;102:897-901.
21. Backlund EO. A new instrument for stereotaxic brain tumour biopsy.
Technical note. Acta Chir Scand. 1971;137:825-827.
22. Dellaretti M, Reyns N, Touzet G, et al. Stereotactic biopsy for brainstem tumors: comparison of transcerebellar with transfrontal approach. Stereotact Funct Neurosurg. 2012;90:79-83.
23. Pirotte BJ, Lubansu A, Massager N, et al. Results of positron emission
tomography guidance and reassessment of the utility of and indications for
stereotactic biopsy in children with infiltrative brainstem tumors. J Neurosurg. 2007;107:392-399.
24. Khatua S, Moore KR, Vats TS, et al. Diffuse intrinsic pontine gliomacurrent status and future strategies. Childs Nerv Syst. 27:1391-1397.
25. Lonser RR, Warren KE, Butman JA, et al. Real-time image-guided
direct convective perfusion of intrinsic brainstem lesions. Technical note.
J Neurosurg. 2007;107:190-197.
26. Grill J, Puget S, Andreiuolo F, et al. Critical oncogenic mutations in
newly diagnosed pediatric diffuse intrinsic pontine glioma. Pediatr Blood
Cancer. 2012;58:489-491.
27. Wilkinson R, Harris J. Moral and legal reasons for altruism in the case
of brainstem biopsy in diffuse glioma. Br J Neurosurg. 2008;22:617-618.
28. Thirant C, Bessette B, Varlet P, et al. Clinical relevance of tumor cells
with stem-like properties in pediatric brain tumors. PLoS One. 2011;6:e16375.
29. Puget S, Philippe C, Bax DA, et al. Mesenchymal transition and
PDGFRA amplification/mutation are key distinct oncogenic events in pediatric diffuse intrinsic pontine gliomas. PLoS One. 2012;7:e30313.
30. Jansen MH, van Vuurden DG, Vandertop WP, et al. Diffuse intrinsic
pontine gliomas: a systematic update on clinical trials and biology. Cancer
Treat Rev. 2012;38:27-35.
31. Bartels U, Hawkins C, Vezina G, et al. Proceedings of the diffuse
intrinsic pontine glioma (DIPG) Toronto Think Tank: advancing basic and
translational research and cooperation in DIPG. J Neurooncol. 2011;105:119125.
633
Ethics of Biopsy in Children with Newly

Diagnosed Diffuse Intrinsic Pontine Glioma
By Nicholas K. Foreman, MD
Overview: To understand the ethical dilemmas that beset

this issue of biopsy in children with newly diagnosed diffuse
intrinsic pontine glioma, one must understand both the history
behind it and the current dominant interpretation of ethics
through the medium of the institutional review boards. It is
N DIFFUSE intrinsic pontine glioma (DIPG), the stage

for the controversy over biopsy was set in the early 1990s.
The early optimism engendered by the often dramatic responses to radiation had faded and the fatal nature of DIPGs
had become all too obvious. The morbidity from biopsy in
that era was low, but there was no apparent benefit in
looking at the tissue over diagnosis by radiology. Indeed,
histology seemed more confusing than radiology given that
tumor grade did not influence survival.1 The standard diagnosis then became the typicality of the scan. If the scan
was typical there was no need to biopsy. Oddly enough, this
standard method of making a diagnosis was never standardized and there remained a huge variation in what was
considered typical or atypical.2
DIPGs were then diagnosed by scan. Radiation was administered as a standard therapy and the tumors would
shrink. The quality of life for many would improve for a few
months. However, all children with typical (and, indeed,
most with atypical scans) would die. Time to death was
stunningly uniform across all studies, with a median of 9 to
10 months.
Given the lack of curative therapy, the children with this
terrible disease became the subject of literally dozens of
phase I and II trials. These were passed easily by institutional review boards (IRBs) because there was a possibility
of direct benefit. No IRB seemed to have reviewed the
increasing evidence, added by each consecutive trial, that
there was no benefit from these non biology-directed phase
I and II trials.3 Children with this terrible disease suffered
again and again from vomiting and diarrhea from yet
another fruitless trial. Perhaps more than 1,000 children
worldwide were entered onto these trials with no benefit and
no questioning from the IRBs.
By the turn of this century, DIPG deaths were making up
to 30% to 40% of the deaths in the pediatric neuro-oncology
program as prognosis improved for other brain tumors.
There had been an explosion of biologic information about
most pediatric brain tumors, but virtually nothing was
known about the biology of DIPGs. There were discussions
with parents and parental lead advocacy groups. A proposal
was made by me at the Childrens Hospital Colorado in 2008
to biopsy a limited number of children with all the costs
being borne by institutional money. In the proposal it was
stated that there was no direct benefit to the children
concerned, instead of saying the benefit was in the order of
phase I or II trials. The local IRB declined to approve it
citing federal regulations that procedures done on children
should offer the possibility of benefit if they were not
minimal risk.4 The IRB asked for a panel of experts to be
assembled and this led to a national panel being assembled
by the U.S. Food and Drug Administration (FDA). Although
634
also important to understand that this article represents the

authors personal viewpoint. At a consensus meeting to discuss the issue of biopsies in Diffuse Intrinsic Pontine Glioma
(DIPG) at the National institutes of Health held in the fall of
2011, there were a variety of opinions expressed.
the proposal was supported by most members of the panel,

the ethicists on the panel were opposed and the local IRB
rejected the proposal.5
This brought up some interesting ethical issues by various
panel members and participants. I am going to give some
personal comments on these in the following sections
First, That There Has to Be the Possibility of Some
Direct Benefit to the Child to Allow a Procedure or
a Therapy That May Cause More Than Minimal Harm
This does not have to be quantified and may in fact be

immeasurably small (such as the continuing proposals for
non biology-based phase I and II trials in this tumor).
Indeed, it would appear it could even be speculative. IRBs
(and apparently ethicists) considered surgical procedures to
be in a different class from chemotherapy regimens, although the latter may inflict weeks of misery on a child. The
French data showing a low risk to the procedure was
insufficient in the view of the ethicists.6
So, if one said that the biopsy would include looking for
v600e mutation of BRAF or a H3F3A mutation and that the
child could at progression be considered a candidate for
therapy targeting these mutations, this might pass muster even though there is no evidence these agents are
effective in childhood brain tumors.
Second, That Societal or Family Benefit Should Not
Be Regarded in a Discussion of the Ethics of
Performing Biopsy on a Child
One of the ethicists said that consideration of societal or

family benefit was more in line with European ethics than
American. The basis for this statement is difficult to find and
perhaps lies more in the interpretation of ethics. Family
members at the meeting said the death of their child seemed
in vain, that nothing was learned and more children would
go on to die as a result of this terrible disease. They felt that
both they and the child (when older) would have had comfort
from knowing that others would be less likely to suffer. One
of the panel members implicitly stated that the family had
no rights in this matter and could not consent to their childs
undergoing a procedure without direct benefit even when
the child had a fatal disease. Some of the parents were
From the University of Colorado, Aurora, CO.

Address reprint requests to Nicholas K. Foreman, MD, Professor of Pediatrics, SeebaumTschetter Chair of Neuro-Oncology, University of Colorado, 13123 E. 16th Ave., B115,
Aurora, CO 80045; email: nicholas.foreman@childrenscolorado.org.
1092-9118/10/1-10
CHILDREN WITH DIFFUSE INTRINSIC PONTINE GLIOMA
outraged by this and considered this a paternalistic void of

their rights.
Third, That All Alternatives to Performing Biopsies on
Children Should Be Considered
Basically what was being asked was whether there some

way of advancing knowledge and therapy without performing biopsies, which would seem absolutely uncontroversial.
This statement and the failure of this proposal put many of
the parental advocacy groups firmly behind an effort to
obtain autopsy samples. This effort was strikingly successful
and has resulted in great advances in the knowledge base for
this disease. However it should be noted even in the recent
Nature Genetics article showing a potential driving muta-
KEY POINTS
Diffuse intrinsic pontine gliomas are almost uniformly fatal.

There is considerable controversy over the ethics of
biopsying the tumor.
The issue of direct benefit is central to this controversy.
The issue of whether familial or societal benefit
should be considered when a biopsy on a child is
contemplated is controversial.
The issue of who declares a procedure standard is
uncertain.
tion in a specific histone gene in many autopsy samples, a

small number of upfront biopsies were used to suggest that
this was a driving mutation not a passenger mutation.7 The
article did not say how these specimens were obtained.
Fourth, Declaration that a Procedure Is Standard
A participant who serves on an IRB pointed out that most

IRBs relied on local expertise to declare what was standard.
If an institutions neurosurgeons declared that biopsy was
standard, then it was not in fact an IRB issue. This brings up
the whole question of what is a standard and whether its
determination at a local level should be influenced by national consensus. Is there an obligation on the part of local
investigators when considering a procedure standard to
discuss with their IRB whether there is national controversy
about that standard? Do IRBs have an ethical obligation to
investigate what is standard or whether a therapy has a
realistic possibility of benefit?
As a Thought-Provoking Exercise
Lets say that there had never been a statement in an

influential journal indicating that biopsies of typical DIPG
were not needed for diagnosis and should not be done.
Lets say there were biopsies performed throughout the
1990s and an H3F3A mutation was identified in 1999
(purely hypothetical given technical considerations).
Lets say that targeted therapy doubled survival time and
cured a small number (say 10% to 20%) by 2005.
Is it possible that fear of a biopsy (with a known very small
morbidity and mortality risk) resulted in shortening the
lives of many and even the loss of life?
Author
Nicholas K. Foreman*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Albright AL, Packer RJ, Zimmerman R, et al. Magnetic resonance
scans should replace biopsies for the diagnosis of diffuse brain stem
gliomas: a report from the Childrens Cancer Group. Neurosurgery. 1993:
33:1026-1030.
2. Hankinson TC, Campagna EJ, Foreman NK, et al. Interpretation of
magnetic resonance images in diffuse intrinsic pontine glioma: a survey of
pediatric neurosurgeons. J Neurosurg Pediatr. 2011; 8:97-102.
3. Frazier JL, Lee J, Thomale UW, et al. Treatment of diffuse intrinsic
brainstem gliomas: Failed approaches and future strategies. J Neurosurg
Pediatr. 2009;3:259-269.
4. Anderson BD, Adamson PC, Weiner SL et al. Tissue collection for
correlative studies in childhood cancer clinical trials: ethical considerations

and special imperatives. J Clin Oncol. 2004;1:4846-4850.
5. U.S Food and Drug Administration. Minutes of the Joint Meeting of
the Pediatric and Oncology/Advisory Committee, April 27, 2009. http://www.fda.
gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/PediatricAdvisoryCommittee/UCM171523.pdfBrainstemgliomas. Accessed March 2, 2012.
pontine lesions in children. J Neurosurg. 2007;107(1 Suppl):1-4.
7. Wu G, Broniscer A, McEachron TA, et al. Somatic histone H3 alterations
in pediatric diffuse intrinsic pontine gliomas and non-brainstem glioblastomas. Nat Genet. 2012; 44:251-253.
635
TRANSITION AND DECISION MAKING FOR

PALLIATIVE CARE OF PATIENTS WITH
PEDIATRIC CANCER
CHAIR
Jennifer W. Mack, MD, MPH
Boston, MA
SPEAKERS
Chris Feudtner, MD, PhD, MPH
Philadelphia, PA
Pamela S. Hinds, PhD, RN
Childrens National Medical Center
Washington, DC
Communication and Decision Support for

Children with Advanced Cancer and
Their Families
By Jennifer W. Mack, MD, MPH, Chris Feudtner, MD, PhD, MPH,
and Pamela S. Hinds, PhD, RN
Overview: Clinician communication related to treatment decision making is a fundamentally important health care intervention and is often reported by parents of seriously ill
children to be the most valued of clinician skills. Since
different children and families have different communication
styles and expectations, and since these may change over the
course of the illness experience, one of the early and recurring
tasks is to clarify and work with these diverse styles and
expectations. Adopting a stance of compassionate desire to
know more about patients and families, in addition to imparting information, is vital, and can be facilitated by following a
general strategy of ask, tell, ask. In addition to the exchange
of information, communication between clinician and patient
ALLIATIVE CARE is centered on children with lifethreatening illnesses and their families. As such, the
values and goals held by individual children and families
guide plans for care, and the childs quality of life and
symptomsphysical, psychologic, spiritualare of primary
importance. These principles affirm that the child is valued
and guide care from diagnosis through all phases of illness.
For patients with progressive cancer or serious complications, communication practices founded on these principles
set the stage for palliative care when it is needed. In this
article, we offer practical guidance for clinicians who care
for children with advanced cancer, addressing various aspects of communication and supportive decision making
across the illness trajectory (see Fig. 1).
Communication across the Illness Experience
Communication, central to the work of caring for children

with advanced cancer, has three primary purposes. First,
communication allows for the exchange of information and
the development of shared knowledge. Communication goes
both ways: from the family and child as well as to them, and
sometimes wordlessly, in shared silence. Second, communication serves as the foundation for a relationship between
the child, parent, and clinician, in which the parents and
child can feel known and understood. Third, communication
provides a forum for decision making. The type of decision
may change over time, but even when parents and children
approach painful decisions about end-of-life care, they can
use familiar processes and relationships as an anchor.
and family also involves the signaling and exchange of emotions, in which the pace, verbal inflection, and body language
of the conversation are fundamental. Discussions about prognosis and goals of care, while needing to be handled in a
gentle manner, should start early in the illness experience and
be revised whenever there is a relapse or major complication.
Children often want to participate in these conversations to a
degree of their own choosing, which they themselves can
clarify. Effective and empathetic clinician communication can
greatly facilitate decision making and care for children with
advanced cancer and their families, and provide a substantial
source of comfort.
Their answer helps the clinician to know where to start, and

correct any misconceptions. After communicating information, the clinician can use a final question to check understanding (I want to make sure that Ive explained things.
Can you tell me what you are taking away from this
conversation?) and assess the emotional effect (How are
you feeling now?) of the news.
Even at diagnosis, conversations may be incomplete without honest discussion about the childs future, including the
nature of the illness and the prognosis. These conversations
are hard cognitively, emotionally, socially, and spirituallyand clinicians may understandably avoid them.1,2 Yet
most parents worry about their childs future from the
day they hear the word cancer, and want prognostic
information, even if they find it upsetting.3 In addition,
starting such conversations at diagnosis helps with transition to palliative care if treatment fails.
We recommend the following steps when talking about
prognosis,4,5 in line with the general strategy of ask, tell,
ask:
Consider what the child or family needs to know about
prognosis, and when they need to know it. In some situations, if the family does not want prognostic information,
then the clinician need not discuss it. In other situations,
such as a very poor prognosis or acute deterioration, true
informed consent may not be possible without it.
Ask the family and/or child what they understand about
what lies ahead.
Unless you have already decided that the family needs
Sharing Information
The first communication goal that spans the illness trajectory is the sharing of information. Clinicians must both
seek information from and impart information to children
and their families, and what we choose to discuss sets a tone
for the central issues of care.
A useful general communication strategy is ask, tell,
ask, in which communication from the clinician is framed
by the child and family. For example, when discussing a
cancer diagnosis, one might start by asking the child or
family, What is your understanding of the diagnosis so far?
From the Department of Pediatric Oncology and the Center for Outcomes and Policy
Research, Dana-Farber Cancer Institute, Boston, MA; Harvard Medical School, Boston,
MA; Pediatric Advanced Care Team, Department of Medical Ethics, PolicyLab, The
Childrens Hospital of Philadelphia, Philadelphia, PA; Department of Pediatrics, The
Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA; Childrens National Medical Center, Washington, DC; The George Washington University,
Washington, DC.
Address reprint requests to Jennifer Mack, MD MPH, 450 Brookline Ave., Boston, MA
02215; email: jennifer_mack@dfci.harvard.edu.
1092-9118/10/1-10
637
MACK, FEUDTNER, AND HINDS
detailed prognostic information, ask them what they want to

know. Some patients and families will wait for the physician
to offer such information. Clinicians who do not routinely
offer prognostic information may leave some children and
families without information they want.
For families who want information, provide an honest
estimate, without euphemisms. Adult patients with cancer
are more likely to understand the prognosis with one negative fact; although providing reasons to be more hopeful
than the mortality or morbidity data suggest may make
physicians feel better, it can cause families to be overly
optimistic.6
Ask the parents and the child how they are feeling, and
respond appropriately.
Remind the child and/or family that you will be see them
through. Although we cannot ensure that the course will be
easy, we can help them to feel less alone.
When treatment fails or results in a life-threatening
complication, this information must also be shared. As
outlined above, we recommend assessing the childs and
familys understanding of the situation, providing information, then checking understanding and asking how they
are doing (ask, tell, ask). Importantly, families value clear
language, without euphemisms. Clinicians sometimes focus
on treatment rather than prognosis,7 perhaps because this
is something we can offer in a difficult situation. Conveying
a poor prognosis with honesty, however, can help parents
make the best decisions for their children. Many families
also appreciate honest expressions of empathy or sadness.
While there are no right words for this situation, one might
say, I wish the news was different. I am afraid that the
scans show the cancer has returned. Unfortunately we know
that when the cancer comes back, it is not curable. This is
such sad news. Often the precise words used are less
important than the relationship between the clinician and
the patient and family, since communication works best in a
context of caring and trust.
Developing Relationships
The second communication goal, the development of a

relationship between the parents, child, and clinician, also
begins at diagnosis. Children and families need to know that
their caregiver cares about them, can be trusted, respects
KEY POINTS
638
Ask, tell, ask is a useful general communication

strategy.
Each child and family has a different style and set of
expectations regarding communication and making
decisions, and clinicians need to clarify and work with
these styles and expectations.
Emotions play a vital role in communicationand
are often the most important substance being communicated.
Asking a child What makes you happiest? can be a
gateway to many important conversations.
Asking a parent What do you need to do to feel like
you are being the best parent you can be for your
child? can likewise open up deep lines of discussion.
and listens to them, and will continue to be there as the

illness evolves. These relationships develop organically and
in diverse directions, rather than following a prescribed
path. Nevertheless, clinicians can foster such relationship
with questions and behaviors that engage the child and
parents in teaching the clinician about who they are.
Most importantly, this work is fostered by listening.
Children especially may not always wish to talk about issues
related to their illness, but sometimes, they may leave the
conversational door ajar. Clinicians who are patiently and
consistently listening are more likely to detect and use that
opening. Furthermore, by listening, clinicians reassure children and parents that their words have meaning.
Questions can help build relationships, and often have
dual roles: they allow the child and family to express
themselves, and give clinicians information that helps them
to provide the best possible care. Posing these questions
and listening to the answers does not necessarily require
much time; sitting by the bedside for a few minutes at the
end of the day may suffice to learn about one or two of these
issues, and reinforces the message that those in the room
are valued.
What kind of a person are you? How would you describe
yourself? These questions unfold across multiple conversations, with many possible prompts: What are you proudest
of ? How do you like to be thought of ? Where do you find
strength and support? The goal is to learn about this child,
and this family, as unique individuals. Early conversations
(about family members, school, and interests, for example)
may evolve over time into deeper ones about spirituality
or the meaning of illness (Do you ever wonder why this
happened to you?).
Tell me about a time when you were the happiest. This
question has particular value in getting to know the child
as a person. Children with cancer share a lot of tough times
with their medical team. This question allows them to share
what is good and meaningful to them, which is good in itself,
for the child, family, and clinician. Additionally, this can
aid clinicians in conversations about care when treatment
options have failedRemember when you told me about
. . . ? I am wondering if we can find our way to more times
like that.
How does communication work in your family? How do
you want it to work? Asking this question shows respect for
the family system, which may be influenced by cultural
values, past experience, and the unique ways that a family
comes together around a childs illness. Asking questions
and listening are the best tools to understand these differences. Some families have very different preferences from
those of the clinician; for example, a family may wish to
exclude an adolescent patient from conversations about the
diagnosis. By asking at the beginning, the clinician can
learn what matters to the family and start a conversation
about how to proceed. A plan that works well for everyone
requires ongoing conversation, but this question reminds
families and patients that their preferences are paramount.
As you think about the future, what do you worry about the
most? Worries about the future are nearly universal at the
time of diagnosis, yet often go unspoken. This question
allows the clinician to acknowledge these difficult emotions.
Unrealistic worries can be corrected, and reasonable worries
can be recognized, understood, and met with empathy. This
COMMUNICATION AND DECISION SUPPORT FOR PEDIATRIC ADVANCED CANCER
Fig. 1.
Overview of communication issues and strategies.
639
question also sets the stage for formulating of goals of care,

which may address fears about outcomes.
After a relapse or serious complication, communication
can continue to affirm and deepen the relationship between
child, family, and clinician. Useful questions focus on the
emotional and existential experience of illness and hopes
and fears for the future, and allow children and families to
reflect on the childs life and its meaning. Some examples
are below, although not every question is appropriate for
every child; as always, careful listening may help identify
good times for these conversations.
Do you ever think about what all of this means?
As you think about the future, is there anything that
you are especially afraid of? Is there anything that you are
especially hoping for?
I know you have pain, but is there anything that is even
worse than the pain?8
Is there anything youd especially like for your family to
know about you?
How do you like to be thought of ?
Making Decisions
The third goal of communication, shared decision making,

is also relevant throughout illness. Starting treatment with
language based on goals of care can help to make the
transition to palliative care feel more natural. For most
children, the initial goal of treatment is cure. Yet evidence
suggests that parents consider palliation a priority even
during initial, cure-focused care.9 Thus, we recommend making goals an explicit part of all conversations about cancer
treatment: As you think about your (your childs) illness,
what is most important to you? and Aside from this most
important goal, what else are you hoping may be possible?
Once goals have been identified, clinicians can respond by
framing care, including treatment options, in the context of
these goals. Rather than relying on parents to decide on
their own, clinicians may make recommendations about care
consistent with parents goals and values. For example, one
might say, You just told me how important it is for her to
have a good quality of life. Given that, I wouldnt recommend
intensive chemotherapy, because I am worried that she will
spend a lot of time in the hospital and the clinic, instead of
doing things she enjoys. Id like to recommend that we offer
her any treatments we think may make her feel better,
but stay away from treatment that is likely to be difficult or
make her feel unwell. In doing so, clinicians can join
parents in what may be very difficult decisions. Individual
parents have a wide range of preferences for involvement in
decision making, and preferred roles can change over time
as parental experience and the nature of decisions change.
Thus clinicians may wish to ask parents how they want
decision making to look, and work to support their preferences.
Starting Conversations Early
Conversations about prognosis and goals of care are appropriate for all children with cancer and their families.
They address fears about the future, help clinicians learn
what the child and family consider important, and allow the
child, family, and clinician to know one another as people.
Clinicians can return to these discussions and, in doing so,
reassure children and their families that they matter, that
640
they are known, and that their care team will do everything
possible to uphold their wishes.
Emotions and Communication
In all conversations, emotions play several roles in shaping the interaction.10 First, emotions affect the way people
think and behave. Emotions like fear, sadness, anger, joy,
happiness, surprise, relief, guilt, shame, disgust, or contempt (which constitute a core set of emotions that many
psychologists identify as primary emotional responses) orient a person toward different aspects of a situation and color
their interpretation of it. Second, whether in the background
or foreground of any particular interaction, emotions are
part of what individuals communicate to each other, wittingly or not: by a combination of word choice, vocal inflection, body language, and other cues, people show how they
feel. Third, what people show each other may or may not
accurately reflect how they truly feel, and resultant misunderstandings can profoundly alter the tone and outcome of
particular conversations and future interactions. Fourth,
when children are seriously ill, parents often have both
strong negative feelings (their child is so ill they are afraid
or angry) and strong positive feelings (they love their child
with boundless affection and pride), further complicating
the handling of emotions, communication, and decision making.11
Because of these and other important effects of emotions
on communication, clinicians who care for children with
advanced cancer and their families must become aware of
and seek to improve their own emotional communication
skills. As yet, there are no well-established evidenced-based
best practices regarding the emotional side of communication;12 acknowledging this, we recommend the following
techniques:
Keep tabs on your own emotions. Even before a conversation begins, and periodically during the encounter, clinicians
should spend a moment focusing on their own feelings, and
name the emotions they have; labeling how one is feeling
helps in self-regulating the effect that emotions have on
ones own thinking and behavior.
Engage in a common purpose. At the outset of the conversation, after offering a personal greeting, work to quickly
establish with the patient and family what you are all trying
to get out of the discussion: Thanks for meeting with me.
There are a few things I know I want to discuss with you.
What do you want to talk about, what would be helpful?
Slow down. Often clinicians, feeling either time pressure
or the need to discuss lots of information, move the conversation forward with at their own quickened pace. Slowing
down is a cardinal way to show empathy and caring. Even
if fewer facts are covered in a given period of time, the
emotional quality of the conversation is enhanced.
Listen and summarize. Observational studies of physicians talking with patients show that physicians do most
of the talking, but that patients leave encounters far more
satisfied when physicians do less talking and more listening.
A particularly effective technique is for the clinician to ask
a question, listen for 30 seconds to a minute or two, then
summarize what has been said: Okay, let me see if I heard
you correctly, you are most concerned about . . .
Solicit permission. Before addressing topics that may be
difficult to discuss, solicit permission to do so: There is
something else, something important, that may be hard to

talk about but I think we need to talk about, okay?13
Ask about what they are hoping for. When confronting
serious illness, people often become focused on making it
go away. When asked, In trying to take the best care of you
that I can, it will help me to know what you are hoping
for can you share your thoughts with me?, the first
answer will often be a cure or miracle. Be patient and
supportive. I also hope that can happen. What else are you
hoping for? The subsequent hopes are likely to be very
helpful in understanding distressing symptoms or previously unarticulated fears or goals that can be addressed by
good comprehensive care.14
Provide a warning shot when bad news is coming. If you
are aware that you about to share some bad news, provide
the patient or family with a warning shot: I wish the news
was different. The test shows that the cancer has come
back.
When people cry or shout, stop. People experience another
person as being empathic if this other person seems to get
how they are feeling and respond in an appropriate manner.
Continuing to talk when someone cries, shouts, or becomes
visibly distraught, is a recipe for being perceived as lacking
empathy.
Observe and name. People often show their feelings but
may not verbalize how they are feeling. In many situations,
simply stating that you are aware that they have strong
feelings can be helpful: I see how much this news has upset
you.
Before ending, be clear about next steps. Do not end an
emotionally charged conversation without a clear sense of
what will happen next and when you will talk again. Map
out a plan. Make a commitment about the time and place of
your next conversation.
The only way to incorporate these skills into routine
clinical practice is to practice. Faculty as well as trainees, in
our experience, benefit from role-playing scenarios. And
although most clinicians dread being videotaped during a
scenario, no other educational method matches direct observation of ones own conversational style and habits to motivate and guide positive change.
Specific Communication Issues Often Encountered
during Advanced Illness
When a childs cancer is no longer curable, the child and

family are likely to experience additional significant treatment or care decisions such as whether to participate in a
phase I trial, implement a do-not-attempt resuscitation
order, or end all treatment efforts and focus on end-of-life
options.15 The child and family members come to end-of-life
treatment decision making affected by exposure to other ill
children and their families who preceded them in reaching
such a decision point, and by their own personal experience
with treatment. Both of these sources of experience produce
profound, lingering images for the child and family and
affect how they engage in end-of-life decision making. It is
therefore helpful to clinicians to directly ask the child and
parents about the treatment experiences they find themselves reflecting on when making end-of-life treatment decisions.
Engagement in end-of-life discussions. A proportion of
these families will initiate the end-of-life discussion because
they have been carefully (and typically, privately) considering what they would do if their ill childs illness became
terminal. A separate proportion of families will not be able to
engage in this kind of discussion because doing so is culturally unacceptable. To do so means that the families must
face the risk of both losing their child and of being excommunicated from their larger community of relatives, friends,
or fellow worshippers for giving up on their child. It is too
much to expect these families to engage in end-of-life decision making. But the majority of families of children whose
disease cannot be cured, particularly those comfortable in
the culture of North America, will expect to be part of
discussions about their ill childs end-of-life care options.16,17
This majority of parents also anticipates having their preferences for end-of-life care honored by the health care
system because they are the childs parent until the childs
last breath. Even though the majority of parents indicate
a preference for involvement, they also readily identify
end-of-life decision making to be the most difficult of all their
care decisions.15
Change over time. One of the striking findings from studies involving parents of children with relapsed, advanced,
and incurable cancer is their self-observation of not being
the same individuals the care team knew at the point of
diagnosis or during cure-oriented treatment. Instead, the
parents describe being more watchful and wary, or being
less inclined to know all of the treatment details and
preferring to focus on only the one most pressing concern at
any given time.18,19 The first implication of this parent
self-observation is that clinicians should anticipate being
surprised at some point during end-of-life decision making
or care by parental responses to the discussion. The second
implication for clinicians is that developing a strategy for
keeping aligned with the parents evolving care preferences
and their expressions of those preferences is beneficial. Such
a strategy can include occasionally and directly asking
parents to describe their self-observed changes and their
current care preferences.
Observing signs of dying. Parents indicate that to be
ready to participate in end-of-life discussions, they must
first come to believe that their child will not recover. Factors
that help parents to believe that their child is going to die
include the information and care recommendations received
from clinicians, and statements that their ill child has made
about not continuing treatment if it became pointless. Parents beliefs are also influenced by their childs physical
changes such as swelling, bleeding, or changes in breathing
patterns, and by concerns about adverse events that could
result from further anticancer treatment.15,20 For these
parents, clinicians can facilitate their participation in decision making by mentioning the childs actual changes, including those physically apparent and those apparent only
on scans, and by interpreting these changes for the parents,
making it clear that all that can be done has been done and
done well.
Parents personal sense of being a good parent. Parent
participation in end-of-life discussions is also influenced by
the parent-named factor of being a good parent.21,22 This
factor represents an internal definition held by parents that
is their personal benchmark of how well they did by their
child before and during end-of-life care. Although there are
aspects of this definition that are unique to each parent,
641
there are commonalities as well. Clinicians may anticipate

these commonalities to include the parents desire to be
certain that their ill child feels loved by the parent, and that
the parent has been a positive role model, and has continued
to provide care and to meet the childs needs, including
protection from suffering and maintaining vestiges of health
for as long as is possible. Finally, parents want to be certain
that they have made prudent decisions on behalf of their
children. Clinician recognition of the existence of this internal definition and inviting parents to share their own
definition of being a good parent to their seriously ill child
will help parents to trust that the end-of-life decisionmaking process includes thoughtful recognition of their
emotional well-being as parents.
Child participation in conversations. We find that not all
parents are comfortable with their child being included in
end-of-life decision making. Limited research supports the
ability of children treated for cancer to participate in such
discussions and, indeed, a preference for inclusion. A careful
exploration by clinicians with the parents regarding their
preferences for including their child is warranted before
initiating actual discussion with the child. In this clinicianparent preparatory discussion, parents can identify the
terms they are comfortable using with their ill child so that
clinicians, parents, and the ill child can use the same
language to facilitate shared understanding of the seriousness of the clinical situation. The process of ask, tell, ask is
very fitting for this preparatory phase of end-of-life discussions.
Although the research evidence is limited, the findings
that have been reported coupled with reported clinical
experience support the ability of children (age 6 and older)
to indicate their preferences for continuing or discontinuing
cure-oriented treatment.23 Additionally, children age 10 and
older have been studied in terms of their stated preferences
for enrolling on a phase I trial, not implementing a resuscitation order, or initiating terminal care.24,25 These children
were able to state their reasons for their end-of-life prefer-
ences, with the most commonly reported factor being concern for loved ones such as family members and favorite
clinicians. Documented fears of children regarding end of life
include having pain and being alone. Assurances from parents and clinicians about being available and determined to
help are of comfort to a seriously ill child.
Treatment team communication and dynamics. It is also
important to involve the childs health care team in end-oflife discussions and treatment decision making. Team tension can emerge when members of the team do not feel
included in decision making. Patients and parents report
being able to sense team tension. Teams that establish time
to discuss end-of-life care options among themselves and to
review child and family preferences related to end-of-life
options may be able to prevent or diminish this kind of
tension.
Conclusion
Clinician communication related to treatment decision

making is a fundamentally important health care intervention and is often reported by parents of seriously ill children
to be the most valued of clinician skills. Words and the
manner in which clinicians convey their concerns about the
ill child and about the family during the periods of diagnosis,
treatment, and transition, devoting themselves to supporting patient and parental decision making, become sources
of substantial influence on child and parent responses to
challenging times. Clinician attention to this powerful skill
set will likely be a source of special and well-remembered
comfort for the child and family.
Acknowledgments
This study was supported in part by Grants No. R21
NR008634, R21 NR010026, and RO1 NR012026 from the National Institute of Nursing Research, Cancer Center Support
Grant No. P30 CA21765 from the National Cancer Institute,
and an American Cancer Society Mentored Research Scholar
Grant.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Jennifer W. Mack*
Chris Feudtner*
Pamela S. Hinds*
REFERENCES
1. Miyaji NT. The power of compassion: truth-telling among American
doctors in the care of dying patients. Soc Sci Med. 1993;36:249-264.
2. Christakis NA, Iwashyna TJ. Attitude and self-reported practice regarding prognostication in a national sample of internists. Arch Intern Med.
1998;158:2389-2395.
3. Mack JW, Wolfe J, Grier HE, et al. Communication about prognosis
between parents and physicians of children with cancer: parent preferences
and the impact of prognostic information. J Clin Oncol. 2006;24:5265-5270.
4. Back AL, Arnold RM. Discussing prognosis: how much do you want to
know? talking to patients who do not want information or who are ambivalent. J Clin Oncol. 2006;24:4214-4217.
5. Back AL, Arnold RM. Discussing prognosis: how much do you want to
know? talking to patients who are prepared for explicit information. J Clin
Oncol. 2006;24:4209-4213.
6. Robinson TM, Alexander SC, Hays M, et al. Patient-oncologist commu-
642
nication in advanced cancer: Predictors of patient perception of prognosis.

Support Care Cancer. 2008;16:1049-1057.
7. The AM, Hak T, Koeter G, et al. Collusion in doctor-patient communication
about imminent death: an ethnographic study. BMJ. 2000;321:1376-1381.
8. Cassel EJ. The nature of suffering and the goals of medicine. N Engl
J Med. 1982;306:639-645.
9. Wolfe J, Klar N, Grier HE, et al. Understanding of prognosis among
parents of children who died of cancer: impact on treatment goals and
integration of palliative care. JAMA. 2000;284:2469-2475.
10. Feudtner C. Collaborative communication in pediatric palliative care: a
foundation for problem-solving and decision-making. Pediatr Clin North Am.
2007;54:583-607, ix.
11. Feudtner C, Carroll KW, Hexem KR, et al. Parental hopeful patterns of
thinking, emotions, and pediatric palliative care decision making: a prospective cohort study. Arch Pediatr Adolesc Med. 2010;164:831-839.

12. Stone D, Patton B, Heen S. Difficult Conversations: How to Discuss
What Matters Most. New York: Viking; 1999.
13. Feudtner C. Partnering leadership. Arch Pediatr Adolesc Med. 2011;
165:487-488.
14. Feudtner C. The breadth of hopes. N Engl J Med. 2009;361:23062307.
15. Hinds PS, Oakes L, Furman W, et al. Decision making by parents and
healthcare professionals when considering continued care for pediatric patients with cancer. Oncol Nurs Forum. 1997;24:1523-1528.
16. Hinds PS, Oakes L, Quargnenti A, et al. An international feasibility
study of parental decision making in pediatric oncology. Oncol Nurs Forum.
2000;27:1233-1243.
17. Michelson KN, Koogler TK, Skipton K, et al. Parents reactions to
participating in interviews about end-of-life decision making. J Palliat Med.
2006;9:1329-1338.
18. Hinds PS, Birenbaum LK, Clarke-Steffen L, et al. Coming to terms:
parents response to a first cancer recurrence in their child. Nurs Res.
1996;45:148-153.
19. Hinds PS, Birenbaum LK, Pedrosa AM, et al. Guidelines for the
recurrence of pediatric cancer. Semin Oncol Nurs. 2002;18:50-59.
20. Hinds PS OL, Furman W. End of life decision-making in pediatric
oncology. In: Ferrell B CN (ed). Oxford Textbook of Palliative Nursing Care. 2d
ed. New York: Oxford University Press, 2006;945-958.
21. Hinds PS, Oakes LL, Hicks J, et al. Trying to be a good parent as
defined by interviews with parents who made phase I, terminal care, and
resuscitation decisions for their children. J Clin Oncol. 2009;27:5979-5985.
22. Maurer SH, Hinds PS, Spunt SL, et al. Decision making by parents
of children with incurable cancer who opt for enrollment on a phase I trial
compared with choosing a do not resuscitate/terminal care option. J Clin
Oncol. 2010;28:3292-3298.
23. Nitschke R, Humphrey GB, Sexauer CL, et al. Therapeutic choices
made by patients with end-stage cancer. J Pediatr. 1982;101:471-476.
24. Hinds PS, Drew D, Oakes LL, et al. End-of-life care preferences of
pediatric patients with cancer. J Clin Oncol. 2005;23:9146-9154.
25. Freyer DR. Care of the dying adolescent: special considerations. Pediatrics. 2004;113:381-388.
643
ADDRESSING THE IMBALANCE OF SUPPLY AND

DEMAND: INTEGRATING ADVANCED PRACTICE
PROVIDERS INTO SURVIVORSHIP CARE
CHAIR
Michael Goldstein, MD
Beth Israel Deaconess Medical Center and Harvard Medical School
Boston, MA
SPEAKERS
James Ross Waisman, MD
Breastlink Medical Group, Inc.
Todd Alan Pickard, PA-C, MMSc
University of Texas M. D. Anderson Medical Center
Houston, TX
Mary S. McCabe, RN, MA
New York, NY
Planning for the Future: The Role of Nurse

Practitioners and Physician Assistants in
Survivorship Care
By Mary S. McCabe, RN, MA, and Todd Alan Pickard, PA-C, MMSc
Overview: The number of cancer survivors in the United

States now approaches 12 million individuals, with an estimated 7.2% of the general population aged 18 years or older
reporting a previous cancer diagnosis. These figures highlight
a number of questions about the care of survivors how
patients at risk for a known set of health problems should be
followed, by whom, and for how long. At the same time that
oncologists are developing strategies to provide services to
this growing population, there are economic and systems
challenges that have relevance to the previous questions,
including a predicted national shortage of physicians to provide oncology services. Nurse practitioners (NPs) and physician assistants (PAs) have been identified as members of the
S A RESULT of the ongoing advances in early detection and treatment, the number of cancer survivors in
the United States now approaches 12 million individuals,
with an estimated 7.2% of the general population aged 18
years or older reporting a previous cancer diagnosis.1,2 The
5-year relative survival rate for adult cancer survivors has
reached 67%, with the largest number of survivors having
been treated for breast, prostate, and colorectal cancers.2
These optimistic figures also highlight the challenges and
opportunities facing the oncology community as we work to
make survivorship a formal period of care. Such a focus
requires not only appropriate surveillance for recurrence,
but also the comprehensive rehabilitation of the posttreatment patient: 1) follow-up medical care tailored to the
problems of specific populations, 2) psychosocial support,
and 3) health promotion education that includes diet, exercise, and screening for new primary cancers. These core
services are coupled together with the imperative that we
share the care of these individuals with their community
primary care physician (PCP), thus assuring a coordination
of care that leads to effective communication and improved
quality of life for cancer survivors (Sidebar 1).
In 2005, the Institute of Medicine issued a seminal report,
From Cancer Patient to Cancer Survivor: Lost in Transition,
which served as an early guide to the development of
survivorship-specific services and models of care.3 It was the
first comprehensive proposal detailing the follow-up care
needs of survivors and the novel provider arrangements that
could be implemented to provide these services. This report,
and others that followed, refocused our thinking about how
survivorship care can and should be delivered. Initially,
survivorship research focused on the long-term consequences of cancer therapy in the pediatric cancer survivor,
in particular the serious morbidity and premature mortality
resulting from exposure to radiation and chemotherapeutic
agents in developing organs and normal tissue. And now,
over the past 10 years, research has increasingly focused on
the long-term and late effects experienced by individuals
treated for adult-onset cancers as well.4,5,6 Traditionally,
the follow-up care of the survivor of an adult-onset cancer
was primarily focused on an evaluation of recurrence, but
this single focus is insufficient. We now know that survivors
e56
health care team who can help reduce the oncology supply and
demand gap in a number of ways. The ASCO Study of Collaborative Practice Arrangements (SCPA) in 2011 concluded that
oncology patients were aware and satisfied when their care
was provided by NPs and PAs; there was an increase in
productivity in practices that utilized NPs and PAs; utilizing
the full scope of practice of NPs and PAs was financially
advantageous; and, physicians, NPs, and PAs are highly satisfied with their collaborative practices. Increasingly, the
oncology and health policy literature contains evidence supporting innovative provider models. There is still much work to
be done to move beyond pilot data to establish the true value
of these models.
face a variety of health risks that are dependent on treatment exposures, genetic predisposition, comorbid health
conditions, and lifestyle behaviors and that the identified
issues cross multiple domains, including the medical, psychological, and social.7,8,9,10 Because breast cancer survivors
are the most widely studied group to date, they serve as an
excellent example of the range of late effects a survivor can
face. Medical late effects include anthracycline-related cardiomyopathy, osteoporosis, cognitive dysfunction, and infertility; psychological effects include fear of recurrence, sexual
dysfunction, and fatigue; and, social issues relate to return
to work and changes in role functioning. These examples
raise the question of how patients at risk for a known set of
health problems should be followed, by whom, and for how
long. At the same time that oncologists are developing
strategies to provide services to this growing population
formally extending oncology care through the survivorship
periodthere are other economic and systems challenges
that have relevance to the previous questions.
Drivers of Workforce Change
There are a number of factors in the U.S. health care

system today that will affect the ability for oncology physicians to provide care for patients with cancer. The growing
population of individuals over the age of 65 with an increasing incidence of cancer is a significant driver for an increasing demand for oncology care.11 The improvement in cancer
therapies and the resulting reduction in mortality rates
mean that there will be an increase in the number of years
that a patient will live with cancer through all phases of
the disease. Essentially, there will be a larger population of
individuals diagnosed with cancer, and they are likely to live
longer with the disease. This drives the increased demand
From the Memorial Sloan-Kettering Cancer Center and University of Texas M. D.

Anderson Cancer Center.
Address reprint requests Mary S. McCabe, 1275 York Avenue, New York, NY 10065;
email: mccabem@mskcc.org.
1092-9118/10/1-10
NPS AND PAS IN SURVIVORSHIP CARE
Sidebar 1. Essential Components of Survivorship Care
Surveillance for recurrence

Screening for new cancers
Identification and management of the consequences
of the cancer and its treatment
Health promotion strategies
Coordination/communication between oncology specialists and primary care providers
for oncology services up and will continue to do so for the

foreseeable future.12
The number of physicians entering the field of oncology
has decreased, and the supply of oncologists will not be able
to meet the demand for oncology services by 2020. In fact,
there is a predicted shortage of more than 4,000 physicians
to provide oncology services. Several factors for this gap
have been identified and include fewer residents entering
oncology training programs, a growing population of
retirement-age oncologists, a shift in the health care system
to a focus on primary care, and decreasing reimbursement
for chemotherapy infusion and other services that oncology
practices require to remain fiscally sound. The American
Society of Clinical Oncology (ASCO) Workforce Study completed in 2007 identified the gap between supply of oncology
physicians and demand for oncology services. There will be
an estimated demand of 57.2 to 60.7 million visits per year
by 2020 and only an estimated supply of 45.6 to 47.8 million
visits per year. Thus, there is a gap of 11.6 to 12.9 million
visits per year by 2020 (Fig. 1). This study highlighted the
fact that the increasing gap in supply and demand requires
new ways of providing oncology services such as increasing
oncology training programs to produce more oncologists and
shifting services traditionally provided by oncologists, such
as survivorship, to the primary care setting.13
Nurse practitioners (NPs) and physician assistants (PAs)
have been identified as members of the health care team
who can help reduce the oncology supply and demand gap in
a number of ways.14 Fortunately, there is already a significant number of NPs and PAs practicing in oncology settings
working with physicians to provide diagnostic evaluation,
patient education, infusion services, treatment monitoring,
KEY POINTS
There is an increasing number of survivors who have

been successfully treated for cancer.
The number of oncologists will not be able to meet the
demand for oncology services by 2020.
Nurse practitioners (NP) and physician assistants
(PA) are professionally prepared to help reduce the
gap in care.
Innovative care models are being implemented internationally using NPs and PAs as important colleagues in the care of cancer survivors.
Additional research is needed to evaluate the most
efficient, highest quality care models in a variety of
settings and health systems.
surveillance, and symptom management.15-20 The ASCO

Workforce Study identified that NPs and PAs can extend the
services that oncology physicians provide and allow physicians to focus on evaluating new patients, creating treatment plans, and addressing changes in patient condition.
NPs and PAs can free up oncologists from the more routine
needs of patients with cancer so that they can focus on more
complex care.21 The Workforce Study also identified transitioning survivorship care from oncology physicians to the
primary care setting as a means to close the gap in supply
and demand. It was estimated that a reduction of 10% to
20% in the number of patients being seen by an oncologist
during the monitoring phase could be achieved.13 Although
the Workforce Study focused on transitioning survivorship
care from oncologists to primary care physicians, it can be
argued that NPs and PAs can also provide this care. Many
cancer centers and oncology practices have established survivorship programs that utilize NPs and PAs. These programs have enabled oncologists to shift post-treatment care
to other members of the team while keeping the patients
within the institution or practice.22
Drivers of Collaborative Practice
After the ASCO Workforce Study was completed, it was

apparent that NPs and PAs would be an important part of
the solution to close the gap in supply and demand. However, there has been a limited understanding of the NP and
PA workforce in oncology as well as the scope of services they
provide, their effect on productivity, patient satisfaction,
and physician satisfaction in the utilization of NPs and PAs.
To address these issues, ASCO released the results for a
Study of Collaborative Practice Arrangements (SCPA) in
2011. The study had five main conclusions: oncology patients
were aware and satisfied when the care they received was
provided by NPs and PAs; there was an increase in productivity in practices that utilized NPs and PAs; utilizing the
full scope of practice of NPs and PAs was financially advantageous and was a main driver of the model of collaborative
practice chosen; perceptions of workload for oncologists,
NPs, and PAs did not correlate to objective measures of work
production; and physicians, NPs, and PAs are highly satisfied with their collaborative practices.
Another interesting result of the SCPA was the list of
services provided by NPs and PAs in oncology practices. The
top three services as identified by more than 80% of the
practices surveyed were assessing patients during treatment visits, pain and symptom management, and follow-up
care for patients in remission. The bottom three services
with 20% or less of practices indicating these services were
NPs and PAs providing night or weekend call, survivorship
clinics, and other services. What is striking about these
results is that NPs and PAs provided a large portion of
follow-up care for patients in remission, but this was not
within a formal survivorship program. This seems to indicate that NPs and PAs have the skills needed to take care of
cancer survivors but that they are not yet doing so within
structured programs.
There have been three main collaborative practice models
described between oncology physicians, NPs, and PAs. Each
model has a different level of physician interaction with the
patient and has a different effect on productivity and patient
volume. The first model, incident-to-practice, is defined as
autonomous practice of an NP/PA while the physician is
e57
MCCABE AND PICKARD
Fig. 1. Growth of gap between supply

and demand for oncologist visits, 2005 to
2020.13
present in the office suite but does not routinely see patients.
This practice model allows for the physician and NP/PA to
work in parallel, thus increasing productivity and patient
volume while allowing the physician to be available for
patient care as needed. This model was the most prevalent
in the SCPA. The second model, shared practice, is defined
as physician and NP/PA care that is given together. The
physician always sees the patient during the clinical visit.
This model has a modest effect on productivity and patient
volume. This model was the second most prevalent in the
SCPA. The third model, independent practice, is defined as
NP/PA care that is provided to a separate panel of patients
without physician participation or presence. This model
increased productivity and patient volume, but it is generally reimbursed at 85% of physician rates. This model was
the least prevalent in the SCPA.
NPs and PAs: Who Are They?
Acknowledging that oncology has always been a multidisciplinary specialty, it is not difficult to ask the question of
who is the appropriate provider for various groups of survivors, based on a risk model of recurrence and late effects. As
we frequently see, NPs and PAs can and are increasingly
filling this role.
NPs are licensed advanced practice nurses who have
training as registered nurses before taking postgraduate
training. Their education programs include degree-granting
and postgraduate education programs (masters or doctoral
degree) and are accredited. To become licensed/certified as
an NP, individuals must pass a certification examination
that assesses national competencies of the overall role and
at least one specialty area of practice, such as adultgerontology or pediatrics. NPs may also specialize in oncology, and the education for this expertise is supplemental
to the core curriculum. Overall, the NP curriculum is broadbased and includes three separate level graduate-level
courses in advanced physiology/pathophysiology, heath assessment, and pharmacology, as well as a variety of clinical
experiences. Extensive clinical training provides the graduate with strong physical assessment skills and focuses on a
comprehensive approach to patient care with an emphasis
e58
on health promotion. This expertise is a strong complement

to the skills of both the physician oncologist and the primary
care physician. In support of the NP role as a key member of
the survivorship care team, the 2011 Institute of Medicine
Report, The Future of Nursing: Leading Change, Advancing
Health, highlights that nurses should practice to the full
extent of their education and trainingsomething the survivorship provider role offers.23 A second key message in
this report is that nurses should be full partners with
physicians and other health professionals in redesigning
health care in the United States. This is in synch with the
ASCO SCPA that calls for the greater inclusion of NPs in
oncology practice models and reports that such models
result in increasing productivity for the practice and provide
high physician and NP satisfaction, as well as satisfied
patients.
First proposed in the mid 1960s, PAs are professional
members of the health care team who are trained in the
medical model. They practice in physician-PA teams
throughout the United States, Canada, the United Kingdom, Australia, and the Netherlands. The PA educational
model includes a general medical education for a period of
24 28 months (there are a handful of programs lasting
36 months). The first year is spent in didactic training in
gross anatomy, pathophysiology, medical terminology, microbiology, pharmacology, principals and practices of medicine, history and physical examination skills, bedside
procedures, and major diseases of all body systems. The
second year is spent in clinical rotations in pediatrics,
internal medicine, family practice, emergency medicine,
general surgery, obstetrics and gynecology, and a variety of
electives. The remaining coursework in the last semester of
training includes community health, epidemiology, biostatistics, professional practice, and medical research. The
masters degree is currently the terminal degree for the PA
profession with a few programs offering a clinical doctorate.
There are a significant number of PA programs housed
within medical schools where PA students and medical
students take basic science courses together. Most PA students perform clinical rotations side by side with medical
students.
The educational curriculum and certification for the PA

profession is based on a single national standard. Education
accreditation for all PA programs is through the Accreditation Review Commission on Education for the Physician
Assistant (ARC-PA). The ARC-PA protects the interests of
the public and PA profession by defining the standards for
PA education and evaluating PA educational programs
within the territorial United States to ensure their compliance with those standards. Graduation from an ARC-PA
accredited PA program is a requirement for licensure in
every state. There is a single national certification examination through the National Commission on the Certification
of Physician Assistants (NCCPA) and is a requirement for
licensure in every state.
Practice as a PA is collaborative by definition. The profession has defined itself and is based on a commitment to
working with physicians in collaborative teams.24 PAs do
not practice medicine independent of their relationship with
physicians. However, autonomous practice is very common,
as physicians are not required to be present when services
are provided by PAs. Prescriptive authority is granted to
PAs in all 50 states.
Integrating NP and PA Providers into
Survivorship Programs
The professional training of both NPs and PAs lends itself

very well to the comprehensive, holistic care proposed for
survivors. This skill set allows them to provide the multiple
services needed in survivorship care including age-specific
cancer screening, general wellness, disease prevention, patient and caregiver counseling, nutrition counseling, and
psychosocial assessment. Because many NPs and PAs have
demonstrated the ability to develop expertise in oncology
and are involved in the care of patients with cancer in active
treatment, it will be important to assess the value of the
inclusion of NPs and PAs as providers of survivorship care
and determine whether such care can be transitioned immediately post-treatment to primary care providers as some
have suggested.25,26,27 And although NPs and PAs are
increasingly taking on central roles in survivorship care, the
question remains whether oncologists, patients, and payors
will embrace this model nationally.
An important issue related to the transition of survivorship care to other providers is the establishment of survivorship eligibility criteria and surveillance guidelines. Such
guidance is a foundational step toward determining which
patients are appropriate for survivorship care and when
they can be transitioned from the oncologist to an NP or PA.
Having clearly defined eligibility criteria based on stage of
disease and risk of recurrence offers a consistent approach
and important direction for those professionals providing
survivorship care, as well as cancer survivors themselves. In
addition, guidance about the identification and management
of long-term toxicities and late effects of cancer treatments
are critical to the effective transition to other providers.
Although limited guidance currently exists for patients with
adult onset cancers, groups such as ASCO, the American
Cancer Society, and the National Comprehensive Cancer
Network have all convened national groups of experts to
address this important need.
Although NPs and PAs have the training and experience
to provide quality care to cancer survivors, it is important to
realize that there a number of factors that should be ad-
dressed before proceeding. Beyond the need for direction

from oncologists through survivorship eligibility criteria and
survivorship algorithms, there are basic regulatory considerations and logistics to consider. Unlike physicians, who
enjoy a relatively stable scope of practice from state to state,
the legal scope of practice for NPs and PAs can have
significant variation. Some states allow broad prescriptive
authority and other states tightly regulate this authority
such that NPs and PAs cannot prescribe certain medications. The definition of physician supervision, collaboration,
and delegation can also vary from state to state. This is
important when determining if physicians have to be physically present or review a certain percentage of charts.
Understanding the practice act for NPs and PAs in each
state is crucial to ensure that they have the support from
physicians they need as well as a defined operational structure in any survivorship program.28 These logistical issues
are not insurmountable nor are they prohibitive to incorporating NPs and PAs into survivorship programs. They are
simply issues that need to be addressed to create an effective
and efficient survivorship program. NPs and PAs have the
training, medical knowledge, and skills to provide survivorship care; they are committed to collaborative relationships with physicians; and they have a demonstrated role in
oncology with a commitment to lifelong learning.29,21
Models of Survivorship Care
Internationally, as oncology physicians in hospitals, clinics, and private practices begin to develop more efficient
approaches to the follow up of cancer survivors, they increasingly look to the incorporation of NPs and PAs as partners.
There are well-established practice arrangements based on
the Wagner Chronic Care Model from which to choose, as
well as the multidisciplinary care model that has long been
used in the care of pediatric cancer survivors.30 The choice of
model is most often based on the complexity of the patient
(risk for long-term and late effects) and the optimal financial
reimbursement arrangement for the medical group. The
Shared Survivorship Visit is common and has evolved into
two types of survivorship care based on the diagnosis of the
survivors being seen. First, is the multidisciplinary model,
where survivors with a variety of diagnoses are cared for,
and the survivor is seen by both the physician and NP or PA,
with each provider focusing on particular aspects of the visit.
This model is very useful for complicated patients, and in
the general medicine experience, studies have shown that
NP/PAs spend more time counseling the patient than if the
patient were seen only by the physician. The second type of
shared visit is the disease-specific model in which the
physician-NP/PA team focuses on one disease, such as
breast cancer survivors. Many groups start with this type of
care model when setting up formal survivorship services as
a way to pilot the sharing of care among providers and to
assess the financial returns. Although the billing specifics
may differ by state and payor, both types of shared visit
models offer the opportunity for joint billing in the physicians name. The second category of visit is the Independent
Survivorship Visit, which includes two types of care models
where the NP or PA sees the survivors independently of the
physician. The first type of independent visit is the consultative model; this common approach is easily established as
a one-time or annual visit with an NP or PA during which
general survivorship issues are discussed and a Treatment
e59
MCCABE AND PICKARD
Sidebar 2. Models of Survivorship Care: Physician

Nurse Practitioner/Physician Assistant
Shared-Visit Model
Multidisciplinary clinic
Disease-specific clinic
Independent Visit Model

Consultative clinic
Integrated clinic
Summary and Care Plan are provided. When specific issues

are identified, the NP or PA makes referrals to appropriate
specialists, such as medical subspecialists, physical therapists, or psychiatry. Although one advantage of this model is
that the provider can serve an unrestricted survivor population, it can be difficult to be expert in the long-term
follow-up issues of many types of survivors. Another independent visit type that is becoming increasingly adopted is
the integrated model, where the NP or PA is imbedded with
the treatment team, but sees the patient independently.
The strength of this model is that it frees up the treating
physician and allows the patient to continue to be seen in a
familiar location with familiar staff, providing continuity of
care because of the proximity and relationship among providers. Both types of independent care models require a
cultural change for some oncology physician groups, but as
discussed in the ASCO workforce report, there are considerable pressures to encourage change in how care is delivered (Sidebar 2).31
Future Directions in Caring for Survivors
Increasingly, the oncology and health policy literature

contains papers providing support for innovative provider
models for the follow-up care of cancer survivors. Coupled

with these early study results that show equivalent outcomes and excellent patient satisfaction when survivors
receive care from NPs and PAs, are the economic and quality
care incentives for change. Oncology has long been a multidisciplinary specialty, and reconfiguring the responsibilities
of the team members can and should result in better, more
efficient care at a time when the number of cancer survivors
is rapidly increasing. Such new models focus on the key
elements of a tailored visit utilizing a risk-based approach to
care that allows the specific needs of the survivor to be
addressed. In addition, such models all incorporate a communication link with the community primary care provider.
If we are to overcome the barriers to shared-care between
specialists and generalists, we need to incorporate a bidirectional communication plan that designates which provider is
responsible for what specific medical services. This sharing
of care has long been used for other disease groups such as
patients with cardiac disease and diabetes. It is time for us
to move out of our subspecialty isolation and assure coordinated care for our survivors. In addition, using NPs and PAs
as a communication link can assist with the implementation
of risk-based care and the transition of survivors back to the
PCP when they are at low risk of recurrence and low risk of
late effects, not needing cancer follow-up. There is still much
work to be done to move beyond pilot data and clinical
hunches about the value of these models. We need to
evaluate the patient health outcomes and the financial
efficiency of these new models as well. Just as oncology has
been a leader in many aspects of health care, it is an
opportunity and necessity to do so in this important field of
cancer survivorship. We owe it to our patients to assure
them the highest quality of life possible after cancer.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Mary S. McCabe*
Todd Alan Pickard
CEMINES
CEMINES (I)
REFERENCES
1. Rowland JH. Cancer Survivorship: Rethinking the cancer control continuum. Semin Oncol Nurs. 2008;24:145-152.
2. Underwood JM, Townsend JS, Stewart ST, et al. Surveillance of demographic characteristics and health behaviors among adult cancer survivors behavioral risk factor surveillance system, United States, 2009 CDC Surveillance Summaries. 2012;61(SS01):1-23.
3. Hewitt J, Greenfield S, Stovall E. (eds). From cancer patient to cancer
survivor: Lost in transition. Washington, DC: National Academies Press; 2005.
4. Bhatia S. Role of genetic susceptibility in development of treatmentrelated adverse outcomes in cancer survivors. Cancer Epidemiol Biomarkers
Prev. 2011;20:2048-2067.
5. Carmack CL, Basen-Engquist K, Gritz ER. Survivors at higher risk for
adverse late outcomes due to psychosocial and behavioral risk factors. Cancer
Epidemiol Biomarkers Prev. 2011;20:2068-2077.
6. Oeffinger KC, Tonorezos ES. The cancer is over, now what? Cancer.
2011;117(10 suppl):2250-2257.
7. Miller KD, Triano LR. Medical issues in cancer survivors - a review.
Cancer. 2008;14:375-387.
8. Oeffinger KC, Mertens AC, Sklar CA, et al. Chronic health conditions in
adult survivors of pediatric cancer. N Engl J Med. 2006;355:1572-1582.
9. Rao AO, Demark-Wahnefried W. The older cancer survivor. Crit Rev Onc
Hematol. 2006;60:131-143.
e60
10. Holland J, Weiss T. The new standard of quality cancer care: Integrating the psychosocial aspects in routine cancer from diagnosis through survivorship. Cancer. 2008;14:425-428.
11. Parry C, Kent EE, Mariotto AE, et al. Cancer survivors: A booming
population. Cancer Epidemiol Biomarkers Prev. 2011;20:1996-2005.
12. Bunnell CA, Shulman LN. Will we be able to care for cancer patients in
the future? Oncology. 2010;24:1343-1352.
13. Erikson C, Salsberg E, Forte G, et al. Future supply and demand for
oncologists. J Oncol Pract. 2007;3:79-86.
14. Street D, Cossman JS. Does familiarity breed respect? physician
attitudes toward nurse practitioners in a medically underserved state. J Am
Acad Nurse Prac. 2010;22:431-439.
15. American Academy of Physician Assistants. Specialty Practice: PAs in
Oncology. March 2011:1-4.
16. Buswell LA, Ponte PR, Shulman LN. Provider practice models in
ambulatory oncology practice: Analysis of productivity, revenue, and provider
and patient satisfaction. J Oncol Pract. 2009;5:188-192.
17. Friese CR, Hawley ST, Griggs JJ, et al. Employment of nurse practitioners and physician assistants in breast cancer care. J Oncol Pract.
2010;6:312-316.
18. Hinkel JM, Vandergrift JL, Perkel SJ, et al. Practice and productivity
of physician assistants and nurse practitioners in outpatient oncology clinics

at national comprehensive cancer network institutions. J Oncol Pract. 2010;
6:182-187.
19. Hylton H. Clinical partnership: The role of physician assistants in
oncology practice. ASCO Daily News, June 4, 2011:21B.
20. Ross AC, Polansky MN, Parker PA, et al. Understanding the role of
physician assistants in oncology. J Oncol Pract. 2010;6:26-30.
21. Polansky M. Strategies for workplace learning used by entry-level
physician assistants. J Physician Assist Educ. 2011;22:43-50.
22. Towle EL, Barr TR, Goldstein M, et al. Results of the ASCO study of
collaborative practice arrangements. J Oncol Pract. 2011;7:278-282.
23. IOM (Institutes of Medicine). The future of nursing: Leading change,
advancing health. Washington, DC: The National Academies Press; 2011.
24. American Academy of Physician Assistants. Professional Issues: The
Physician-PA Team. October 2011; 1-4.
25. Watts SA, Gee J, ODay ME, et al. Nurse practitioner-led multidisciplinary teams to improve chronic illness care: The unique strengths of nurse
practitioners applied to shared medical appointments/group visits. J Am Acad

Nurse Pract. 2009;21:167-172.
26. Lewis R, Neal RD, Williams NH, et al. Nurse-led vs. conventional
physician-led follow-up for patients with cancer: Systematic review. J Adv
Nurs. 2009;65:706-723.
27. Coniglio D, Pickard T, Wei S. Commentary: Physician Assistant Perspective on the Results of the ASCO Study of Collaborative Practice Arrangements. J Oncol Pract. 2011;7:283-284.
28. Gage EA, Pailler M, Zevon MA, et al. Structuring survivorship care:
Discipline-specific clinician perspectives. J Cancer Surv. 2011;5:217-225.
29. Korber K. How do NPs and PAs get information and learn? Adv NPs
PAs. 2011;2:14.
30. Wagner E. Chronic disease management; what will it take to improve
care for chronic illness? Effective Clinical Practice. 1998;1:2-4.
31. Gilbert SM, Miller DC, Hollenbeck BK, et al. Cancer survivorship:
Challenges and changing paradigms. J Urol. 2008;179:431-438.
e61
DOING IT RIGHT, AND FOR LESS: IMPLEMENTING

PRACTICE CHANGES TO MANAGE THE GROWING
COMPLEXITIES, INEFFICIENCIES, AND
COSTS OF CANCER CARE
CHAIR
Adam Brufsky, MD, PhD
University of Pittsburgh School of Medicine
Pittsburgh, PA
SPEAKERS
Bruce E. Hillner, MD
Virginia Commonwealth University
Richmond, VA
Henry Orlando Otero, MD
Virginia Mason Medical Center
Seattle, WA
Driving Evidence-Based Standardization of

Care within a Framework of Personalized
Medicine
By Adam Brufsky, MD, PhD, Kathleen Lokay, BBA, and Melissa McDonald, MS
Overview: Cancer care in the United States faces a number of

key challenges today that are causing payers, referring physicians, and patients alike to question the value of the care, in
terms of both outcomes and costs. New technologies in the
form of pharmaceuticals and biologics, prognostic tests, and
new radiation therapy tools and techniques offer the promise
of improved outcomes, but their cost-effectiveness is often
unclear. Oncologists themselves are caught in the middle, as
they are the prescribers of such technologies and the entity
billing for such services but have only limited ability to
influence the pricing models for these services. Finally, as
oncology care becomes more complex because of increased

understanding of the pathogenesis of the many subtypes of
cancer, the community-based oncologist who cares for patients
with all cancer subtypes is confronted with maintaining an
up-to-date knowledge base that is expanding rapidly. Although
no single solution for these issues exists in oncology today,
the experience at the University of Pittsburgh Medical Center
has demonstrated that a clinical pathways program can reduce unwarranted variability in both treatment and outcomes,
drive adherence to evidence-based medicine, and, in the process, reduce the growth rate in the total costs of cancer care.
benefit from a certain level of standardization for the majority of clinical presentations.
Y 2004, the University of Pittsburgh Medical Center

(UPMC) Cancer Centers had expanded to approximately 40 sites of services in a 100-mile radius in Western
Pennsylvania. Concerns over the consistency and quality of
care across such a diverse network were validated through
the results of internal surveys that demonstrated wide variability in approaches to cancer care. Pressures and concerns
from the large payers in the region were also driving an
imperative to collaborate around a solution to containing the
rising costs. In addition, with the University of Pittsburgh
Cancer Institute (UPCI) as its NCI-designated cancer center, UPMC needed tools for increasing awareness of, and
accrual to, clinical trials. The solution for all of these needs
was the development and implementation of the Via Oncology Pathways program to improve quality, ensure consistency of care, reduce hospital admissions, and reduce total
cost of care. To date, the program has served UPMC well for
more than 7 years and provides a firm foundation for UPMCs
overall health care reform strategy for accountable care.
Differences between Via Oncology Pathways and
Oncology Guidelines
Although excellent sources of oncology guidelines exist

today, adherence to these guidelines is more difficult to
assess, especially in community-based oncology practices,
where more than 80% of cancer care is delivered. A recent
analysis by IntrinsiQ of its data set of more than 17,000
patients per month suggests that, for example, for non
small cell lung cancer, adherence to guidelines is 100% for
first-line therapy but only 60% for second- and third-line
therapies (Ed Kissell, IntrinsiQ, email communication, September 2009). Even within guidelines, case studies routinely
collected through physician surveys by reputable third parties suggest that a substantial amount of unexplained variability exists, in large part due to the inclusive nature of
multiple options as standards of care. Case study surveys
conducted in 2005 within the UPMC Cancer Centers
(UPMC-CC) revealed a high amount of variability within
guidelines that could not be easily explained. No one will
dispute that cancer is a very complex disease or that the
nature of decision making is highly dependent on physician
judgment for each unique patient. However, experiences in
other fields of business suggest that oncology care will
e62
Development and Maintenance of Via

Oncology Pathways
Starting in 2005, UPMC developed and maintained algorithms (Via Oncology Pathways) for oncology clinical decision making that UPMC physicians use to determine the
best management for any given state and stage of cancer.
This effort has been painstaking and has involved the
cooperation of the majority of the academic and clinical
experts at UPMC as well as numerous physicians from other
practices. A committee exists for each major disease category (eg, colorectal) and consists of two chairpersons: one
academically based oncologist specializing in that disease
and one community-based oncologist with a background and
patient concentration in that disease. The disease committees are composed of practicing oncologists from UPMC as
well as all other practices that use the Via Oncology Pathways program. The committees convene quarterly to review
new clinical literature, pathway results, and the appropriateness of the granularity of the algorithms (eg, defining
the states and stages of disease and unique patient scenarios
at which decisions should be made). Through their collaborative work, a single best treatment for the majority of
clinical scenarios in oncology care is defined. These best
treatments are based on reviewing the literature in a consistent decision hierarchy. First, the committees look for the
most effective treatment based on the current literature. In
cases where there is a single best (i.e., most effective)
treatment, that becomes the Via Oncology Pathway for that
case. However, if there is more than one treatment with
comparable efficacy, then the committees look for the least
toxic therapy with the goal of maximizing patient quality of
life and outcomes and minimizing unnecessary costs of
From the University of Pittsburgh, Pittsburgh, PA; D3 Oncology Solutions (Via Oncology
Pathways), an affiliate of the University of Pittsburgh Medical Center, Pittsburgh, PA.
Address reprint requests Adam Brufsky, MD, PhD, Division of Hematology/Oncology,
University of Pittsburgh Cancer Center, 300 Halket St, Pittsburgh, PA 15213; email:
brufskyam@msx.upmc.edu.
1092-9118/10/1-10
CLINICAL PATHWAYS STANDARDIZING PERSONALIZED MEDICINE
toxicity management, such as hospitalizations. Finally, if there

is more than one treatment with comparable efficacy and
toxicity, the committees look for the least costly alternative to
reduce unnecessary health care expenditures without compromising clinical benefit. In addition to a single best therapy
as the primary pathway, the committees include options for
common scenarios, such as neuropathy, poor performance
status, drug shortages, etc. However, such options are only
presented and counted as on pathway when the physician
notes the specific scenarios. For those patient scenarios not
addressed by the pathway, it is anticipated and expected that
physicians will choose a treatment off pathway. There is
no penalty for such decisions and, in fact, adherence rates
approaching 100% would cause concern. The expectation of
the Via Oncology Pathways programs disease committees is
adherence rates in the 70% to 90% range overall.
Imbedding Data-Driven Personalized Medicine within
the Via Oncology Pathways
As part of the quarterly process for the Via Oncology

Pathways, the disease committees review not only data
regarding alternative treatments but also biomarkers and
other prognostic testing to drive to personalized medicine
where the data are appropriate. The same level of rigor is
applied to all alternatives with the additional requirement
placed on biomarkers and prognostic tests of whether the
results actually drive care decisions. If the data are robust
and show a positive effect on care decisions, the pathways
are updated to include these tests as recommendations (eg,
anaplastic lymphoma kinase translocationpositive drives
to crizotinib, epidermal growth factor receptor mutation
drives to erlotinib, etc). For those tests without adequate
data or firm clinical relevance, the committees develop
literature-based explanations to be included in the pathways to discourage the ordering of these high-cost tests. As
a result, the Via Oncology Pathways drive evidence-based
personalized medicine only when it has been shown to make
a difference in patient care.
Physician Engagement Strategies
A number of strategies have been successful in engaging

oncologists both at UPMC and in other markets in the
adoption of and adherence to the Via Oncology Pathways.

Key to their support is an understanding of the realities of
the rising costs of cancer care and the likely changes that
will be imposed by payers if no alternative solutions are
proposed. Beyond these compelling reasons, other factors
that drive physician acceptance include (1) creating an open
and transparent disease committee process that allows
participation by all physicians, (2) allowing unique patient
scenarios to be addressed in the pathways, and (3) emphasizing that treating off-pathway is not a negative outcome
but an expected one because of the nature of the unique
patient scenarios encountered daily.
Decision Support Tools Are Critical for the Use and
Measurement of Clinical Pathways
The development and maintenance of the clinical content

of the Via Oncology Pathways are certainly the most critical
components in reducing unwarranted variability and adhering to evidence-based medicine. However, without tools to
either deliver such content to the oncologists or measure
their adherence to the pathways, the clinical content alone is
no more valuable than other online resources or reference
textbooks. Within UPMC, the need for such tools became
evident quickly after the development of the clinical pathways and the implementation of a cumbersome paper-based
process. Substantial investment was made by UPMC in
developing a Web-based software application for the delivery of the Via Oncology Pathways via a decision-support tool
to the physicians. The delivery format is patient-specific and
provided in real time. The Via Oncology Pathways are
navigated through a question-and-answer format where the
critical questions that drive that disease-specific pathway
are presented and the physician is navigated to the appropriate branch of the decision tree based on his or her
response. At the end of each node of the decision tree, local
clinical trial options are presented, followed by the pathway
treatment option including the full details of the order (ie,
drugs, doses, schedule, other medications, etc). An easy-touse process is also available for selecting an off-pathway
treatment. Physicians are never prevented from going off
pathway but, rather, are asked to describe the reason for
going off pathway (from a predefined list).
Results of Via Oncology Pathways Use
KEY POINTS
Issues in oncology today of quality, variability, and

cost are driving the need for physician-led solutions.
A rigorous clinical pathways program includes physician experts who meet routinely to evaluate evidence and develop care algorithms that standardize
personalized medicine.
Strategies to engage physicians must be employed,
including involvement in the development and maintenance of the pathways content.
Decision support tools are critical components to
achieving adherence and measuring results.
Early results demonstrate the success of clinical
pathways programs in terms of adoption, adherence,
reducing variability in care, patient outcomes, and
cost containment.
Currently at UPMC, approximately 90 medical oncologists and 30 radiation oncologists use the Via Oncology
Pathways in their daily patient care. These oncologists
practice at 40 sites of service, including the flagship academic center in the heart of Pittsburgh and communitybased sites over a 100-mile radius throughout Western
Pennsylvania. In 2011, the UPMC physicians confirmed a
pathways status for 94% of their patient visits (195,000
visits) and achieved an on-pathway rate of 82%2 for their
treatment decisions (18,000 treatment decisions). The original premise of Via Oncology Pathways was to find the
minimum number of therapies to meet the needs of the
majority of patient scenarios. This result seems to reflect
that the goals of reducing unwarranted variability, adhering
to evidence-based medicine, and ensuring that each patients care is personalized, have been achieved. Other
practices using the Via Oncology Pathways have generated
comparable results.
e63
BRUFSKY, LOKAY, AND MCDONALD
A key question in the implementation of a pathways

program is whether such reductions in variability also
reduce or slow the rate of growth in cancer care costs. UPMC
has collaborated with two external parties to explore these
questions in three separate studies. First, through a multiyear effort with Highmark Blue Cross Blue Shield of Pennsylvania (Highmark), the largest commercial payer for
UPMC, two studies were conducted to compare the effects
of implementing the Via Oncology Pathways (then called
the UPMC Pathways) program on growth rate in the total
costs of cancer care with providers without such a program
(Robert Wanovich, BOCP, Highmark Blue Cross Blue
Shield, email communication, July 2007 and August 2009).
To determine if a growth-rate reduction occurred, it was
necessary to also study the rate of growth of cancer costs in
a control arm of patients being treated by non-UPMC
physicians who do not use any type of formal pathways
program. The two studies examined the effects of implementing the pathways for nonsmall cell lung cancer and
breast cancer. In both studies, two periods were measured:
the 12-month period before the implementation of the pathway within UPMC and the 12-month period after the implementation. Total costs of care were calculated by Highmark
for the control arm and the experimental arm, using data
sets of actual claims payments for inpatient, outpatient, and
pharmacy benefit claims. The results in both studies demonstrated a positive difference in the rate of growth favoring
the experimental (i.e., UPMC) arm.
In the study of the breast cancer pathway, preliminary
results showed a 16% growth rate of costs for the nonpathways arm between the two 12-month periods compared with
7% for the UPMC arm. In addition, hospital admissions per
100 patients decreased by 15% in the UPMC arm compared
with a 2% increase in the nonpathways arm. This difference
suggests that reductions in unwarranted variability combined with a prioritization of treatments with lower toxicities can affect the rate of hospitalizations for patients with
cancer. Such a reduction is not only a source of cost savings
but also a likely improvement in patient quality of life and
overall outcomes.
In the study of the nonsmall cell lung cancer pathway,
actual growth rates between the two 12-month periods were
6% for the control arm and 1% for the UPMC arm. Included
in this growth rate difference were reduced hospitalization
costs of 12% in the UPMC arm, compared with a 4% increase
in the control arm.
Although the Highmark studies focused on the effect of
the Via Oncology Pathways program on the rate of growth
in cancer costs in Western Pennsylvania, a third study with
IntrinsiQ examined the possible implications of the Via
Oncology Pathways program on the care provided elsewhere
in the United States (Ed Kissell, IntrinsiQ, email communication, September 2009). IntrinsiQ is a national leader in
the field of oncology information management and their
proprietary chemotherapy ordering software, Intellidose, is
used by approximately 700 oncologists in more than 100
practices. Their data set is widely used by key information
consumers such as pharmaceutical companies because of its
comprehensive nature, both in terms of clinical granularity
and its ability to accurately project patterns of care for the
entire United States. By reviewing the actual treatment
decisions made by its oncologist customers and comparing
those decisions to the Via Oncology Pathways recommenda-
e64
Table 1. Adjuvant Chemotherapy Usage and Clinical Trial

Participation Rates for 174 Patients Node-Negative,
HER2/Neu-Negative/Unknown, ER-Positive Breast Cancer
(On-Pathway Rate: 90.8%)
Treatment Selected, %
Oncotype Risk
High
Intermediate
Low
Not Ranked
TC
AC
Tamoxifen
Anastrozole
Clinical
Trial
34
73
40
27
76.5
80.8
59.3
11.8
2.7
7.4
6.8
55.0
14.8
6.8
30.0
3.7
5.9
1.4
7.5
7.4
Abbreviations: AC, doxorubicin/cyclophosphamide; ER, estrogen receptor;

HER2/neu, human epidermal growth factor receptor-2; TC, docetaxel/cyclophosphamide.
tions, IntrinsiQ was able to calculate the financial effects on

the costs of drugs to payers if actual care had instead
complied with the Via Oncology Pathways. By its calculations, payers could save approximately 40% on oncology
drug costs if the Via Oncology Pathways had been followed
in all retrospective prescribing decisions. Assuming that
actual Via Oncology Pathways are followed in only 80% to
85% of patient scenarios, such savings would likely be in the
24% to 32% range.
Finally, a pilot study of Via Oncology Pathways and
Horizon Blue Cross Blue Shield of New Jersey (Horizon) has
generated encouraging results in an early analysis (Richard
Popiel, MD, Horizon Healthcare Innovations, and Richard
Weininger, MD, CareCore National, email communication,
February 2012). Two large practices in northern and southern New Jersey implemented the Via Oncology Pathways in
the third quarter of 2010. The study compares total cost of
care (excluding radiation oncology, because the participating
practices provided only medical oncology services) between
these Via Oncology Pathways practices (experimental arm)
and the remaining practices in New Jersey (control arm)
over the same periods. The initial results are being validated
by Horizon with possible publication in the second half of
2012.
Breast Cancer Treatment Patterns at UPMC-CC
An analysis of the patterns of care for patients with breast

cancer within UPMC-CC for the 12 months that ended May
31, 2011, was performed to describe utilization patterns and
concordance with the Via Oncology Pathways. This period
was selected to avoid the confounding changes in care
patterns in the second half of 2011 because of concerns from
the United States Food and Drug Administration over the
use of bevacuzimab to treat patients with breast cancer. The
analysis included new chemotherapy treatment decisions for
Table 2. Frequency and Description of Treatment Decisions
According to the Via Oncology Pathways and Accrual to Clinical
Trials for 104 Patients with HER2/Neu-Negative/Unknown,
ER-Negative, PR-Negative Breast Cancer Receiving Adjuvant
Chemotherapy (On-Pathway Rate: 79.8%)
Node Statis
No. of
Patients
TC
AC
TAC
Other
Clinical
Trial
Negative
Positive (13 nodes)
82
22
62.2
18.2
17.1
59.1
4.5
11
4.5
9.8
13.6

HER2/neu, human epidermal growth factor receptor-2; PR, progesterone receptor; TAC, docetaxel/doxorubicin/cyclophosphamide; TC, docetaxel/cyclophosphamide.
CLINICAL PATHWAYS STANDARDIZING PERSONALIZED MEDICINE

Trials for 109 Patients with HER2/Neu-Positive, ER-Positive
Breast Cancer Receiving Adjuvant Chemotherapy
(On Pathway Rate: 82.6%)
Node Status
No. of
Patients
TCH
AC
Other
Clinical
Trial
70
39
78.6
84.6
2.9
14.3
7.7
4.3
7.7
Negative
Positive

HER2/neu, human epidermal growth factor receptor-2; TCH, docetaxel/carboplatin with concurrent trastuzumab, followed by trastuzumab.
patients with metastatic breast cancer receiving adjuvant

and first-line treatment as documented by the UPMC-CC
oncologists in the Via Oncology Pathways system. Excluded
from the analysis were hormone therapy decisions (except
for patients with node-negative or HER2-negative disease
receiving adjuvant treatment) and biologic drugs prescribed
as monotherapy (these are assumed to be maintenance
therapy following combination therapy).
For this period and this population of decisions, the
on-pathway rate (ie, treatment decisions per the Via Oncology Pathways recommendation divided by all treatment
decisions) was 85.7% (505 treatment decisions). Tables 1
through 4 describe the frequency and description of treatment decisions according to the Via Oncology Pathways and
Trials for 118 Patients with HER2/Neu-Negative/Unknown,
ER-Positive, PR-Positive Metastatic Breast Cancer Receiving
First-Line Chemotherapy (On-Pathway Rate: 86.4%)
Treatment
Selected (%)
Capecitabine
Paclitaxel/Bevacizumab
Paclitaxel (protein-bound)/Bevacizumab
Paclitaxel (protein-bound)
Paclitaxel
Docetaxel/Gemcitabine
Other
Clinical Trial
27.1
22.9
19.5
8.5
5.9
2.5
8.5
5.1
Abbreviations: ER, estrogen receptor; HER2/neu, human epidermal growth

factor receptor-2; PR, progesterone receptor.
accrual to clinical trials. All off-pathways decisions are

aggregated and described as Other. These data are reported as is based on physician answers to questions
within the Via Oncology Pathways software application and
have not been validated against original medical records.
In the adjuvant setting, the pathway for breast cancer
provides guidance for the use of Oncotype DX recurrence
scores (RS) in both node-positive and node-negative settings.
Table 1 shows a breakdown of chemotherapy usage for
patients with node-negative, HER2/Neu-negative disease.
Chemotherapy usage is high ( 88%) in the high-risk group
and was not reported in the low-risk group. This usage is
consistent with retrospective data that suggest patients
with node-negative disease with high RS scores benefit from
chemotherapy and those with low RS scores ( 18) do not.1
Intermediate-risk patients showed more variability, with
84% receiving chemotherapy and 14% receiving hormonal
treatment alone.
In the first-line HER2/Neu-negative metastatic setting
(Table 4), usage reflects an 86% concordance with the Via
Oncology Pathways. This presentation provides an example
of how the disease committees provide options for numerous
patient scenarios, each with different associated treatments.
Conclusion
The results from the UPMC experience with the Via

Oncology Pathways program and those reported by other
clinical pathways programs2 suggest that these are effective
models for improving quality, reducing unwarranted variability in care, and reducing the rate of growth in the cost of
cancer care. A robust pathways program that reduces unwarranted variability can serve as the vehicle to improve the
value of cancer care to patients, payers, and providers
through increasing quality and decreasing costs. Oncologists
must take an active role in defining and implementing
cost-effective care for patients, payers, and referring physicians, or suffer the alternatives that potentially compromise
quality, access to care, and practice viability. If appropriately implemented, clinical practice guidelines and pathways are possible solutions to improving the quality and
cost-effectiveness of cancer care that preserves physician
decision-making, ensures access to evidence-based personalized medicine, and eliminates nonvalue-added administrative hurdles such as prior authorizations.
Author
Adam Brufsky
Employment or
Leadership
Positions
Kathleen Lokay
D3 Oncology
Solutions
Melissa McDonald
D3 Oncology
Solutions
Consultant or
Advisory Role
Celgene;
Genentech;
Novartis; Roche
Stock
Ownership
Honoraria
Celgene;
Genentech; Lilly;
Novartis; Roche
Research
Funding
Celgene;
Genentech; Lilly;
Novartis; Roche
Expert
Testimony
Other
Remuneration
REFERENCES
1. Paik S, Tang G, Shak S, et al. Gene expression and benefit of chemotherapy in women with node-negative, estrogen receptor-positive breast

cancer in the community setting. JOP. 2010;6(1):12-18.
e65
PATIENT PORTALS IN ONCOLOGY

CHAIR
Elizabeth S. Rodriguez, DNP, RN, OCN
New York, NY
SPEAKERS
Henry Feldman, MD
Division of Clinical Informatics
Brookline, MA
The Future of Oncology Care with Personal

Health Records
By Henry Feldman, MD, and Elizabeth S. Rodriguez, DNP, RN, OCN
Overview: Personal health records (PHRs) and patients access to their own clinical information through a patient portal
are changing the patient-physician relationship. Historically,
health care providers have been gatekeepers of patients
medical records. Now, these portals provide patients access
to clinical information, electronic messaging with the clinical
team, and appointment and billing information. This type of
access supports patient empowerment by engaging patients
in their own care.
Patients desire online access to information. The health
care industry, like any other, must respond to the needs of its
consumers. Oncology practices face unique challenges to
meeting this need because of the complex nature of medical
records of patients with cancer . Health care providers worry
about the consequences of patients receiving bad news
online, thereby increasing patient anxiety. This anxiety may, in
ERSONAL HRs, also referred to as patient portals,

began in 2000 with Beth Israel Deaconess Medical
Center launching PatientSite online,1 which allowed patients to view laboratory results, medication lists, imaging
reports, pathology reports, and financial reports, and to send
messages to their health care providers. Early during the
implementation, there was much controversy about whether
patients would be frightened or overwhelmed by the technical aspects of data and whether they would inundate their
providers with questions about the data in their record.
These concerns were most strongly raised by two groups of
providers: oncologists and psychiatrists. These fears were
not borne out over the 12 years since deployment of that
system, and in the intervening years, many large medical
centers and health care systems have developed and deployed large-scale PHRs. As of this writing, tens of millions
of patients have access to PHRs sharing select clinical data
over the Internet in a secure manner.
As part of The American Recovery and Reinvestment Act
of 2009, the federal government, under the auspices of
the Department of Health and Human Services decided
that the US health care system would benefit from the
increased use of electronic health records (EHRs). To ensure that government funds were allocated for useful progress, a set of phased meaningful use guidelines were
enacted by the Centers for Medicare & Medicaid Services
that vendors and health care providers would have to meet,
under the auspices of the Office of the National Coordinator
for Health Information Technology (ONC-HIT). One of these
requirements in phase II is the use of PHRs. The required
features of PHRs include inpatient and outpatient data that
must be shared. As such, practices must start addressing
these issues in relation to their practice areas, including
oncology.
PHRs and patient portals are rapidly being viewed as a
way to encourage patient empowerment and engage patients in their own care. Multiple studies have shown poor
retention of complex technical information by patients during office visits.2,3 In subspecialty fields, such as oncology,
extremely complex care plans are presented, with many
options and complications, and it is unreasonable to expect
e66
turn, increase providers workload by creating additional calls

or visits to the office.
These valid concerns require careful consideration when
implementing a PHR or patient portal into a practice. Providers will benefit from a clear understanding of actual compared
with potential risks and benefits. Much of the concerns about
the negative effect on providers workload and the potential
increase in patients anxiety have not been borne out. On the
other hand, the implementation strategy, governance structure, and end-user education are crucial components to ensuring success.
Successful implementation of a PHR or patient portal affords the opportunity to improve patient satisfaction and
increase efficiency in provider workflow. The possibility exists
to improve patient outcomes by engaging the patient in
decision making and follow through.
any patient to recall everything from an office or inpatient

visit. These tools can help leverage your time and improve
patient-physician communication.4,5 However, like any intervention, these tools have a risk-benefit ratio and side
effects and must be managed carefully.
Governance
One of the most difficult issues practices face in implementing this type of project, and particularly one that
crosses many stakeholder areas and needs, is governance.
PHRs involve a particularly complex area of governance, as
much of the data are regulated by various government
agencies, the data serve multiple purposes, the information
has not traditionally been intended for lay readers, and data
are being shared between content experts (providers) and
lay readers (patients and surrogates) who have different
ideas of immediacy, privacy, and explanatory material. All of
these factors need to be brought to the table in planning and
monitoring the implementation and use of a PHR.6,7 Clearly,
representation by the facilitys stakeholders is crucial; these
stakeholders include the technical staff of the practice;
clinical staff, including physicians and nurses; informatics
staff; business operations; and, most importantly, patients.
Practices may also consider addinglegal staff, as PHRs are
becoming more regulated.
This governance structure has several key tasks to initiate
and monitor the project:
Ensure that all regulatory issues are addressed
Work with patients and caregivers to determine what
data need to be shared, when, and in what formats
From Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA;
Ambulatory Care Services, Department of Nursing, Memorial Sloan-Kettering Cancer
Center, New York, NY.
Address reprint requests to Henry Feldman, MD, Division of Clinical Informatics, Beth
Israel Deaconess Medical Center, 1330 Beacon St., Suite 400, Brookline, MA 02446; email:
hfeldman@bidmc.harvard.edu.
1092-9118/10/1-10
PERSONAL HEALTH RECORDS
Determine what resources are needed at all phases of

the project
Set policies that all providers can work with to support
clinical care and patient empowerment
Decide what data are shared immediately, embargoed,
or blocked
Provide end-user support and education
Develop communication policies to prevent messaging
overload or under-alerting for providers and patients
These governing bodies must be given authority to make

changes to the applications as needs arise and to enact
policies for both staff and patients. If a practice has multiple
specialties, it is recommended that representatives from
several specialty areas be consulted, at least initially about
the design and needs analysis.
Embargos Compared with Blocking
One of the commonly voiced concerns about displaying

clinical data directly to patients is that a patient may see a
report with ominous findings before his or her clinician has
seen it. The most common phrase is similar to I do not want
my patient finding out their cancer has recurred from a
website! This is a reasonable concern and, although such a
situation may be rare, most clinicians understand that
pathology, radiology, and some laboratory results come back
with a differential rather than a simple diagnosis, which
may be quite concerning. This problem is similar to notes by
medical students, which often have academic leanings
rather than concise clinical pictures (that purpuric rash
could be Ebola but more likely is idiopathic thrombocytopenic purpura). Such notes can produce anxiety in patients,
whereas clinicians have the expertise to know the likelihoods of specific outcomes and can filter out much of the
noise to summarize for the patient.
As such, the governance structure needs to closely look at
the evidence for how to treat a given piece of clinical
information. The basic three choices are to share it immediately, embargo it for a specified time period, or block it
(never show it). Most medical centers have chosen to show
KEY POINTS
Personal health records (PHRs) and patient portals

provide patients with increased access to their own
medical information.
Important considerations exist for the use of PHRs
and portals related to the types of data to display, the
timeliness of data release, electronic communication
with patients, and end-user education.
A governance structure is essential for successful
system implementation and sustainability while
keeping providers engaged and protecting patient
safety and privacy.
PHRs and patient portals can serve as a platform for
online patient communities and as a way to share
clinical trial information.
Evaluation of how patients and providers use these
sites can be easily accomplished through analysis on
the backend of the system.
routine labs (such as basic chemistry profiles) immediately,

embargo pathology and radiology results, and to block such
results as HIV antibody testing. Most embargos last 1 to 6
days, and there is a safety net in knowing that if a clinician
fails to see a report for some reason, the patient will see it
after 6 days, and an important finding can be addressed at
that time. We address the special topic of clinical notes in
the section on Types of Data to Share.
Communication Strategies
Patient-physician communication is hard enough in person, as any clinician can tell you, and it becomes infinitely
more complex when delivered through electronic media.
Furthermore, because health care, and oncology in particular, is a team sport, it is often overwhelming to a patient to
figure out who to message about a given need. Front-line
providers are also overwhelmed by information overload and
do not welcome additional channels of communication. Interestingly, patients are also surprisingly cautious about
adding to clinicians workloads, and so messaging strategies
need to work with both the practice style and patients
information needs.8 Whatever strategy is adopted, it is
crucial that it be tightly integrated into practice workflows,
or it will not be advantageous to the clinician.
Messages typically can be divided into administrative
(medication refills, scheduling, referrals, financial matters)
and clinical (new medication request, educational needs,
and symptom/follow-up inquiries). Dividing basic tasks between clinical and nonclinical messaging makes sense and is
a good start. However, many lines are blurred and because
of this, some practices have adopted a system in which a
central person is responsible for triaging all messages to a
practice, which can be very successful. One subspecialty
clinic on PatientSite uses a total communication strategy in
this manner, where all communication to and from the clinic
is through the PHR and this single point of communication,
with a guaranteed response time. This system has been
quite successful for this clinic, but it required a large
investment (a full-time equivalent staff) by the clinic and
policy makers. A corollary to this is that every person who
receives a message is a single point of failure if he or she
either does not check messages or is unavailable to do so,
and so coverage schemes are required. A strategy that most
clinics adopt for this is a two-level process, where most
clinical messages are received by all clinical providers on
that patients team (physicians, nurse practitioners, registered nurses, and administrative staff) and whoever responds to the patient removes the message from the other
users queues. As a backup, each individual can denote
another user to provide coverage of his or her inbox.
Whatever your strategy is, it is vital to manage expectations; that is, to explain to patients what the turnaround
time is for messaging and what can be handled through
messaging and what should be handled through call-in or
other mechanisms. Closed-loop communication also helps
eliminate patients anxiety about the status of messages.
Typically, this type of communication can be performed
by the system automatically (for example, sending readreceipts or notification once a prescription has been sent
to the pharmacy). Patients understand that clinicians are
busy, and do not mind waiting for most responses for a
reasonable period of time, as long as that time is a known
e67
FELDMAN AND RODRIGUEZ
quantity. Determining the appropriate time period is an

important task for the governance committee.
One common complaint has been that time spent messaging with patients is nonreimbursed. However, in the upcoming capitated Accountable Care Organization (ACO) era,
time communicating electronically is time well spent if it can
save an office visit, admission, or test. Unlike telephone or
face-to-face communication, the asynchronous nature of
electronic communication allows the provider to prioritize
the communication, increasing efficiency.9 The economics of
PHRs will change over the next few years, and funding of
these projects need to take this into effect.
Types of Data to Share
The most controversial area of PHRs is what the patient

should be allowed to see and do. Most PHR providers agree
that financial information is noncontroversial to share, as is
medication and scheduling information; in fact, sharing this
information through a PHR can often reduce phone calls to
the clinic. Radiology and pathology reports are much more
controversial, but many practices now have several years of
experience with little untoward events, with use of embargos.
One area of concern is the question of releasing clinician
notes to patients. There are two large-scale examples of
sharing notes have demonstrated little downside and have
been well received. The OpenNotes trial in 2011 included
20,000 patients across three medical centers, and opened
all primary care notes to the patients.10,11 (You can view
videos about the experience at myopennotes.org.) The
oncology-specific example is from the University of Texas
M. D. Anderson Cancer Center in Texas, which opened all
clinician notes to patients, and approximately 60,000 patients have viewed their notes through their PHRs at
myMDAnderson.org. The standard concern is that patients
will be scared by what they see in the chart; however, hiding
medical issues from patients stands on shaky ethical
grounds. Furthermore, under federal Health Insurance Portability and Accountability Act (HIPAA) laws and some state
laws, patients must be able to view their medical records.
Your governance committee should work closely with
patients, providers, and other stakeholders to decide what
will and will not be shown and whether there will be
embargos, and have clear policies with transparent reasoning for each decision. These policies should be posted clearly
on the PHR for patients to see. You should also be prepared
to readdress policies as evidence and technology advance.
Provider Education
Like any clinical tool, full PHR use requires education for
the team that will be using it. Clinical champions will
increase your success of adoption, as providers are more
willing to accept advice from one of their own than from
dedicated training staff. If your PHR is so disruptive to
clinical workflows that a major training effort is required,
you may wish to reexamine the system design. Education is
needed about how to manage expectations of patients and
about the use of inappropriate phrases in documentation.
For instance, clinicians should use dyspnea rather than
SOB (or shortness of breath), as patients may not understand the abbreviation. Although patients have the right to
see their records, few do so in reality, and clinician documentation often reflects this.
e68
Patient Education
Patient education is needed in several areas, some technologic and some personal. Low technical literacy may be a
problem if your patient population is primarily older and low
health or English literacy may be a problem if you have a
large immigrant population. How you will address specific
groups of users and help them utilize your systems to their
best advantage should be worked out during the design
phase of the project. You also must develop a process for
educating patients about the rollout of the PHR itself, and
about communication expectations of the implementation. If
the patient education needs become too cumbersome, it
would be wise to reexamine the design.
Once patients are using the PHR, your practice needs to
decide how much patient education materials you will include for them to be self-sufficient. Patients are very selfreliant, and will use Google and other search engines to try
and understand what they read in their PHR and will
overcome the jargon.8 Your practice may decide on either
pre-made or new materials for patients. You should also
decide if you will send materials to patients (based on
diagnoses or on clinician selection) or simply provide libraries for them to browse/search. For instance, when you
display laboratory results, do you show the internal name
for the tests (such as Na for sodium) or do you show the full
names with links to basic explanations of each lab type?
Dana-Farber Cancer Center and Memorial Sloan-Kettering
Cancer Center have opted to use www.labtestsonline.org
to provide patients a reliable source of information in understandable terms.12,13 This public website is for patients
and caregivers and offers a large library of definitions and
explanations of common lab tests as a resource.
Communities
Many medical practices have considered implementing a

patient community function in their PHRs. Communities
are similar to social networking chat areas, where patients
discuss symptoms, therapies, and other topics of mutual
concern. The use of a community function is legally complex,
and needs to be carefully considered by any practice. Although many of these chat areas already exist on the
Internet, they are not hosted at medical practices and
HIPAA-covered entities.
The challenge with implementing communities is that
patients could breach each others confidential health information. Because the communities are hosted and provided
by a practice, patient breaches could be viewed as privacy
breaches, which, under current HIPAA and other privacy
regulations, can carry dire consequences. Also, patients
could give each other incorrect medical advice; again, because the community is a service provided by the practice,
there is potential for liability for the practice. Some institutions have dedicated health care providers to scan or moderate chat functions, which is clearly expensive, and it is
unclear that practice-hosted communities offer benefit over
similar functions already on the Internet outside of a practice. State laws can vary around liability and should be
carefully watched.
Protocols and Studies
PHRs also provide a powerful tool for monitoring patients

who are already on protocols and for recruiting potential
PERSONAL HEALTH RECORDS
study participants. One key problem in smaller practices is

that both patients and clinicians may be unaware of ongoing
studies in a particular disease; technology can effectively
match up clinicians, patients, and protocols. The PHR is also
an efficient way to securely gather information from patients
on protocols, such as symptoms, quality-of life scores, and
other research information. These uses have to be implemented in an ethical way that meets health services research guidelines but can be a powerful tool when done well.
Evaluating the Intervention
Most PHRs include some logging function to record what a

user did and saw while using the system. These data are rich
for mining to determine what patients are doing with the
system, what functions are used most frequently, and what
aspects are never used. Regular reporting and analysis of
usage patterns are crucial, as is periodic re-engagement
with stakeholders to address shifting needs and technologies. For example, in 2006, few designers thought about the
need to access patient portals from cell phones other than
through text messaging, but after the introduction of the

iPhone and other smartphones, this access became an important area for PHR designers to consider, as patients have
became much more sophisticated in their demands for mobile interaction.
Conclusion
PHRs are coming to practices around the United States,

and oncology practices large and small will have to adopt
these technologies. Practices will confront many of the
issues described in this chapter, and will need careful
governance, stakeholder engagement, and constant reassessment. PHRs serve many masters across many domains
and needs. Patients are already demanding access to their
records, and government regulations are requiring these in
upcoming years. Thus, instead of thinking of PHRs as a
burden, oncologists should view them as an opportunity to
promote their practice, engage and educate their patients,
and, possibly, in the ACO era, lower the utilization in their
practice.
Author
Henry Feldman*
Elizabeth S. Rodriguez*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Sands DZ, Halamka JD, Pellaton D. PatientSite: a Web-based clinical
communication and health education tool. HIMSS Proc. 2001;3:1-6.
2. Tarn DM, Flocke A. New prescriptions: how well do patients remember
important information? Fam Med. 2011;43:254-259.
3. Nightingale SL. Do physicians tell patients enough about prescription
drugs? Do patients think so? Postgrad Med. 1983;74:169-175.
4. Siteman E, Businger A, Gandhi T, et al. Patient review of selected
electronic health record data improves visit experience. AMIA Annu Symp
Proc. 2006:1101.
5. Siteman E, Businger A, Gandhi T, et al. Clinicians recognize value of
patient review of their electronic health record data, AMIA Annu Symp Proc.
2006:1101.
6. Reti SR, Feldman HJ, Ross SE, et al. Governance for personal health
records. J Am Med Inform Assoc. 2009;16:14-17.
7. Halamka J, Aranow M, Ascenzo C, et al. Health care IT collaboration in
Massachusetts: the experience of creating regional connectivity. J Am Med
Inform Assoc. 2005;12:596-601.
8. Earnest MA, Ross SE, Wittevrongel L, et al Use of a patient-accessible

electronic medical record in a practice for congestive heart failure: patient and
physician experiences. J Am Med Inform Assoc. 2004;11:410-417.
9. Rodriguez ES. Using a patient portal for electronic communication with
oncology patients: implications for nurses. Oncol Nurs Forum. 2010;37:667671.
10. Walker J, et al. Inviting patients to read their doctors notes: patients
and doctors look ahead - patient and physician surveys. Ann Intern Med.
2011;155:811-819, 2011.
11. Delbanco T, Walker J, Darer JD, et al. Open notes: doctors and patients
signing on. Ann Intern Med. 2010;153:121-125.
12. Wald JS, Burk K, Gardner K, et al. Sharing electronic laboratory
results in a patient portal-a feasibility pilot. Stud Health Technol Inform.
2007;129(Pt 1):18-22.
13. Rodriguez ES, Thom B, Schneider SM. Nurse and physician perspectives of cancer patients having online access to their lab results. Oncol Nurs
Forum. 2011;38:476-482.
e69
THE ASCO QUALITY ONCOLOGY PRACTICE

INITIATIVE AND BEYOND
CHAIR
Joseph O. Jacobson, MD
North Shore Medical Center
Salem, MA
SPEAKERS
Michael N. Neuss, MD
Nashville, TN
Robert Hauser, PhD, PharmD
American Society of Clinical Oncology
Alexandria, VA
Measuring and Improving Value of Care in

Oncology Practices: ASCO Programs from
Quality Oncology Practice Initiative to the
Rapid Learning System
By Joseph O. Jacobson, MD, MSc, Michael N. Neuss, MD, and
Robert Hauser, PharmD, PhD
Overview: Rising cancer care costs are no longer sustainable. Medical oncologists must focus on providing the
maximum value to their patients; improving short-term, intermediate and long-term outcomes; and managing overall costs.
Accurate measurement of outcomes and overall cost is essential to informing providers and institutions and in the quest for
continuous improvement in value. The ASCO Quality Oncology
Practice Initiative (QOPI) is an excellent tool for sampling
HEN PUBLISHED in 2001, Crossing the Quality

Chasm: A New Health System for the Twenty-first
Century defined quality of care as the degree to which
health services for individuals and populations increase the
likelihood of desired health outcomes and are consistent
with current professional knowledge.1 Six core components
of quality were identified: safety, effectiveness, patientcenteredness, timeliness, efficiency, and equitability. The
publication received widespread attention and served as a
clarion call to action to providers and health care organizations to begin a relentless focus on quality of care.
The Need to Focus on the Value of Cancer Care
At the time of publication of Crossing the Quality Chasm,

health care costs as a percentage of gross domestic product
had exceeded 14%. In 2010, the percentage had increased to
17.9%, and total health care expenditures had reached $2.6
trillion, according to the Center for Medicare and Medicaid
Services. Further rises in health care costs are now recognized as unsustainable. The Patient Protection and Affordable Care Act was signed into law by President Obama in
2010 at least partly in response to the recognition that
spiraling health care costs threatened the economic health of
the nation.
In the United States, the cost of cancer care is increasing
at a faster rate than nonmalignant conditions. If annual
direct cancer care costs cannot be contained, they are projected to reach $173 billion by 2020, representing a 39%
increase compared with 2010.2 Increases are caused by rises
in both the cost of therapy and the extent of care.3 Rises in
cost of therapy are partly justified by dramatic advances in
the management of cancer that have occurred in the past
decade. For example, a new generation of targeted chemotherapeutic agents has emerged with unparalleled activity
and with reduced toxicity compared with standard chemotherapy.4 However, the benefit of other new and expensive
technologies such as robotic surgery for early-stage prostate
cancer is still largely unproven.5
Porter and Teisberg argue that quality alone is insufficient to justify the incorporation of a new technology or
agent into routine use.6 They argue persuasively that value
is a far better means to assess the effect of a change in care,
defining value simply as outcomes achieved per dollar
spent. Dollars spent is intended to include the cost of care
e70
processes of care in medical oncology practices. To achieve

the larger goal of improving the value of cancer care, ASCO is
investing in the development of a Rapid Learning System,
which will leverage emerging information technologies to
more accurately measure outcomes (including those reported
by the patient) and costs, resulting in highly efficient, effective, and safe cancer care.
over a full set of interventions needed to manage a specific

medical condition.7 In this paradigm, Porter describes three
tiers of outcome. Tier 1 includes measures of success familiar to oncologists, including survival and response to treatment. Tier 2 assesses the process of recovery and focuses on
disutility of care including treatment delays, toxicities, adverse events, and errors. The long-term outcome and late
treatment effects constitute Tier 3 outcomes.
Defining the value of the care that we provide to our
patients is vital for medical oncologists as we prepare for
new reimbursement models. As Porter notes, valueneither an abstract ideal nor a code word for cost reduction
should define the framework for performance improvement
in health care.7 For medical oncologists, this requires new
attention to delivery of cancer care with a focus on the full
spectrum of services and on providing care in which value
can be quantified. Various models for providing value-based
cancer care have been described.8-10
Providing High-Value Care Requires an Organized
Health Care Delivery System
Bohmer has observed that high-value health care organizations have four common habits: (1) specification and
planning; (2) intentional infrastructure design; (3) measurement and oversight and; (4) self-study.11 He observes that
many health care organizations succeed at accomplishing
some of these goals, but only a handful manage them all; it
is these few that have succeeded in delivering high-value
cancer care. Examples include Intermountain Healthcare
and Mayo Clinic.
Specification and Planning
Bohmer defines specification as separating heterogeneous

patient populations into clinically meaningful subsets. As he
notes, Many hospitals and clinicians do not plan care
From the Dana-Farber Cancer Institute, Boston, MA; Vanderbilt-Ingram Cancer Center,
Nashville, TN; Department of Quality and Guidelines, American Society of Clinical
Oncology, Alexandria, VA.
Address reprint requests Joseph Jacobson, MD, MSc, 450 Brookline Ave., Dana-Farber
Cancer Institute, Boston, MA; email: joseph_jacobson@dcfi.harvard.edu.
1092-9118/10/1-10
IMPROVING VALUE OF CARE IN ONCOLOGY
processes in advance in such detail; instead, they treat each

new patient or problem as a random draw from a heterogeneous population and must, therefore, reinvent the strategy
for solving it. For much of its history, medical oncology has
approached care this way. Fortunately, advances in genomics have begun to identify clinically meaningful disease
subsets. Patients with adenocarcinoma of the lung, for
example, are now routinely tested for the presence of activating mutations of epidermal growth factor or translocations of anaplastic lymphoma kinase (ALK).12,13 Patients
with lung cancer with these genotypes now have access to
targeted therapies that have transformed the natural history of their disease. By planning, Bohmer argues for the
development of pre-emptive treatment approaches for patients using clinical practice guidelines or clinical pathways.
Intentional Infrastructure Design
The gradual shift of medicine from guild to profession is

embodied by this habit.14 This entails designing care to
maximize the use of each team members skills and expertise and requires engagement of the patient and family. The
medical home model now being implemented widely in
primary care is an example of effective infrastructure design, and a pilot in one medical oncology practice suggests
that the model can be extended.9 Multidisciplinary cancer
carewhen well designedis a powerful tool to leverage
personnel and limited resources.15
Measurement and Oversight
High-value organizations go far beyond measurement

required for external reporting to the development of a set of
internal measurements that become integral to accountability and organizational management.
Self-Study
Bohmer notes that high-value organizations treat clinical

knowledge as an organizational as well as individual property . . . these organizations deliberately nurture a culture
that supports learning. . . . In this environment, in which
KEY POINTS
Medical oncologists must strive for high value of

cancer care defined as outcomes achieved per dollars spent.
Reliable, accurate means to assess the quality and
cost across the spectrum of cancer services must be
created to drive high-value care.
The ASCO Quality Oncology Practice Initiative
(QOPI) is available to all U.S. medical oncologists to
sample care processes and to compare their performance over time and against national benchmarks.
The ASCO QOPI Certification Program (QCP) allows
practices to demonstrate high quality of care and safe
chemotherapy administration practices.
ASCO is developing a Rapid Learning System that
leverages new information technologies with the goal
of improving the value of cancer care by focusing on
effectiveness, safety, efficiency, and quality.
specification and planning are built into systems, variations

from guidelines and pathways are studied to understand
and learn from variation. This is the clinical equivalent of
the managerial approach of managing by exception.
Managing What We Measure
At the 1962 Yale University Commencement Address,

John F. Kennedy noted that for the great enemy of truth is
very often not the lie deliberate, contrived, and dishonest but the mythpersistent, persuasive, and unrealistic.
This admonition is as true in medicine as it is in politics.
Without data to inform our delivery of care, we are dependent on subjectivity and biases and at risk of squandering
precious resources and mistreating our patients.16 The collection of accurate, granular, and timely data is vital for
organizational and provider growth and in the quest for
excellence. Berwick, James, and Coye describe two overlapping types of performance assessment, measurement for
selection and measurement for improvement.17
Measurement for selection. This approach to improving
the quality of care is based on the premise that health care
outcomes studied in any setting (e.g., individual practitioner
or health care organization) will have a distribution of
performance levels. In principle, this consumer-based approach to value simply requires the selection of the desired
outcome, review of the provider or organizational performance, and selection based on identifying the best
performerget better care by choosing better care.17 A
major shortcoming of this approach is that it does not
improve the overall quality of care; care is simply shifted
toward high performing practitioners or organizations.18 In
addition, measurement for selection remains an inexact
science in medicine. Outcome variation across providers and
organizations is often because of the complex interactions of
patient-specific factors (which are often imperfectly measured) with the health care system. In addition, measurement for selection often relies on a series of summary
statistics lacking sufficient specificity or granularity to draw
precise conclusions for most medical conditions or procedures.
Measurement for improvement. Using this approach, focus is shifted toward the processes of care rather than the
providers, and, when successful, the entire distribution of
performance levels is shifted to a higher level. Institutions
such as Intermountain Healthcare (IHC) that have experienced major strides in improving quality over the past
decade have consciously shifted from measuring for selection to measuring for improvement.18 IHC success has
depended on building a comprehensive clinical information
system, engaging clinicians to develop standardized approaches (care process models [CPMs]) to the management
of common conditions, encouraging clinicians to vary from
these approaches when justified by circumstances, and continuously modifying the CPMs based on analysis of variations and on emerging literature.
The Emergence of the Quality Oncology Practice
Initiative (QOPI)
In response to the Institute of Medicines National Cancer

Policy Board recommendations published in 1999,19 ASCO
commissioned researchers at the Harvard School of Public
Health and at the RAND Corporation to undertake a study
e71
JACOBSON, NEUSS, AND HAUSER
Fig. 1. Growth of the Quality Oncology Practice Initiative (QOPI) Since 2006.
of oncology care in the United States. The National Initiative for Cancer Care Quality (NICCQ) surveyed breast and
colorectal cancer care in five U.S. metropolitan areas.20 This
cross-sectional analysis of processes of care identified widely
varying rates of adherence to recommended care and identified significant opportunities for improvement.21
In 2002, Joseph Simone, MD, co-chair of the National
Cancer Policy Board, recruited seven volunteer medical
oncologists from community practices to consider alternative methods for measuring cancer quality in ambulatory
practices. This alpha group of clinicians created a measurement system, now known as the Quality Oncology
Practice Initiative (QOPI), based on the goal of promoting
excellence in cancer care by helping oncology practices
create a culture of self-examination and improvement.22
Through a series of pilot projects, a collection of consensusbased and evidence-based performance indicators was selected, a chart abstraction methodology was created, and a
system for rapid feedback of results to practices was created.
In 2006, based on the success of a series of pilot projects,
ASCO offered all U.S. members the opportunity to participate in QOPI. QOPI has grown rapidly since the initial
national rollout. Nearly 100 performance indicators have
been created, spanning multiple domains and diseases.
More than 700 practices have enrolled in QOPI, and semiannual data collections now routinely include 250 to 300
practices (Fig. 1). A rigorous QOPI Certification Program
was introduced in 2010.23 At the end of 2011, 105 practices
had achieved certification.
ASCO remains at the forefront of physician specialty
societies in its quest to continuously improve the value of
care provided to patients with cancer. This manuscript
describes the current state of the QOPI program and preliminary planning for an entirely new approach to measuring and improving cancer care based on principles recently
developed by the Institute of Medicine for rapid learning.24
How QOPI WorksA Consideration of Effort,
Expense, and Rewards
The American Society of Clinical Oncology (ASCO) Quality Oncology Practice Initiative (QOPI) has two major com-
e72
ponents. The self-reported QOPI practice survey provides a

structured method and benchmark comparison designed to
help practices understand their areas of strength and weakness in patient care, both in absolute and comparative
terms. By giving practices a way to measure their performance on a variety of process measures designed to look at
specific domains of care, practitioners can identify areas of
potential opportunity. By providing comparative data, they
allow practices a way to understand if their weaknesses are
similar to or different from comparable groups.25,26 Through
the QOPI certification process, they can obtain external
verification of their performance as well as compliance with
policies and procedures consistent with the ASCO/ Oncology
Nursing Society (ONS) chemotherapy safety standards.23
QOPI participation is open to all U.S. ASCO members
twice yearly. An estimated 10% to 20% of members have
submitted records. Most of the practices that have participated have participated repeatedly. The chart abstraction
and data submission process seems daunting at first, but
becomes easier over time. Identifying charts for analysis is
easier with electronic billing or charting systems, which
allow sequential lists of patients with specific diagnoses to
be arranged and facilitate chart review without physically
finding and handling a paper chart, though both systems are
equally acceptable. Interestingly, some practices have found
that review in paper records is faster than electronic formats. Our experience, having reviewed charts in both, is
that the electronic format allows much quicker abstraction
as reviewers become more familiar with the data elements,
but it still takes 20 40 minutes per chart, with the time
dependent on the number and complexity of modules. At the
Vanderbilt-Ingram Cancer Center, physicians and nurses
perform the chart reviews. In private practices, abstraction
is often done by both nurses and/or research data technicians. Assuming 40 100 charts are reviewed at an average
employee expense of from $40 to $200 (the latter for physicians) per hour, the cost for a round of data submission
ranges from $1,600 to $20,000. Based on reports of participants, it seems reasonable to estimate $4,000 per practice
per round; the expense of the effort is dependent on the pay
scale of those performing abstraction.

Table 1. QOPI Participation and QOPI Certification Compared
Program
QOPI Participation
Cost to ASCO Members
Free
Measures
89 questions in 7
modules [1]
Standards
Not applicable
External Audit
Duration of Certification
Comments
No
Not applicable
Comparative data are
available to help a
practice measure its
achievement
QOPI Certification
$300015,000 (Depending
on practice size and
number of locations)
Passing score on 5 modules
(Currently, score to pass is
72% on general measures,
80% on appropriate
adjuvant treatment)
Documented policies
consistent with 17 of
ASCO/ONS chemotherapy
administration [i]standards
Yes
3 years
Pricing is variable for larger
groups
Abbreviations: QOPI, Quality Oncology Practice Initiative; ONS, Oncology

Nursing Society.
ASCO requests that each QOPI practice identify a practice steward who will take responsibility to educate practitioners about their achievement in QOPI. Our experience in
distributing these results in both community and academic
settings has been that physicians accept the good results
with a tremendous amount of pride, and simultaneously
doubt the reliability or importance of measures where their
performance lags. Nevertheless, they almost always improve their performance in these areas, although it is very
difficult to determine whether true performance or documentation of performance is really changing. From personal
experience, one of the authors (MNN) notes that he has
learned much and changed how he interacts with patients.
For example, he is much more attentive to infertility counseling and interventions before starting chemotherapy and
trying to prevent constipation in patients on narcotic analgesics than previously.
QOPI certification of a large academic practice such as
Vanderbilt-Ingram Cancer Center poses challenges. Just
like preparing for any serious certification exercise, such as
the Commission on Cancer, Joint Commission, or Magnet
Nursing Certification, it is a significant team effort that
requires commitment of time and resources. Achieving certification requires that a practice surpass a threshold
achievement level during a round of data submission and
demonstrate compliance with a subset of the chemotherapy
administration safety standards developed by ASCO and
ONS.27 The number of charts needed for review is dependent on the practice size (Table 1).
The benefits of participation and certification vary by
institution. The intrinsic rewardassessing practice or
practitioner performanceis great. Other benefits are possible. Listing of certification has likely attracted some patients. More tangible advantages have been realized by some
practices. For example, some practices have been provided
payment bonuses from payers. Some exclusive quality programs recognize participation or certification as part of
inclusion standards. Finally, some practices have negotiated
rate premiums from payers based on the achievement of
certification (though antitrust concerns preclude any specific
detailing of these arrangements).
In summary, QOPI participation and certification require

significant effort associated with nontrivial expenses for
practices. There are intrinsic rewards and the opportunity
for self-examination and improvement with even a partial
data submission, although full and broad participation and
benchmark comparisons offer more.
The ASCO Rapid Learning System (RLS)
The field of oncology is entering an unprecedented time

where advances in technology, drug development, treatment
techniques, genomics, and research are driving health care
costs to unsustainable levels.28 As a result, there is an
increasing need for higher quality care, at lower costs, with
better outcomes. To meet these needs and demonstrate
success, each component (quality, cost, and outcomes) must
be accurately defined, measured, and reported. Fortunately,
we have defined quality measures in oncology (QOPI), costs
can be captured via claims and other means, and outcomes
can be defined and reported though electronic medical record (EMR) systems. Unfortunately, all of these data reside
in many different electronic (sometimes paper) systems that
do not interconnect. If these important data continue to
reside in many disparate systems, reporting high-value
cancer care will be burdensome or impossible. However,
there is a solution on the horizon: a learning health care
system. As defined by the Institute of Medicine, a learning
health care system is grounded in the principles of improving effectiveness, safety, efficiency, and quality, paired with
evaluating the processes shared by engineering and medicine.29 ASCO is taking the lead to develop a rapid learning
health system in oncology to accomplish the goal of improving effectiveness, safety, efficiency, and quality in oncology.
The ASCO RLS will have the ability to utilize evidencebased medicine in the form of publications, guidelines, and
measures, pair it with patient-level data from multiple
Health Information Technology systems, aggregate it, analyze it, and turn the data into usable knowledge.
QOPI provides a uniquely strong foundation, but ASCO
recognizes that members and the oncology community will
benefit from an initiative with even greater sophistication.
In 2011, the ASCO Board directed that immediate work
begin on construction of a cutting-edge rapid learning system for oncology.
What Is a Rapid Learning System?
In the most basic form, an RLS is a technology platform

that allows for the collection of data from different health
information technology systems to be collected, aggregated,
analyzed, and turned into usable information to improve
treatment and drive scientific learning and discovery from
every patient at every encounter (Fig. 2).30 From a culture
standpoint, an RLS speeds and improves how we learn
(Fig. 3).
What Can We Do With an RLS?
The output from an RLS is potentially unlimited. The

community will have the ability to obtain real-time clinical
decision support, identify patients for clinical trials, conduct
population health research, conduct health outcomes research, conduct comparative effectiveness research, provide
quality benchmarking, possibly use for REMS reporting,
and identify rare adverse events by detecting early warning
signals.
e73
Fig. 2.
Proposed Model for an Oncology Rapid Learning System (RLS).
One example of the power of an oncology RLS can be

illustrated with the example of erythropoietin stimulating
agents (ESAs). Based on studies demonstrating reduced
need for blood transfusion in chemotherapy patients, the
first of these agents was approved for oncology use in 1993.
ESAs subsequently became a commonly used component
of cancer supportive care during active treatment. Beginning in 2004, emerging safety concerns led to product
label warnings. In 2007, additional studies suggesting
negative effects on survival and disease progression caused
the United States Food and Drug Administration (FDA) and
CMS to issue further product label warnings and to restrict
coverage (Product Package Insert). However, if we had
access to data from an RLS at the time, with real-time
capture of millions of clinical data points and patient outcomes, it is reasonable to conclude that we could have
identified these safety signals before 2004. If the use of
ESAs from 2004 -2007 had been at levels seen between
2008 and 2011, the quality of health care provided to our
patients would have been improved, and CMS could have
saved more than $3 billion (assuming a conservative estimate of 6,000 practicing oncologists prescribing ESAs during these years).
Why Did ASCO Undertake an RLS?
ASCO has been committed to quality of cancer care since

the original founders met in 1964 to share ideas and research on how they were improving patient care and quality outcomes in the field. ASCO continues to meet its
quality driven mission with the QOPI program, and an
RLS is just the next iteration in how we learn, share
knowledge, and improve quality outcomes. ASCO has a
patient-driven mission, which also focuses on education and
quality. ASCO has built trust with its membership and will
always keep members best interest and outcomes front and
center.
How Is ASCO Proceeding? What Will It Take to Build
an RLS?
ASCO has developed a plan, led by the Board of Directors

and an RLS Advisory Group, in which a solid foundation is
built on trust and the following seven guiding principles:
1. Rigorousthe hallmarks of RLS will be methodological
rigor and comprehensiveness
2. Patient-focusedRLS will reduce morbidity and mortality and incorporate the humane values of the profession
Fig. 3. Comparison between the Current Model of

Care and a Proposed Model Based on a Fully Functional
Rapid Learning System (RLS).
e74
3. Transparentthe methodology of RLS and the incentives of every collaborator and contributor will be
visible
4. Independentthe credibility of RLS depends on oncology peer decision-making and collaboration without
compromise of principles. There will be a firewall
between industry and guideline and standards development
5. InclusiveASCO is open to collaborating with any organization, public or private, that accepts its collaboration principles. RLS will be accessible and usable
from non-ASCO platforms
6. StreamlinedRLS data collection will be efficient and
not burdensome to providers
7. SustainableASCO will make a substantial investment
in RLS and needs a sustainable economic model. Any
commercialization of RLS will be supported by ASCOs
educational and quality endeavors
The plan for building the RLS is to begin with small pilot
projects to build the foundation of the system. The foundation of the RLS is comprised of three key components:
1. Data Governance
2. Data Standards
3. Flexible Technology Platform.
A second key project in the process is to define data

standards. ASCO will work with and convene panels of
experts to begin to build on work in medical data standards
and enhance the oncology data standards already developed.
For data to be easily aggregated, common definitions are
needed in which all systems report. Developing this data
ontology in oncology, will be paramount to the success of
the system.
ASCO will soon be selecting a technology platform on
which to build the RLS. Choosing the correct technology
platform is critical as ASCO develops the RLS. The platform
will need to have the ability to work within the constraints
of data and technology today, but be flexible enough to
adjust to the health information technology systems and
software of the future. Again, ASCO will engage experts in
health information technology, oncology, and developers to
help guide its decisions in this arena. Once the foundation of
the RLS is built, then portals and applications will be built.
ASCO currently envisions having physician portals, clinical
decision support portals, patient portals, and numerous
applications to enhance the quality of oncology care.
Conclusion
ASCO will convene experts in the area of data governance

to develop a defined set of policies and procedures on how
data are collected, secured, accessed, aggregated, and reported. The Society will engage experts from government
agencies, law firms, health systems, academic centers, community practices, and other industries to ensure the RLS is
built with the strongest and most transparent data governance polices possible. These policies will be maintained by
a board of experts who will be responsible for stewarding
and updating the documents.
ASCO has developed and earned the trust of its members

since 1964 and has a dedicated mission to improve quality
and a history of successful projects and programs. QOPI and
the QOPI Certification Program allow ASCO participants to
measure and improve practice performance, to compare
themselves to their peers, and to guarantee that they adhere
to safe chemotherapy administration practices. To continue
to lead the field, and to respond to the urgent need to
improve the value of cancer care, ASCO is investing in the
development of a comprehensive information technologybased Rapid Learning System.
Author
Joseph O. Jacobson*
Michael N. Neuss
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
EMeRged Systems
(U), (L)
Robert Hauser
Amgen; Novartis
REFERENCES
1. Institute of Medicine: Crossing the Quality Chasm: A New Health
System for the Twenty-first Century. Washington DC: National Academies
Press, 2001.
2. Mariotto AB, Robin Yabroff K, Shao Y, et al. Projections of the Cost of
Cancer Care in the United States: 2010. J Natl Cancer Inst. 2011;103:117128.
3. Smith TJ, Hillner BE. Bending the Cost Curve in Cancer Care. N Engl
J Med. 2011;364:2060-2065.
4. Stegmeier F, Warmuth M, Sellers WR, et al. Targeted Cancer Therapies
in the Twenty-First Century: Lessons From Imatinib. Clin Pharmacol Ther.
2010;87:543-552.
5. Barbash GI, Glied SA. New technology and health care costs-the case of
robot-assisted surgery. N Engl J Med. 2010;363:701-704.
6. Porter ME, Teisberg EO. Redefining health care: Creating value-based
competition on results. Boston, MA: Harvard Business School Press, 2006.
7. Porter ME. What is value in health care? N Engl J Med. 2010;363:24772481.
8. Pollock RE. Value-based health care: The MD Anderson experience. Ann
Surg. 2008;248:510-518.
9. Sprandio J. Oncology patient-centered medical home and accountable

cancer care. Community Oncology. 2010;7:565-572.
10. Bach PB, Mirkin JN, Luke JJ. Episode-based payment for cancer care:
A proposed pilot for Medicare. Health Aff (Millwood). 2011;30:500-509.
11. Bohmer RM. The four habits of high-value health care organizations.
N Engl J Med. 2011;365:2045-2047.
12. Landi L, Cappuzzo F. Targeted therapies: Front-line therapy in lung
cancer with mutations in EGFR. Nat Rev Clin Oncol. 2011;8:571-573.
13. Shaw AT, Yeap BY, Solomon BJ, et al. Effect of crizotinib on overall
survival in patients with advanced non-small-cell lung cancer harbouring
ALK gene rearrangement: A retrospective analysis. Lancet Oncol. 2011;12:
1004-1012.
14. Sox HC. Medical Professionalism and the Parable of the Craft Guilds.
Ann Intern Med. 2007;147:809-810.
15. Jacobson JO. Multidisciplinary Cancer Management: A Systems-Based
Approach to Deliver Complex Care. J Oncol Pract. 2010;6:274-275.
16. Auerbach AD, Landefeld CS, Shojania KG. The Tension between
Needing to Improve Care and Knowing How to Do It. N Engl J Med.
2007;357:608-613.
e75

17. Berwick DM, James B, Coye MJ. Connections between quality measurement and improvement. Medical Care. 2003;41:130-138.
18. James BC, Savitz LA. How Intermountain Trimmed Health Care Costs
Through Robust Quality Improvement Efforts. Health Affairs. 2011;30:11851191.
19. Hewitt ME, Simone JV. Ensuring Quality Cancer Care. Washington
DC: National Academies Press, 1999.
20. Schneider EC, Malin JL, Kahn KL, et al. Developing a system to assess
the quality of cancer care: ASCOs national initiative on cancer care quality.
[erratum appears in J Clin Oncol 2004 Nov 15;22:4656]. J Clin Oncol.
2004;22:2985-2991.
21. Malin JL, Schneider EC, Epstein AM, et al. Results of the National
Initiative for Cancer Care Quality: How can we improve the quality of cancer
care in the United States? [erratum appears in J Clin Oncol 2006;24:1966].
J Clin Oncol. 2006;24:626-634.
22. Neuss MN, Desch CE, McNiff KK, et al. A process for measuring the
quality of cancer care: The Quality Oncology Practice Initiative. J Clin Oncol.
2005;23:6233-6239.
23. McNiff K, Bonelli KR, Jacobson JO. Quality Oncology Practice Initiative Certification Program: Overview, Measure Scoring Methodology, and Site
Assessment Standards. J Oncol Pract. 2009;5:270-276.
e76
24. Institute of Medicine: A Foundation for Evidence-Driven Practice: A

Rapid Learning System for Cancer Care: Workshop Summary. Washington
DC: The National Academies Press, 2010.
25. Neuss MN, Desch CE, McNiff KK, et al. A Process for Measuring the
Quality of Cancer Care: The Quality Oncology Practice Initiative. J Clin
Oncol. 2005;23:6233-6239.
26. Neuss M, Jacobson J, McNiff K, et al. Evolution and elements of the
Quality Oncology Practice Initiative measure set. Cancer Control. 2009;16:
312-317.
27. Jacobson JO, Polovich M, McNiff KK, et al. American Society of Clinical
Oncology/Oncology Nursing Society Chemotherapy Administration Safety
Standards. J Clin Oncol. 2009;27:5469-5475.
28. Garber AM, Skinner J. Is American health care uniquely inefficient.
Journal of Economic Perspectives. 2008;22:27-50.
29. Grossmann C, Goolsby WA, Olsen L, et al. Engineering a Learning
Healthcare System: A Look at the Future - Workshop Summary, The
Learning Health System Series Roundtable on Value and Science-Driven
Healthcare. Washington DC: National Academies Press, 2011.
for cancer care. J Clin Oncol. 2010;28:4268-4274.
PHYSICIAN WELLNESS: COPING WITH REPETITIVE

LOSS AND WORK-RELATED STRESS
CHAIR
Michael J. Fisch, MD, MPH
Houston, TX
SPEAKERS
Teresa Gilewski, MD
New York, NY
Joshua Hauser, MD
Chicago, IL
B u o y a n c y : A M o d e l f o r S e l f - R e fl e c t i o n a b o u t
Stress and Burnout in Oncology Providers
By Michael J. Fisch, MD, MPH
Overview: Burnout is a prevalent syndrome among oncology

providers marked by exhaustion, a sense of ineffectiveness,
and depersonalization. This syndrome can have enormous
influence on patient care as well as the providers career
fulfillment and personal and family well-being. A buoyancy
URNOUT IS a familiar syndrome in health care.

Maslach defines it as a prolonged response to chronic
emotional and interpersonal stressors on the job, and [it] is
defined by the three dimensions of exhaustion, cynicism, and
inefficacy.1 It is well described in medical oncologists,2,3
pediatric oncologists,4,5 surgical oncologists,6,7 primary care
physicians,8 nurses,9 residents,10-12 medical students,13 and
others, including physician leaders.14,15 Provider burnout
has been associated with varied factors, including higher
workloads, less time with patients, lower experience and
training, young marriages, having young children, lower
reimbursements, higher levels of debt, work-family conflict,
poor health, unrealistic expectations, and anger issues. The
adverse consequences of burnout are not difficult to predict,
and include despair, lower fulfillment, low productivity,
increased turnover, loss of boundaries with patients, increased medical errors, and suboptimal care. Several excellent reviews about ways to understand and combat burnout
in oncology have been published.16,17
The focus of this manuscript is to describe a practical
paradigm for the opposite of burnout: buoyancy. Buoyancy is
a force that keeps one afloat, in contrast to sinking. Posing
this concept as the dependent variable in a multivariable
model, 10 parameters that might be expected to absorb most
of the variance in a model of buoyancy are depicted in Figure
1. They are listed in no particular order, do not have equal
importance within persons or among individuals, and are
expected to vary over time. These theoretical parameters
are: (1) autonomy, (2) exercise of skill, (3) establishing and
maintaining meaningful relationships, (4) being awake to
the present reality (mindfulness), (5) gratitude, (6) courage,
(7) appreciation of impermanence, (8) compassionate mind
frame, (9) finding and keeping ones sense of safety and
security, and (10) answering the question, Do I matter?
The parameters of buoyancy can be compared with ones
finances, in which the different parts of a portfolio contribute to the bottom line. For instance, if managing fear and
finding and keeping your safety and security stocks take a
brief hit, they may rebound splendidly, or perhaps ones
overall buoyancy will trend upward by virtue of a compensatory increase in other areas. With this in mind, one can see
how buoyancy would be expected to vary day to day and year
to year. However, a solid manager of the metaphoric portfolio (i.e., you) could keep the net performance in a consistently good range and find ways to hedge against extreme
underperformance. Extreme underperformance in buoyancy
would likely manifest as burnout.
A detailed examination of some of the parameters contributing to buoyancy will reveal how they can manifest and
change over time.
model is proposed as a method to take stock of key parameters that may contribute to happiness and resiliency. Selfmonitoring of personal buoyancy parameters may help
oncology providers prevent burnout.
Autonomy (Freedom to Choose)
Oncology providers are well aware of autonomy as an

ethical principle and routinely apply it to medical decision
making. For clinicians, it is worth reflecting on whether one
has a modicum of control over the work environment and
schedule as well as how one proceeds to undertake various
projects. With changes in health care and the trend toward
larger organizational structures, some providers may feel
that they have lost ground in this dimension. It is worthwhile to consider what has not been lost, to determine which
expectations are realistic, and to have a dialogue and negotiation with colleagues on key points of autonomy when
necessary.
Exercise of Skill (Doing Your Thing)
This parameter is focused on realizing what ones

strengths and passions are and making sure one does not
compromise too much in this area. For example, people who
enjoy taking care of patients with complex conditions and
mastering difficult diagnoses would not fare that well in a
high-volume clinic focused on breast cancer survivorship
care. This extends not only to the obvious focus of care but
also to the aspects of work that make it most fulfilling. For
example, the nontechnical, humanistic aspects of work (especially in settings associated with medical failure) are often
some of the most gratifying experiences.18 Understanding
that these aspects can be incorporated into ones focus, while
continuing to build on skills in communication and compassionate care, can render any patient care setting or individual encounter more enriching.19
Early in my career, I received this advice: Do not let
medical school ruin your education. Medicine, with its
potentially endless layers, is sufficiently absorbing that it
can overgrow other dimensions of skill and interest if allowed. Think deeply about your other skills and interests,
perhaps music, reading mystery novels, cooking, boating,
helping others in the community, teaching or coaching
children, martial arts, and so on. Make the opportunity to
exercise skills in some of these areas also. Insights derived
from other realms can help us grow as physicians. Davidoff,
for one, described how music enriches our connection to our
own emotions, sensitizes us to being more empathic, and
From the Department of General Oncology, University of Texas M. D. Anderson Cancer

Address reprint requests to Michael J. Fisch, MD, MPH, Department of General
Oncology, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Unit
0410, Houston, TX 77030; email: mfisch@mdanderson.org.
1092-9118/10/1-10
e77
MICHAEL J. FISCH
Fig 1.
also provides insight on how clinical practice might be

learned, taught, and performed more effectively.20
Establishment of Meaningful Relationships
Seventeenth-century poet John Donne realized that no

man is an island entire of itself. Recognizing ones connections to others and reflecting on how to tap into and nurture
key relationships can get lost in the course of daily routine,
and we clinicians are indeed busy. Providers may face
burnout when they focus too heavily on their relationship
with their patients and lose sight of their colleagues, other
staff, and collaborators who are critical to success and
well-being. This increases vulnerability to mistakes, loss of
boundaries, feelings of professional impotence, and being
perceived as ineffective. The demands on providers are
particularly high early in their careers when they are trying
to establish their niche, skills, and reputation. This frequently coincides with young marriages and young children.
The work-family conflict that often ensues is clearly a risk
factor for burnout, and physicians must be aware that
attention, time, and open communication are necessary
KEY POINTS
e78
Stress and burnout are prevalent problems among

oncology providers.
The syndrome of burnout can adversely affect patient
care outcomes and reduce the health and fulfillment
of clinicians.
Buoyancy refers to the force, in opposition to burnout,
that uplifts the provider.
The individual parameters that constitute overall
buoyancy vary between individuals and also vary
within individuals over time (and sometimes day to
day).
Examining ones own buoyancy factors can be useful
to guide choices and habits in such a way that
maximizes buoyancy and minimizes risk of burnout.
Dimensions of buoyancy.
outside the work setting to sustain growth and fulfillment in

the workplace.21 One way to enhance success in this dimension is to learn communication skills and leadership skills.
Workshops on these topics are often available to early career
physicians but overlooked in favor of content-focused learning. These behavior-oriented workshops help with skills
such as authentic listening, giving and receiving feedback,
reserving judgment, noticing ones own emotions, negotiation, and other critical aspects of relating to others and
mindful awareness of oneself. Finally, some practices and
programs have created forums for self-disclosure and dialogue among physicians (see http://theheartofmedicine.org
for ideas and resources).
Being Awake to Your Present Reality
Oncology providers generally are good achievers and are

quite adept at striving for specific goals for themselves and
their patients. Being awake to your present reality amounts
to checking in with yourself about basic conditions that are
frequently nearly out of viewin ones life blind spot, as it
were. This might include simple matters such as I am tired.
I should rest now or I am frustrated. I should take a break
for a while, but an assessment could encompass larger
matters such as I am ill. I should see a doctor at this point
or even an appraisal of what one really wants out of life (not
just work or home in isolation). Honest appraisal and
regular assessment of your health, how you think and work,
and your expectations can pay major dividends in buoyancy.
Regarding health, taking note of your diet, the amount
and quality of your sleep, and whether aerobic and resistance exercise are part of your routine is a good starting
point. Willingness to get care from another physician is
important too, as this allows preventive care and health
maintenance as well as treatment for existing conditions.
Neglected health issues such as depression and substance
abuse are common among providers who suffer from impairment, disruptive behavior, or burnout.22
An appraisal of thinking and working habits is also
worthwhile, as these can also escape our notice and undermine well-being. It is worthwhile to consider whether you
work efficiently and delegate appropriately. Or do you tend
BUOYANCY TO PREVENT STRESS AND BURNOUT
to get almost all of your needs met by trying to help others

and neglecting yourself? Do you find yourself being cynical
or intolerant of fools, catastrophizing, or being overly
negative or pessimistic? Self-appraisal in these realms is
healthy. We learn about the concept of equipoise in clinical
trials and scientific experimentation, but it is equally critical
to introduce equipoise in our own ways of looking at things.
In other words, learn to resist becoming too enamored of
your own ideas and perspectives. Take a close look at your
position about an issue and think, Maybe so. Next, try to
explore alternate perspectives, ranging from one extreme to
the opposite and then back toward your own estimate of
reality. You may end up in the same placeyour position
but flexing your mind around an issue can help you find
greater balance and enhance your buoyancy.
Gratitude
The idea that it is healthy to count ones blessings is

widely understood. However, having the discipline and skill
to maintain this particular intention is not simple. One way
to accomplish this is to think deeply about the concept of
buoyancy and notice that there are indeed things that are
working in your life. As a further step, having noticed your
buoyancy and its roots, take a moment to actually feel
gratitude for achieving balance. Finally, act on gratitude by
speaking or writing words of appreciation. In my practice,
good habits are often easiest to internalize when I implement them at a patients bedside for the purpose of improving patient care. In that setting, start by noticing and
appreciating elements of your patients behavior. Patients
do amazing things all the time; it is not difficult to find
behaviors to comment positively about. Learning to say
goodbye to patients toward the end of life23 and writing
letters of condolence to family members24 are examples of
showing gratitude in clinical settings. The next step would
be extending this practice to your colleagues, staff, family,
and friends by saying thank you more frequently. Many
providers are surprisingly stingy when it comes to saying
thank you in professional and personal realms, perhaps as
a result of high expectations from staff and colleagues.
Expressing gratitude more often does not require lowering
expectations, and it can help bolster buoyancy.
Courage (Managing Fear)
Oncology providers, in general, have a great deal of

courage. It takes courage to compete in school and achieve
enough success to get trained in medicine and cancer care. It
takes courage to deal with the sacrifices inherent in medicine and the great responsibility (and privilege) of caring for
persons who are ill. But courage is not generic. One can be a
war hero and be frightened of snakes or bugs. If you think of
courage as fear management, then it makes sense that one
can be successful in some realms and still need work in
others. To guard against burnout and build buoyancy, providers need to learn how to recognize their fears, give those
fears a name, and reflect deeply on how to confront fear.
Common fears include fear of making mistakes, being
blamed or sued, or being judged or taken for granted by
colleagues, friends, or family. There may be fear of losing
income as a result of competition, of not being appropriately
recognized or promoted, or any number of other concerns
that are understandable in our modern environment. Other
issues revolve around countertransference and ones feelings
about personal mortality and death. Becoming aware of our

fears and developing habits and action plans for managing
them is critical. Introspection is critical. For many providers, connecting with others is preferable to tackling these
tasks alone.
Appreciation of Impermanence
To appreciate impermanence is to understand that all

things change and that loss is part of the package in life.
Being flexible and open to change is essential for personal
growth and equanimity. I may not relish change in all
instances (like my aging body or my changing bank account),
but I can see this truth. The idea that change is the nature
of all things is yet another topic that is easy to recite but
difficult to integrate on a practical level. Change is frequently associated with emotions, both positive and negative. People cry at weddings and funerals and graduations,
for instanceall events associated with transitions. How
often to you notice your own emotions? Do you notice when
you are getting angry? Are you aware of your own grieving?
In oncology, dealing with loss and grief is particularly
important.16,25 In his classic article Dealing with Our
Losses, Balfour Mount wrote, To live is to experience
loss.26 A mentor once told me, You have got to learn how to
lose. We experience loss not only with our patients facing
serious illness but also with changes, setbacks, and failures
of every variety: experiments that fail, grants not funded,
papers not accepted, organizational shifts that are not in our
favor, and the list goes on. As oncology providers, we should
strive to maintain our resilience while at the same time
finding a way to notice and manage our emotions. Although
people are variably resilient, there are related skills that can
be learned and practiced. Mindfulness training, for example,
has shown benefits for helping physicians become more
awake to their present reality as well as positioned to accept
the inevitability of change.8
Compassionate Mind Frame
It is a bit embarrassing to have been concerned with the

human problem all ones life and find at the end that one
has no more to offer by way of advice than try to be a little
kinder. Aldous Huxley
I think about compassion as something that starts with
empathy but requires buoyancy (like a coenzyme in a chemical reaction) to come to fruition in an objective way. Without
buoyancy, empathy would pull one toward the same place
as the subject of ones empathy. The good news, then, is that
all the things you do to stay buoyant position you for success
in creating and maintaining a compassionate mind frame.
The aspect of compassion that takes particular practice and
attention is empathy. Once again, a good place to start is
in the clinical realmlearning to act empathetically toward
your patients and, better yet, to feel empathy, too. There
are good reviews on this topic,27,28 and communication
workshops and time spent with master clinicians can help
promulgate this skill. The next step toward maximizing
the buoyancy derived from a compassionate mind frame is to
extend these skills outside the clinical setting to staff,
colleagues, students, family, friends, and strangers, and
then stretch this powerful mind frame all the way to oneself.
Being compassionate toward oneself is a beautiful way to
stay resilient and guard against the forces of burnout. A
e79
MICHAEL J. FISCH
good resource for exploring the principles and some specific

programs related to compassion in health care is the
Schwartz Center (www.theschwartzcenter.org).
Finding and Keeping Your Safety and Security
Some issues are difficult to put into a defined category but

nevertheless have a major effect on personal buoyancy, and
therefore are worthy of careful examination. One example is
money. Some people are more affected by money than
others depending on things like family size, level of preexisting debt, amount of wealth in the family, personal goals
and preferences, etc. Another example is location. For some
oncology providers, living near family (or a spouses family)
makes an enormous difference in the overall well-being of
the family, and thus has influence on the familys tolerance
of long work hours and other aspects of sacrifice that may be
called for within a particular job setting. In addition to
money and location, another factor affecting buoyancy is
religion and spirituality. The ability to integrate with a
religious community and worship on certain days is paramount for some providers. Factors such as these (and many
others that might fit this broad category) are highly variable
between individuals and sometimes vary within individuals
over time.
Answering the Question, Do I Matter?
Oncology providers make considerable sacrifices in the

course of training and also in the course of daily practice. To
maximize a sense of fulfillment, providers need to feel they
are making a difference. They want to matter to their
patients, to their organization or team, and in various
groups and relationships of importance in work, family, and
community settings. A loss of buoyancy is sometimes best
understood by deep reflection and unmasking a Do I matter? crisis in some dimension. Such a crisis can also focus on
the ultimate, existential dimension.
Conclusion
Stress and burnout are prevalent problems among oncology providers, and these problems have a substantial effect
on patient care outcomes and provider health and fulfillment. Examining ones own buoyancy factors can be useful
to guide choices and habits in such as way that maximizes
buoyancy and minimizes risk of burnout.
Acknowledgments
The author gratefully acknowledges Ms. Danielle Walsh for
providing the graphics for the figure, and Ms. Joann Aaron and
Mr. Bryan Tutt for providing editorial assistance.
Author
Michael J. Fisch
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Elsevier (L)
Research
Funding
Expert
Testimony
Other
Remuneration
Biosite
REFERENCES
1. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annual Rev Psych.
2001;52:397-422.
2. Allegra CJ, Hall R, Yothers G. Prevalence of burnout in the U.S.
Oncology community: results of a 2003 survey. J Oncol Pract. 2005;1:140-147.
3. Whippen DA, Canellos GP. Burnout syndrome in the practice of oncology: results of a random survey of 1,000 oncologists. J Clin Oncol. 1991;9:
1916-1920.
4. Liakopoulou M, Panaretaki I, Papadakis V, et al. Burnout, staff support,
and coping in pediatric oncology. Support Care Cancer. 2008;16:143-150.
5. Mukherjee S, Beresford B, Glaser A, et al. Burnout, psychiatric morbidity, and work-related sources of stress in paediatric oncology staff: a review of
the literature. Psycho-oncology. 2009;18:1019-1028.
6. Balch CM, Shanafelt TD, Sloan JA, et al. Distress and career satisfaction
among 14 surgical specialties, comparing academic and private practice
settings. Ann Surg. 2011;254:558-568.
7. Kuerer HM, Eberlein TJ, Pollock RE, et al. Career satisfaction, practice
patterns and burnout among surgical oncologists: report on the quality of life
of members of the Society of Surgical Oncology. Ann Surg Oncol. 2007;14:
3043-3053.
8. Krasner MS, Epstein RM, Beckman H, et al. Association of an educational program in mindful communication with burnout, empathy, and
attitudes among primary care physicians. JAMA. 2009;302:1284-1293.
9. Bram PJ, Katz LF. A study of burnout in nurses working in hospice and
hospital oncology settings. Oncol Nurs Forum. 1989;16:555-560.
10. Blanchard P, Truchot D, Albiges-Sauvin L, et al. Prevalence and causes
of burnout amongst oncology residents: a comprehensive nationwide crosssectional study. Eur J Cancer. 2010;46:2708-2715.
11. Shanafelt TD. Enhancing meaning in work: a prescription for preventing physician burnout and promoting patient-centered care. JAMA. 2009;302:
1338-1340.
12. West CP, Shanafelt TD, Kolars JC. Quality of life, burnout, educational
debt, and medical knowledge among internal medicine residents. JAMA.
2011;306:952-960.
13. Dyrbye LN, Massie FS, Jr., Eacker A, et al. Relationship between
burnout and professional conduct and attitudes among US medical students.
JAMA. 2010;304:1173-1180.
e80
14. Broffman G. Controlled burn! Physician executives must be ready to

handle job burnout, career stress. Physician Executive. 2001;27:42-45.
15. Johns MM III, Ossoff RH. Burnout in academic chairs of otolaryngology: head and neck surgery. Laryngoscope. 2005;115:2056-2061.
16. Lyckholm L. Dealing with stress, burnout, and grief in the practice of
oncology. Lancet Oncol. 2001;2:750-755.
17. Shanafelt T, Chung H, White H, et al. Shaping your career to maximize
personal satisfaction in the practice of oncology. J Clin Oncol. 2006;24:40204026.
18. Horowitz CR, Suchman AL, Branch WT Jr., et al. What do doctors find
meaningful about their work? Ann Intern Med. 2003;138:772-775.
19. Armstrong J, Holland J. Surviving the stresses of clinical oncology by
improving communication. Oncology. 2004;18:363-375.
20. Davidoff F. Music lessons: what musicians can teach doctors (and other
health professionals). Ann Intern Med. 2011;154:426-429.
21. Dyrbye LN, West CP, Satele D, et al. Work/home conflict and burnout
among academic internal medicine physicians. Arch Intern Med. 2011;171:
1207-1209.
22. Brown SD, Goske MJ, Johnson CM. Beyond substance abuse: atress,
burnout, and depression as causes of physician impairment and disruptive
behavior. J Am Coll Radiol. 2009;6:479-485.
23. Back AL, Arnold RM, Tulsky JA, et al. On saying goodbye: acknowledging the end of the patient-physician relationship with patients who are
near death. Ann Intern Med. 2005;142:682-685.
24. Bedell SE, Cadenhead K, Graboys TB. The doctors letter of condolence.
N Engl J Med. 2001;344:1162-1164.
25. Meier DE, Back AL, Morrison RS. The inner life of physicians and care
of the seriously ill. JAMA. 2001;286:3007-3014.
26. Mount BM. Dealing with our losses. J Clin Oncol. 1986;4:1127-1134.
27. Back A, Arnold RM, Tulsky JA. Mastering Communication with Seriously Ill Patients: Balancing Honesty with Empathy and Hope. New York:
Cambridge University Press; 2009.
28. Coulehan JL, Platt FW, Egener B, et al. Let me see if I have this
right . . .: words that help build empathy. Ann Intern Med. 2001;135:221227.
Encountering Grief in Patient Care

By Teresa Gilewski, MD
Overview: Grief is essentially unavoidable and is a normal

reaction to loss. Grief may be experienced by patients and
their loved ones as well as by physicians and members of the
health care team in response to the consequences of illness or
death. Grief is typified by certain indicators that may significantly effect ones emotional and physical well-being. Although these indicators tend to follow a general pattern, there
is variability among individuals. Complicated grief may require
MILY DICKINSON wrote, I measure every grief I

meet with analytic eyesI wonder if it weighs like
mine or has an easier size.1 For most humans, grief
is unavoidable. Physicians encounter grief in their patients
and their patients loved ones but may also experience grief
on a personal level in response to caring for a patient. In the
practice of medicine, grief is often associated with death but
may also be experienced in response to loss from the ravages
of illness. It is valuable for physicians to explore the process
of grief, not only to enhance patient care but also to optimize
their own well-being. Grief has the potential to connect us to
our humanity.
The Meaning of Grief
A common definition of grief is a deep and poignant

distress caused by or as if by bereavement, where bereavement refers to suffering the death of a loved one.2 However, the meaning of grief, may also connote a broader
application. In the Education in Palliative and End-of-Life
Care (EPEC) Project oncology core module regarding loss,
grief, and bereavement, grief is defined as the experience of
a loss and bereavement as the state of living with a loss.3
Although grief is most often associated with a loss from
death, it is important to recognize that loss is present
throughout the course of illness. There may be a multitude
of losses that effect nearly every aspect of life for the patient
and their loved ones.3
Some of these losses may be easily recognizable, such as
the loss of ones usual physical appearance and bodily
functions. With illness, the usual roles of the patient and his
or her loved ones often change. The former caregiver may
now become the patient, thus threatening the equilibrium of
former relationships. The alterations in these social interactions between the patient and members of his or her
community may lead to a sense of loss for the familiarity of
prior responsibilities. In addition, the financial costs of
illness, including expenses and lost wages, may induce a
tangible loss of relative monetary stability. A study of 988
patients with various terminal illnesses (51.8% with cancer)
found that nearly 35% had moderate to high care needs (e.g.,
required assistance with transportation, personal and nursing care, and homemaking).4 These patients reported a
greater economic burden than those with low care needs.
There are also the subtle losses that may develop on an
emotional or philosophical level, such as the loss of expectations for the future, the loss of fulfillment of dreams, and the
loss of a common human perspective that death is somehow
far away. When death becomes imminent or occurs, it tends
to shake the foundation of how we view life.
psychiatric intervention. Caring for the seriously ill or dying

patient may be particularly challenging from an emotional
level and may increase the risk of burnout. Recognition of
these emotions is a critical aspect of providing compassionate
care on a sustainable level. Various strategies may be beneficial in coping with grief, and the exploration of grief may
provide greater insight into the humanistic basis of medicine.
Physicians are privy to some of these losses in their

interactions with patients and their loved ones. However,
physicians may also experience their own grief in response
to caring for patients with such losses. Physicians themselves may feel a sense of loss or distress regarding ineffective treatments, an inability to adequately respond to the
patients suffering, or the death of a patient.
Stages and Types of Grief
Grief consists of multiple reactions, including behavioral,

psychologic/emotional, social, spiritual, and physical changes
(Table 1).3,5 Various characteristics of the mourning processproposed by Sigmund Freud, Eric Lindemann, Colin
Parkes, and John Bowlby have been essential to an understanding of grief and bereavement.6 Elisabeth Kubler-Ross,
MD, popularized five stages of grief in her classic book On
Death and Dying, published in 1969.7 When faced with
death because of an incurable illness, she observed that
patients commonly experienced denial, anger, bargaining,
depression, and acceptance. However, these stages may be
of various durations and intensities. Individuals may not
experience all of these stages, nor may they occur in the
same order. Since then, further evaluation has provided
greater insight into the patterns of grief either normal and
uncomplicated or complicated.
The Yale Bereavement Study queried 233 bereaved individuals whose family member or loved one died of natural
causes.8 Over a period of 24 months, five grief indicators
were rated: disbelief, yearning, anger, depression, and acceptance. Yearning was the most often reported negative
indicator with a peak value at 4 months post-loss. Disbelief
peaked at 1 month post-loss, anger peaked at 5 months
post-loss, and depression at 6 months post-loss. Interestingly, acceptance was the most commonly reported factor
and continued to increase during the duration of the study.
This trajectory is most consistent with an uncomplicated
or normal grief reaction. However, significant expression
of negative emotions after 6 months may indicate a more
complex grief reaction that requires additional assessment.
Complicated grief comprises a unique entity that signifies
a greater level of dysfunction and portends a greater risk of
morbidity. Characteristic features may occur in anticipation

Address reprint requests Teresa Gilewski, MD, Memorial Sloan-Kettering Cancer Center,
300 East 66th St., New York, NY; email: gilewskt@mskcc.org.
1092-9118/10/1-10
e81
TERESA GILEWSKI
Table 1. Some Potential Indicators of Grief
3,5,22
Psychologic/Emotional
Disbelief
Anxiety
Sadness
Numbness
Anger
Behavioral
Searching/yearning
Difficulty with concentration
Physical
Anorexia
Fatigue
Palpitations
Difficulty sleeping
Hair loss
Weight change
Gastrointestinal disturbance
Social
Social withdrawal
Isolation
Spiritual
Questioning beliefs
of death or following death and include excessive bitterness,

yearning, emptiness, numbness, and inability to cope with
the loss.9 An individuals personality traits may contribute
to the development of complicated grief reactions. For some,
grief may be chronic, delayed, exaggerated, or masked and
may necessitate psychiatric intervention.3
In their professional role, physicians awareness of the
emotions experienced by patients and their loved ones in the
grieving process is essential. Physicians may also grieve for
their patients losses or following their patients deaths, in a
series of minigrief reactions.10
The Effect of Patient Care and Grief
on the Physician
General Effect
As human beings, physicians instinctively develop emotions about their patients. It is the responsibility of the
physician to acknowledge that these feelings are an integral
KEY POINTS
e82
In the medical field, grief is a normal response to loss

from the consequences of illness or death.
Grief is characterized by emotional and physical
indicators and may be uncomplicated or complicated;
complicated grief causes significant dysfunction and
requires further psychiatric intervention.
Grief is experienced by the patient and their loved
ones as well as by the physician and other members of
the health care team.
The expectations and goals of the physician in regard
to patient care and the repeated exposure to death
may contribute to burnout.
Coping with grief requires acknowledgement of ones
emotions and a focus on strategies to optimize overall
well-being.
11
part of the doctorpatient relationship. A key challenge is

finding a balance where the physician feels for a patient
but not to the extent that it becomes so overwhelming the
physician cannot empathize with the next patient.12
Meier et al emphasized that these emotions, if left unexamined, may negatively affect patient care and the wellbeing of the physician.13 They propose that recognition of
risk factors that significantly influence an emotional response and awareness of the resultant behaviors and emotions are important. Once the emotions are identified and
accepted, further reflection on the emotion and discussion
with a trustworthy colleague may enhance the ability of
physicians to reach an emotional balance that is sustainable
and, thus, minimize the risk of burnout and stress. This is of
particular relevance when the physician encounters grief on
a frequent basis, as in the field of oncology.
Physician Response to Patient Death
There is a surprisingly limited amount of information on

the physicians response to patient deaths. A cross-sectional
study of two academic hospitals in the United States evaluated the reactions of 188 physicians, including attendings,
residents, and interns to the death of their patients.14 There
were 68 evaluable inpatient deaths that occurred on a
general medicine or intensive care unit service. Only 12% of
the physicians had cared for the patient before this hospitalization; 62% characterized their connection with the patient as not close.
Most physicians had satisfying experiences with their
patients. On average they experienced 2 of 14 indicators of
grief, with feeling upset or numb the most frequent,
while approximately 23% described the death as very
disturbing. Grief was associated with longer periods of
caring for the patient and was more common in female
physicians. There was no significant distinction in emotional
response based on level of training, although interns wanted
more emotional support than attendings. Less than 25% of
the interns and residents viewed the attending as the most
valuable source of assistance in coping.
A study from the United Kingdom surveyed physicians
regarding their last memorable patient death. This was
described as one that caused reflection and happened in
the few months before the study.15 In a mix of house officers
and attending staff, the majority reported coping satisfactorily after the patients death. Nearly 84% favored talking
with others as a method of coping, and approximately 57%
supported additional education on this topic. Approximately
44% reported feelings of moderate to severe sadness.
These studies raise several interesting topics for further
evaluation. Historically, the primary physician often followed the patient as both inpatient and outpatient. However, with the increasing use of hospitalists, the patients
primary physician may have limited contact with the patient during the final inpatient hospitalization. The significance of this factor as well as the location of the patients
death (e.g., hospital, home, hospice center) on the primary
physicians emotional response is unclear. These studies
were not specific to oncology, so oncologists who frequently
care for patients over extended time periods may be at an
increased risk of grief. In addition, the role of the attending
physician in discussing grief with the house staff is likely
underdeveloped.
ENCOUNTERING GRIEF IN PATIENT CARE

Focus on Burnout
Grief and continuous exposure to death may contribute to

burnout. The key features of burnout identified by Maslach
et al include exhaustion, depersonalization (cynicism), and
ineffectiveness, and these symptoms are primarily job related.16 Factors that may increase the risk of burnout
include younger age, single marital status, low self-esteem,
less hardy character trait, avoidant coping approach, belief
that others or chance are in control, and possibly job attitude. Burnout affects patient care as well as the individual
physician.
Whippen and Canellos surveyed 1,000 oncologists regarding burnout, 60% of whom were medical oncologists.17 In the
598 completed responses, overall 56% of oncologists endorsed burnout. The presence of burnout was similar across
the subspecialties with 58% for medical oncologists, 48% for
surgical oncologists, 44% for pediatric oncologists, and 52%
for radiation oncologists. The most common potential causes
of burnout were insufficient personal/vacation time (57%),
continuous exposure to fatal diseases (53%), and frustration
of ineffective therapies (45%). Interestingly, approximately
80% felt that their career fulfilled the expectations they had
in training.
A high level of burnout was observed in another study of
261 house staff, nurses, and oncologists.18 The most common
stressor for burnout was unfavorable work events, such as
frequent exposure to death and controversies over do not
resuscitate status. Therefore, health care workers may be
predisposed to stress and burnout from a variety of factors,
including grief and exposure to patient suffering.19-21
Possible Benefits
Health care professionals can learn much from the patients who face death and their loved ones, if they are open
to listening. Despite the personal toll that results from
encountering grief on a regular basis, physicians and other
health care workers may achieve a profound fulfillment in
the knowledge that their contribution helped ease suffering.
This may favorably affect job satisfaction.22,23 From a sociologic perspective, there is evidence to suggest that giving
support may be more favorable to ones health and wellbeing than receiving support.24
The bereaved may also benefit from engagement of the
physician. In a study of patients with terminal illness and
considerable care needs, their caregivers conveyed the importance of the interaction with the physician.4 If the physician listened to them about the patients illness, they were
less likely to feel depressed. Bereavement outcomes may
also be affected by physician communication.25
Facing Grief: What to Do
The Perspective of the Oncologist and Medical Team
As Mount succinctly noted, To practice oncology . . . is to

augment our losses many-fold, for they are an integral part
of our professional existence.19 Grief, therefore, is a normal
part of an oncologists life. The overall perspective of the
oncologist toward these losses may be a crucial component of
the ability to cope with grief. If the oncologist views a cure or
prolongation of life as the only significant goals of patient
care, then he or she may miss a myriad of opportunities to
alleviate suffering from loss and to achieve greater insight

into the meaning of life. With this viewpoint, the physician
may particularly struggle when dealing with grief.
It is not easy to be reminded of ones mortality on a
frequent basis. It is not easy to grieve or to watch another
grieve. Yet these experiences have the potential to ultimately inspire and lead to a broader understanding of the
consequences of illness. A major challenge is to apply a
realistic approach to attain this idealistic ambition.
Realistic Suggestions to Cope with Grief
Grief is a normal and universal reaction to loss; engaging ones grief may be far more beneficial than avoiding it.
There are numerous strategies that can be used to work
through stress and grief. Several authors suggest that it
is crucial for the physician to be aware of his or her
emotions.13,19 In addition, it is important to set reasonable
work goals, enhance time management skills, and limit
personal involvement with patients.19,20 Integration of other
medical team members (i.e., nurses, chaplains, social workers) may help to optimize care of the whole patient and
alleviate some of the emotional burden for the individual
physician and other medical caregivers. It may be beneficial
to foster a supportive work environment that encourages
discussion of these issues in a group or individual setting,
such as Death Rounds.26
On a more personal level, a variety of suggestions for
coping include habitual exercise, optimal rest, interests
outside of work, writing ones thoughts, reading others ideas
about these issues, laughter, and a focus on inner growth,
such as mindful meditation and expansion of ones view of
life to reflect a more global/philosophic perspective.20,23 In
particular, the field of narrative medicine employs reflective
writing to explore the physicians relationship with self,
patients, colleagues, and society.27 Physicians may also turn
to their own spiritual beliefs for support.
A classic symbol of grief is tears. In response to sadness,
does a physician crying indicate weakness? There may be
various responses to this question.28 However, it may be
completely appropriate to cry for a patient or the loss of a
patient. In fact, it may be viewed by the patient or family as
a sign of compassion.29
Bereavement Care by the Physicians
Bereavement care initiated by the physician may be very

meaningful to patients families.30 In addition, it may improve physician well-being and job satisfaction, although
further research is necessary to definitively confirm an
association.22
Bedell et al noted that in the 1800s, it was common for
physicians to send a letter of condolence to a patients
family.31 However, with the changing times, this practice
has diminished. Yet, for the bereaved, the words of the
physician may provide comfort and exemplify a more humanistic approach to patient care. The authors suggest that
the letter include an articulation of the physicians sympathy, a personal reference to the patient, and an acknowledgment of the familys devotion to the patient. Obviously, only
sincere thoughts and emotions should be included.
A recent Canadian survey of medical oncologists, radiation oncologists, and palliative care physicians focused on
bereavement practices.32 The authors evaluated how often
the 535 physicians practiced one of the following: making a
e83
TERESA GILEWSKI
telephone call to the family, sending a sympathy card, or

attending a funeral. Approximately 33.3% (95% CI 29.3
37.4%) reported usually or always, 30.5% (95% CI 26.5
34.4%) reported sometimes, and 36.2% (95% CI 32.1 40.3%)
reported rarely or never. The most common practice was a
phone call whereas attending a funeral was rare. A small
number of physicians, approximately 9% (95% CI 6.6
11.5%), felt that sending a condolence note was the responsibility of the physician.
The most common identified barrier to contacting the
patients family was lack of time and resources, although
some felt uncomfortable and were not certain whom to
contact. Multivariate analysis revealed that more frequent
bereavement practice was associated with female sex, working in an academic center, conviction that physicians have
a responsibility to write a condolence card, and absence of a
palliative care program, with the most significant being a

palliative care specialty.
For some patients families, it may also be meaningful to
visit with them in person. Physicians may not recognize the
importance of these bereavement practices because of limited feedback from the patients loved ones.
Conclusion
Grief is a normal response to loss and a fundamental

component of the field of oncology. Health care workers in
oncology face the challenge of repeated exposure to serious
illness and death. Enhanced acknowledgment and recognition of the indicators of grief and reasons for grief are
essential to providing sustainable compassionate care. Engaging ones grief may result in considerable insight into the
essence of patient care and the soul of medicine.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Teresa Gilewski*
REFERENCES
1. Dickinson, E. The Complete Poems of Emily Dickinson. Boston: Little
Brown; 1924.
2. Merriam-Webster. www.merriam-webster.com. Accessed February 6,
2012.
3. Emanuel LL, Ferris FD, von Gunten DF, Von Roenn J. Education in
Palliative and End-of-Life Care - Oncology (EPEC-O): Participants Handbook. Module 4: Loss, Grief and Bereavement. EPEC Project. Chicago, IL:
2005;M4-1-M4-24.
4. Emmanuel EJ, Fairclough DL, Slutsman J, et al. Understanding economic and other burdens of terminal illness: The experience of patients and
their caregivers. Ann Inter Med. 2000;132:451-459.
5. Casarett D, Kutner JS, Abrahm J, et al. Life after death: A practical
approach to grief and bereavement. Ann Intern Med. 2001;134:208-215.
6. Woof WR, Carter YH. The grieving adult and the general practitioner:
A literature review in two parts (Part 1). Br J Gen Pract. 1997;47:443-448.
7. Kubler-Ross E. On Death and Dying. New York, NY, Macmillan, 1969.
8. Macicjewski PK, Zhang B, Block SD, et al. An empirical examination of
the stage theory of grief. JAMA. 2007;297:716-723.
9. Tomarken A, Holland J, Schachter S, et al. Factors of complicated grief
pre-death in caregivers of cancer patients. Psychooncology. 2008;17:105-111.
10. Holland JC. Management of grief and loss: Medicines obligation and
challenge. J Am Med Womens Assoc. 2002;57:95-96.
11. Gorlin R, Zucker HD. Physicians reactions to patients: A key to
teaching humanistic medicine. N Engl J Med. 1983;308:1059-1063.
12. Novack DH, Suchman AL, Clark W, et al. Calibrating the physician:
Personal awareness and effective patient care. JAMA. 1997;278:502-509.
13. Meier DE, Back AL, Morrison RS. The inner life of physicians and care
of the seriously ill. JAMA. 2001;286:3007-3014.
14. Redinbaugh EM, Sullivan AM, Block SD, et al. Doctors emotional
reactions to recent death of a patient: Cross sectional study of hospital
doctors. BMJ. 2003;327:185-189.
15. Moores TS, Castle KL, Shaw KL, et al. Memorable patient deaths:
reactions of hospital doctors and their need for support. Med Edu. 2007;41:
942-946.
16. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annu Rev Psychol.
2001;52:397-422.
e84
17. Whippen DA, Canellos GP. Burnout syndrome in the practice of

oncology: Results of a random survey of 1,000 oncologists. J Clin Oncol.
1991;9:1916-1921.
18. Kash KM, Holland JC, Breitbart W, et al. Stress and burnout in
oncology. Oncology. 2000;14:1621-1633.
19. Mount BM. Dealing with our losses. J Clin Oncol. 1986;4:1127-1134.
20. Lyckholm L. Dealing with stress, burnout, and grief in the practice of
oncology. Lancet Oncol. 2001;2:750-755.
21. Ramirez AJ, Graham J, Richards MA, et al. Mental health of hospital
consultants: The effects of stress and satisfaction at work. Lancet. 1996;347:
724-728.
22. Block SD. Psychological considerations, growth, and transcendence at
the end of life: the art of the possible. JAMA. 2001;285:2898-2905.
23. Kearney MK, Weininger RB, Vachon MLS, et al. Self-care of physicians
caring for patients at the end of life: Being connected . . . a key to my
survival. JAMA. 2009;301:1155-1164.
24. Brown SL, Nesse RM, Vinokur AD, et al. Providing social support may
be more beneficial than receiving it: Results from a prospective study of
mortality. Psychol Sci. 2003;14:320-327.
25. Prigerson HG, Jacobs SC. Caring for bereaved patients: All the
Doctors Just Suddenly Go. JAMA. 2001;286:1369-1376.
26. Smith L, Hough CL. Using death rounds to improve end-of-life education for internal medicine residents. J Palliative Medicine. 2011;14:5558.
27. Charon R. Narrative medicine: a model for empathy, reflection, profession, and trust. JAMA. 2001;286:1897-1902.
28. Siegel B, Schultz S. Crying in stairwells: How should we grieve for
dying patients? JAMA. 1994;272:659.
29. Gilewski T. The subtle power of compassion. JAMA. 2001;286:30523053.
30. Penson RT, Green KM, Chabner BA, et al. When does the responsibility
of our care end: Bereavement. Oncologist. 2002;7:251-258.
31. Bedell SE, Cadenhead K, Graboys TB. The doctors letter of condolence.
N Engl J Med. 2001;344:1162-1164.
32. Chau NG, Zimmermann C, Ma C, et al. Bereavement practices of
physicians in oncology and palliative care. Arch Intern Med. 2009;169:963971.
TALKING IN SYNC WITH PATIENTS ACROSS

CULTURAL BARRIERS
CHAIR
Lidia Schapira, MD
Boston, MA
SPEAKERS
William J. Hicks, MD
Columbus, OH
Louise Villejo, MPH
Houston, TX
New Insights in Cross-Cultural Communication

By Lidia Schapira, MD
Overview: Improving clinician-patient communication, improving clinical decision making, and eliminating mistrust
have been identified as three key areas for reducing disparities in care. An important step is the training of cancer
professionals to deliver culturally competent care in clinical
settings as well as increasing the proportion of underrepresented minorities in the health care workforce. Providing care
that is attuned to the patients cultural preferences begins by
talking to the patient about his or her cultural history and
identifying the locus of decision making, preferences for
disclosure of vital health information, and goals of care.
Patients with low literacy and those with poor fluency of the
HERE IS growing interest in providing culturally sensitive care, not only in the United States but also across
the world. Awareness of the global repercussions of cancer
and the enormous burden of suffering imposed mainly on
those with little access to health care and few resources has
kindled the imagination and determination of cancer physicians and led to new global initiatives to close the gap
between low- and middle-income countries and those that
are more affluent.1 Although once considered a problem
exclusive to industrialized and wealthy countries, cancer is a
leading cause of disability and death in the developing
world.1 Multinational and interdisciplinary efforts with support from professional organizations and industry are beginning to target global issues through innovative mechanisms
that involve raising awareness and promoting universal
access to cancer prevention, detection, and care.1 And within
countries, governments and nongovernmental organizations
need to meet the challenge of expanding access for those
who, for complex socioeconomic factors ranging from poverty
to low health literacy, are deprived of access to optimal
health services. Without such efforts, those with few resources will bear a disproportionate burden of misery from
cancer.
Many factors contribute to the unequal distribution and
utilization of health services, and this, in turn, leads to
differences in outcomes. A report commissioned by the U.S.
Congress and published by the Institute of Medicine in 2002
confirmed that members of groups identified as belonging to
ethnic and racial minorities have worse health outcomes as
a result of unequal treatment.2 The report stressed the need
for change and identified opportunities for interventions.
Three key areas were identified as having clinical relevance
and contributing to disparities in health care: providerpatient communication, clinical decision making, and mistrust.3 Recommendations for change addressed many
aspects of both care and health services delivery, and three
are worth highlighting: integrating cross-cultural education
into the training of all health care professionals, supporting
the use of language interpretation services in clinical settings, and increasing the proportion of underrepresented
minorities in the health care workforce.2,3 Other factors that
were identified as contributors to unequal outcomes are
social determinants such as lower levels of education, overall lower socioeconomic status, unsafe housing, and exposure to environmental hazards. The report also stated that
bias, prejudice, and stereotyping on the part of health care
dominant language require additional services. Language interpretation by trained professionals is fundamental to ensure
that patients are able to provide informed consent for treatment. A working definition of culture involves multiple dimensions and levels and must be viewed as both dynamic and
adaptive, rather than simply as a collection of beliefs and
values. Effective cross-cultural education avoids stereotyping
and promotes communication and negotiation to solve problems and minimize tension and conflict. Recent research has
identified that unconscious biases held by clinicians affect
their behavior and recommendations for treatment.
professionals exert important influences in their recommendations and treatment.3 The inherent power asymmetry in
the clinician-patient relationship coupled with unconscious
biases can lead to differential treatment of those seeking
care.
Communication is a complex phenomenon that serves to
transmit messages and construct and maintain relationships. It is one of the pillars of medicine and plays a
fundamental role in the delivery of health care services. Now
a focus of research in social sciences and medicine, it has
gained the status of a necessary competency for medical
trainees in the United States. Although there is consensus
among experts that a basic skill set can be taught and
evaluated, there is still no consistent way to address individual beliefs or deeply held biases that undoubtedly influence practice. Simply put, we can transmit and evaluate
knowledge about communication, construct ethical frameworks, and even model ideal behaviors, but we cannot
control or even modify deeply held beliefs that affect behavior.
Evidence suggests that individual doctors responses tend
to be similar with all patients and that moral judgments and
decision making are driven by automatic emotional
responses.4-8 Rather than resorting to abstract reasoning,
doctors and nurses respond to moral dilemmas with gut
reactions.5-8 Recognizing this fact, we need to address these
instant and unconscious reactions, which most likely result
from intergenerational transmission in families and communities, by designing courses that address cross-cultural communication. In a study of internal medicine and emergency
medicine residents at four academic medical centers in
Atlanta, GA, and Boston, MA, Green and colleagues found
evidence of unconscious bias.9 Most physicians in their
survey did not admit to any racial biases explicitly. However, using validated tools that measure implicit measures,
there were clear differences favoring one group over another, which influenced the physicians recommendations for
treatment in a simulated scenario.9 It seems plausible that
From the Massachusetts General Hospital, Boston, MA.

Address reprint requests to Lidia Schapira, MD, Massachusetts General Hospital, 55
Fruit Street, Boston, MA 02114; email: lschapira@partners.org.
1092-9118/10/1-10
e85
LIDIA SCHAPIRA
unconscious likes and dislikes influence our daily practice.

Feelings of disgust or kinship may affect our judgment and
recommendations.
Culture is a multilevel, multidimensional, dynamic, and
adaptive system and not merely a collection of beliefs and
values.10 Kagawa-Singer reminds us that efforts to apply
the concept of culture accurately and usefully in medical
practice require an assessment at many different levels
and across time in order to capture a composite of the
lived world of the patient and family. Culture is a layered
concept that is shaped by several components such as
environment, economy, technology, religion and worldview,
language, social structure, beliefs, and values.10 From a
purely practical perspective, we can assert that culture
affects the entire way in which cancer is framed in meaning
and response.10 Culture affects the experience and expression of pain, the locus of decision making, the etiquette of
communication with clinicians, the preference for information disclosure and participation in decision making, the
choices for end-of-life care, and expectations.10-12 An appreciation of the role and influence of culture is fundamental in
any model of care that is based on respect for individual
patients.
Societies differ in how they differentiate cultural groups
and how they respond to those who are perceived to be
different or not fully assimilated into the dominant culture.10 Some place great emphasis on religious differences or
nationality. In the United States, often cultural differences
are established by the proxies of racial and ethnic groups.
The Office of Management and Budget Directive codified
these categories but acknowledged that they are socialpolitical categories.10 Just think about the broad category of
non-Hispanic whites, which lumps together immigrants
from Iraq and Israel, from Italy and Russia. What sense can
we possibly make of statistics that aggregate these groups
that represent communities with vastly different historic
traditions, cultures, and beliefs? And yet they have been
used extensively in biomedical research and continue to be
used in medicine.
In a thoughtful essay that reviews the state of knowledge
and thinking about race, ethnicity, and genetics, Francis
Collins points out that race and ethnicity are terms
without generally agreed-on definitions.13 Both carry complex connotations that reflect culture, history, socioeconom-
KEY POINTS
e86
Integrating cross-cultural education into the training

of cancer professionals is crucial to reduce disparities.
Culture affects the way in which cancer is framed and
affects the experience of patients and families.
Cross-cultural communication begins with selfawareness and respect for a patients worldview,
beliefs, and values.
By talking with a patient about his or her cultural
history we can establish that patients preference for
decision making and desire for information.
Care that is culturally sensitive requires excellent
communication and negotiation skills and expert consultation to mediate conflict.
ics, and political status, as well as a variably important

connection to ancestral geographic origins. Those ancestral
origins have at best a hazy connection to current issues of
health disparities, although they may well account for the
unequal distribution of disease-associated alleles for certain
recessive disorders or susceptibility to some cancers.13 The
debate intensifies because race and ethnicity in the view of
many are self-reported constructs and thus the relationship
to disease or genetic risk is at best imperfect. Collins offers
sound and conciliatory advice: Without discounting selfidentified race or ethnicity as a variable correlated with
health, we must strive to move beyond these weak surrogate
relationships and get to the root causes of health and
disease, be they genetic, environmental, or both.13 We can
appreciate and respect the interdisciplinary inquiry and
research into race and ethnicity and move past this debate
in order to focus on interventions that have the potential of
reducing barriers and enhancing the experience of patients
considered underserved.
Cross-cultural communication begins with awareness of
our own moral code and beliefs. As long as medical care is
delivered by humans rather than robots or computers,
communication models need to recognize that health care
professionals are not interchangeable figurines. A patients
experience of care is a composite of many different meetings influenced by emotion and mediated by individual
communication styles. From the patients perspective, the
medical culture itself is challenging to decipher! We may
take for granted a shared respect for the scientific
method and for recommendations based on evidence and are
accustomed to professional hierarchies that vary between
specialties and settings. Our patients, however, even those
with superb education and intellectual abilities, are sometimes surprised by our protocols, rituals, guidelines, and
practices.
Cultural competence begins with respect, curiosity, humility, and empathy. The ideal mindset is that of a first grader,
open to wonder. Specific knowledge about cultures is very
valuable but often unavailable, especially at short notice.
We must rely on learned communication skills to create a
respectful environment where information is exchanged and
dialogue is used to negotiate solutions that serve the patient
well while minimizing conflict, frustration, and tension. In
situations where cultures clash, the best strategy is to
involve cultural brokers who can mediate and help clarify
goals and expectations. Thus doctors need more than a skill
set and ethical framework, they also need inspiration and
imagination to respond to calamity and support patients in
difficult situations so that the emphasis is on what really
matters to the patient.14
Providing care that is attuned to the patients cultural
preferences begins by talking to a patient about his or her
cultural history. This entails eliciting the patients understanding of his or her illness and its causation and course,
his or her preference for involvement in decision making,
and his or her need for varying amounts of information.
Through this process one can also identify the locus of
decision making, which often involves other family members. By showing respect and interest and by providing a
comforting presence, clinicians can build trust in complicated relationships. Without trust, we remain limited in our
ability to provide guidance in times of crisis. Consider for
example that preferences for end-of-life care are often dic-
CROSS-CULTURAL COMMUNICATION
Sidebar. Tips for Practice to Overcome Language

and Literacy Barriers
To overcome language barriers, obtain support from

colleagues or institutional leadership to provide
professional interpreter services. There are national standards for interpreters, but the quality of
services and depth of training are quite variable.
You may want to consider offering training for
interpreters to familiarize them with terms often
used in complex oncology consultations and in clinical research. Remember that using minors for
interpretation is illegal and that family members
often distort the content of conversations. In the
United States, health care organizations are obligated by law to provide language assistance services either in person or via remote interpretation
services at no cost to patients with limited English
proficiency.22-24
To overcome literacy barriers, check the patients
reading level and explore any possible learning
disabilities that could interfere with proper use of
medication or compromise the ability to discuss
options for treatment. About one-half of Americans
are considered to have limited health literacy,
which affects their understanding of basic medical
terms, their ability to follow directions for diagnostic procedures and therapies, give their consent for
research, and engage in a real dialogue about treatment options.
tated by cultural norms and influenced by beliefs in an

afterlife. Knowing and understanding these issues in advance allow physicians to provide direction as death
approaches. Trust has been described as an iterative process, requiring steadiness and honesty.15 In fact, distrust is
often the cause of misunderstandings or miscommunication
and may contribute to disparate outcomes.15 Breaches of
trust committed decades ago by medical researchers in the
United States still haunt relationships with African Ameri-
cans patients, who may harbor suspicions about the integrity of research and refuse to participate in clinical trials.16
Similar doubts about the motives of researchers are also
expressed by patients who have experienced discrimination
or harbor concerns regarding the privacy of personal health
information.
There has been considerable interest among philosophers,
psychologists, anthropologists, and ethicists in studying
truth-telling and disclosure of both diagnosis and prognosis.
In the United States and many Western countries, patient
autonomy and involvement in medical decision making
remains the key driver for full disclosure of health information. Autonomy trumps other ethical and social concerns.17
Research and practice have shown that most patients can
cope with grim information and, with guidance and support, come to a resolution of their emotional pain.18-20
Physicians vary in their level of comfort with such disclosure, and despite the increasing availability of communication skills training, practices are often informed by
instinct and shades of paternalism.12 The same is true of
patients and family members, who often keep important
information from each other. Negotiating these fundamentally private issues remains an important task for
clinicians.
At Harvard Medical School, instruction in cross-cultural
communication begins in the first week of the first year. In
small groups, students discuss their reactions to film clips
and texts chosen to highlight the plight of new immigrants
and the challenges of delivering care in a pluralistic society.
Using case-based learning and with expert facilitation, they
are guided to articulate and identify solutions for complex
communication problems. Courses are designed and delivered by multidisciplinary faculty to drive home the important message that the lessons learned are fundamental for
good practice and broadly applicable. Didactic sessions follow Kleinmans explanatory model of illness, which stimulates inquiry and patient-centeredness and avoids
stereotyping.21 Lessons in cross-cultural communication are
later inserted into other course materials to reinforce the
message that good clinical care is based on understanding
what is fundamentally important to each patient.
Author
Lidia Schapira*
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Farmer P, Frenk J, Knaul FM, et al. Expansion of cancer care and
control in low-income and middle-income countries: a call to action. Lancet.
2010;376:9747.
2. Smedley BD, Stith AY, Nelson AR. Unequal Treatment: Confronting
Ethnic and Racial Disparities in Health Care. Washington, DC: Institute of
Medicine; 2002.
3. Betancourt JR, Renfrew MR. Unequal treatment in the US: lessons and
recommendations for cancer care internationally. J Pediatr Hematol Oncol.
2011;33:S149-S153 (suppl 2).
4. Fallowfield L. Truth sometimes hurts but deceit hurts more. Ann N Y
Acad Sci. 1997;809:525-536.
5. Greene JD, Cushman FA, Stewart LE, et al. Pushing moral buttons: the
interaction between personal force and intention in moral judgment. Cognition. 2009;111:364-371.
6. Greene JD, Morelli SA, Lowenberg K, et al. Cognitive load selectively
interferes with utilitarian moral judgment. Cognition. 2008;107:1144-1154.
7. Greene J. From neural is to moral ought: what are the moral
implications of neuroscientific moral psychology? Nat Rev Neurosci. 2003;
4:846-849.
8. Greene J, Haidt J. How (and where) does moral judgment work?
Trends Cogn Sci. 2002;6:517-523.
9. Green AR, Carney DR, Pallin DJ, et al. Implicit bias among physicians and its prediction of thrombolysis decisions for black and white
patients. J Gen Intern Med. 2007;22:1231-1238.
e87
LIDIA SCHAPIRA
10. Kagawa-Singer M. Impact of culture on health outcomes. J Pediatr
Hematol Oncol. 2011;33:S90-S95.
11. Surbone A. Telling the truth to patients with cancer: what is the
truth? Lancet Oncol. 2006;7:944-950.
12. Cherny NI. Controversies in oncologist-patient communication: a
nuanced approach to autonomy, culture, and paternalism. Oncology. 2012;
26:37-46.
13. Collins F. What we do and dont know about race, ethnicity, genetics
and health at the dawn of the genome era. Nat Genet. 2004;36:S13-S15.
14. Kleinman A. What Really Matters. New York: Oxford University Press,
2006.
15. Corbie-Smith G, Thomas SB, St George DM. Distrust, race, and
research. Arch Intern Med. 2002;162:2458-2463.
16. Corbie-Smith G, Thomas SB, Williams MV, et al. Attitudes and beliefs
of African Americans toward participation in medical research. J Gen Intern
Med. 1999;14:537-546.
17. Schneider C. The Practice of Autonomy: Patients, Doctors, and Medical
Decisions. New York: Oxford University Press, 1998.
e88
18. Parkes CM. Psychological aspects. In Saunders CM (ed). The Management of Terminal Disease. London: Edward Arnold, 1978;44-64.
19. Smith TJ, Low LA, Virago E, et al. Giving honest information to
patients with advanced cancer maintains hope. Oncology (Williston Park).
2010;24:521-525.
20. Mack JW, Wolfe J, Cok ER, et al. Hope and prognostic disclosure. J Clin
Oncol. 2007;25:5636-5642.
21. Kleinman A, Eisenberg L, Good B. Culture, illness, and care: clinical
lessons from anthropologic and cross-cultural research. Ann Intern Med.
1978;88:251-258.
22. National Standards on Culturally and Linguistically Appropriate Standards. http://minorityhealth.hhs.gov/templates/browse.aspx?lvl2&lvl1D
15. Accessed March 15, 2012.
23. Butow PN, Goldstein D, Bell ML, et al. Interpretation in consultations
with immigrant patients with cancer: how accurate is it? J Clin Oncol.
2011;29:2801-2807.
24. Donelan K, Hobrecker K, Schapira L, et al. Medical interpreter knowledge of cancer and cancer clinical trials. Cancer. 2009;115:3283-3292.
THE ONCOLOGIST, THE PATIENT, AND THE MEDIA

CHAIR
Paul R. Helft, MD
Indianapolis, IN
SPEAKERS
Diane Blum, MSW
Lymphoma Research Foundation
New York, NY
Jeremy Manier
Chicago, IL
Patients with Cancer, Internet Information,

and the Clinical Encounter: A Taxonomy of
Patient Users
By Paul R. Helft, MD
Overview: The Internet has changed all of our lives forever

and has certainly changed the way in which patients acquire
information, share their stories, find others in similar circumstances, and analyze their medical situations. It is very clear
that patients have widely adopted the use of online resources
in the face of illness. Access to unfiltered information online
clearly has positive and negative potential effects, and the
introduction of Internet information into the physician-patient
T IS difficult to imagine life without the Internet. Not

only would the practice of medicine be drastically different from the way it is today, but the effect of the Internet on
all aspects of our lives has been so protean as to represent a
true revolution. From the early days of rapid uptake of the
Internet among U.S. adults in the 1990s, use of the Internet
to obtain health information is one of the most common
tasks for Internet users.
Early reports on the use of the Internet among both
patients of multiple specialtists as well as people with
cancer showed that patients from disadvantaged backgrounds used the Internet to obtain health information less
often than patients with higher literacy and socioeconomic
status. For example, in a study our group conducted in the
early 2000s among disadvantaged patients at an urban
public hospital oncology clinic, we found that only 10% of
patients with cancer reported looking up cancer information
on the Internet.1 More recent studies have suggested that
the proportion of patients with cancer who use the Internet
to obtain information about their diseases is closer to twothirds.2
National data from representative samples of U.S. adults
aimed at documenting trends in Internet use have been
continuously collected by the Pew Internet and American
Life Project for more than a decade. In 2011, the Pew survey
found that nearly 80% of all adults now use the Internet.
Among them, 83% reported using the Internet to look up
health information. Important trends documented by this
survey have included a shrinking divide between black,
white, and Hispanic patients in their rates of use; shrinking
gaps between richer and poorer individuals as well as those
who have more or less formal education; and smaller disparities in use between rural and urban/suburban users.3
Early conceptual papers about the effects that access to
Internet information would have on patients hypothesized
that the effects might be profound. It was hypothesized that
in many instances, patients might have as much access to
peer reviewed and other information about their illness as
their physicians would have. Many investigators evaluating
the effects of Internet information have identified the inconsistent quality of Internet information as the primary risk
posed by burgeoning access to information. Indeed, most
studies of the quality and accuracy of information both in
cancer and noncancer health conditions have found substantive issues with health information on the Internet.4,5 Several authors have speculated or attempted to measure the
positive and negative effects Internet information might
encounter may be managed in more or less productive ways.

The means of managing such introductions of information
should vary based on physicians analyses of patients information preferences and styles and their apparent reactions to
the information. Managed well, knowledgeable patients can
offer important opportunities of informed and shared decision
making.
have when introduced into the physician-patient relationship. In a telephone survey, Murray and colleagues found
that patients who had looked up information on the Internet
and discussed it with their physicians were primarily seeking their physicians opinion about the information rather
than requesting a specific intervention.6 Waid and colleagues reported in their literature review that Internet
information was often triangulated in the physicianpatient relationship and that Internet information might
help patients make more informed decisions and strengthen
shared decision making. The authors pointed out that the
dangers include the threat such access to information poses
to the traditional, authoritarian model of the physicianpatient relationship, however.7
Data from the Health Information National Trends Survey (HINTS), sponsored by the U.S. Department of Health
and Human Services, the National Cancer Institute, the
Health Communication and Informatics Research Branch,
and the Division of Cancer Control and Population Sciences,
have suggested that patients initially seek information online and talk with their physicians about the information to
seek approval or assessment of the information.8 Respondents to the HINTS survey between 2002 and 2008 appeared
to have gained greater trust in information from health care
professionals and to have lost trust in health information
from the Internet.9
A less-studied, but inevitably important, area of Internet
use among patients with cancer is the use of blogs and
social networking sites (e.g., www.caringbridge.com; www.
facebook.com). One study in which this issue was examined
showed that, in a small sample of patients with cancer and
their companions, four areas emerged as gratifications of
blog use: prevention and care, problem-solving, emotion
management, and information-sharing.10 This area is certainly important and growing and merits larger and more
systematic study using methods developed to preserve confidentiality of information among subjects.
When our group surveyed U.S. oncologists who were
ASCO members in the early 2000s, respondents median
From the Division of Hematology/Oncology, Indiana University School of Medicine,

Indianapolis, IN.
Address reprint requests to Paul R. Helft, MD, 1800 N. Capitol Ave, Noyes E-130,
Indianapolis, IN 46202; email: phelft@iupui.edu.
1092-9118/10/1-10
e89
PAUL R. HELFT
estimate of the proportion of their patients using the Internet to obtain cancer information was 30%. Oncologists
reported that, on average, 10 minutes were added to each
patient encounter in which Internet information was discussed, and that use of the Internet had the ability to
simultaneously make patients more hopeful, confused, anxious, and knowledgeable.11 One of the striking findings from
this study and others, however, was the suggestion that the
proportion of patients who use the Internet for cancer
information is higher than the proportion who actually
discuss the information they have read with their oncologists. However, there is little doubt that all practicing
oncologists routinely see patients who use the Internet to
obtain information and discuss that information in the
context of clinical encounters.
A Taxonomy of Patients Who Use the Internet
Although the literature on Internet use among patients

with cancer has begun to outline the effects that Internet
information may have on patients decision making, the
relationship with their oncologists, clinical trials enrollment, and on some emotional and psychological outcomes,
the state of the science is not such that the existing literature might be synthesized into an uncontroversial users
guide applicable to patient care. So instead of offering a
systematic analysis or synthesis of the existing literature, I
would instead offer suggestions regarding the management
of various situations in which patients introduce Internet
information into the clinical encounter. Clearly, patients
purposes for doing so, as well as their specific needs from a
respondent, are likely to be as different as each individual
patient. However, my own clinical experience suggests that
Internet information tends to arise in patterned ways. In
order to discuss each of these tendencies, I will offer a
stereotypical patient description, which is in no way meant
to suggest that patients or their information needs or styles
are reducible to simple formulas. Rather, my hope is that
KEY POINTS
e90
Internet information is pervasive and may have both

positive and negative effects on the clinical encounter
and the physician-patient relationship.
An individualized approach to information-sharing
and analysis should be based on an explicit assessment of patients information preferences.
When patients raise questions driven by Internet
information, recognition and validation that the issues raised are important can be an opportunity to
subtly reinforce the physicians competence.
Among patients with high health literacy who seek
Internet information, the main challenge is that they
may raise questions that even expert physicians are
not fully capable of responding to without some preparation.
Appearing to demonstrate pervasive misunderstanding or misinterpretation of Internet information or
clinging only to positive information may be signs of
substantial emotional distress that warrants attention.
these categories will permit patterns to emerge for the sake

of discussion and analysis. These four Internet informationusing patient archetypes are the Worrier, the Seeker, the
Knower, and the Misunderstander.
The Worrier
Mr. A is a 58-year-old man with newly diagnosed pancreatic cancer. When he was first seen, he had pain in the
abdomen and back and was jaundiced. A work-up revealed a
pancreatic adenocarcinoma with involvement of the surrounding vasculature and no evidence of distant spread.
Mr. A, as well as his wife and two children, actively sought
information about pancreatic cancer on the Internet. He
confesses that he stopped looking after the first 24 hours
because the information was too depressing.
Research on information acquisition and information preferences is far too vast to review here, but for the sake of
simplicity, there is a body of research indicating that,
although some patients try to avoid or minimize obtaining
stressful medical information, others seem to search for it
and are highly sensitive to it.12,13 Information avoidance
and information-seeking might thereby be considered two
different psychological strategies for managing anxiety: either by avoiding information or by seeking it out in order to
gain knowledge and control over it. The first mode has been
called monitoring (attention to, scanning for, and amplification of threatening cues), and denotes the extent to which
individuals are alert for and sensitized to the negative,
potentially painful, or dangerous aspects of information and
experience. The second mode, blunting (avoidance of threatening cues), is characterized by individuals distracting
themselves from such information.12,13
In this way, Mr. A might be understood simplistically to be
a blunter, a person whose natural tendency in the face of
threatening or anxiety-provoking information is to avoid
information. Instead of feeling the relief that many people
gain when coming to understand information about their
condition better, Mr. As attempt to seek information conflicted with his personal psychology, and he found it extremely anxiety-provoking. In my own experience, such
patients frequently go through the course of their illness and
ask few questions and engage in little discussion about
potentially worrisome or threatening information. Instead
they speak little or talk about other, less-threatening things.
In dealing with a patient such as Mr. A, I would offer three
suggestions based on the important piece of information he
provided to his oncologist about his experience with Internet
information. First, evidence suggests that such information
styles are deep-seated and pervasive. Being aware of a
patients tendency to avoid threatening information as a
coping mechanism should alert the physician that instances
in which informationparticularly negative informationis to be communicated and shared should be preceded
by frequent assessments of the patients state of mind and
preferences for such information at any given time; for
example, I have a few difficult things to talk with you about
today, Mr. A. How do you feel about tackling some of them
today? Second, blunters are often surrounded by monitors, family members or loved ones who cope with the stress
of the patients situation by seeking out as much information
as possible. Although it is clearly not the physicians role to
manage all related individuals information-seeking behaviors, knowledge that their needs may be different from the
THE INTERNET AND THE CLINICAL ENCOUNTER
patients allows the physician to seek clarification about who

needs or wants to know what when. Finally, because the
information-seeking preferences of blunters may cause
them to avoid information, a strategy for managing critical
information sharing needs to be negotiated early in the
course of the patients illness trajectory. For example, asking
a patient how he or she wants to discuss stressful information ahead of time may help the patient to feel more in
control of information delivery events.
The Seeker
Mr. B is a 63-year-old man with newly diagnosed intrahepatic cholangiocarcinoma. When he was first seen, he had a
palpable mass; he was subsequently found to have a large
adenocarcinoma replacing much of his atrophied left lobe of
the liver, with extension into the right lobe. As time passes
during his treatment, to which he has a major partial
response, he raises issues he has encountered on the Internet,
mostly after prompts from fellow patients he communicates
with on the Cholangiocarcinoma.org message boards.
One of the potential advantages of access to the Internet
in the context of illness is that patients with uncommon
cancers can form relationships online with other patients
despite geographical and other barriers. For many patients,
this ability to share experiences with other patients with the
same disease is valuable and can provide an important
source of emotional support. That said, I have found such
Internet sources frequently lead to questions from patients
that suggest they have difficulty sorting out what applies to
them and what does not, having heard suggestions from
other patients with the same diagnosis but whose circumstances may be radically different. Mr. B thus asks questions at each clinic encounter such as, One of the patients
from my message board had a liver transplant for his
cholangiocarcinoma. Can I have a liver transplant? At other
times, he raises questions about clinical trials, hepatic
intra-arterial infusion therapy, and surgical resection, all
stimulated by comments from other patients. Each time he
considers new information, such comments lead him to seek
further information from other Internet sources and he asks
clarification of his oncologist.
Suggestions for management of Mr. Bs continued questions include focusing on the treatment options he raises and
how they apply to patients in different circumstances, and
explaining how these options have already been considered
and why they do or do not specifically apply to him. The
important message from both of these communicative suggestions is that the physician recognizes and validates the
importance of the issues to him (and thus the discussion of
them validates the sense that the physician is willing to
explore other options with him). This discussion can be an
opportunity to subtly reinforce the physicians competence,
as such questions enable the physician to show that he or
she has already carefully considered the options raised.
Finally, recognizing the importance of such online experiences for the patient can represent a meaningful occasion for
validating the patients efforts to continue to investigate and
understand his disease and options fully.
The Knower
Ms. C is a 52-year-old former medical social worker who

was seen for vague abdominal symptoms. A computerized
tomography scan identifies three separate liver lesions in the

right lobe of the liver. A biopsy is consistent with a low-grade
neuroendocrine carcinoma. She undergoes a complete resection of disease, including an ileal primary tumor. During
multiple clinic encounters, she asks many informed questions
about tumor biology, clinical and serum biomarkers, genetic
evaluation, surveillance measures, and potential future therapies.
This patient represents the stereotypical highly-educated,
informed, and savvy patient who has benefitted greatly from
access to information. To pick up on an earlier theme, she
would likely be characterized as a monitor, in that part of
her coping strategy for dealing with her illness is to learn as
much as possible about her disease and its management.
This patients information preferences represent both a
challenge and an opportunity, however. The main challenge
is that such patients, who have both access to information
and the knowledge and health literacy to understand the
information at a deep level, may at times raise questions
that even expert physicians are not fully capable of responding to without some preparation. The important opportunity
that such patients represent is that they can become sophisticated and rewarding true partners for an interested physician because they are capable of engaging in truly
informed and shared decision making.
An important communicative principle to keep in mind
when interacting with such a patient is that the physician
must make clear that he or she is open to having the patient
engage in such information seeking and that her participation is welcome and encouraged. It is acceptable to ask the
patient explicitly what she hopes to get out of any discussion
that follows from her questions related to the information
she finds (e.g., clarification? perspective? personal application?). Seeking explicit understanding about the patients
goals and then trying to achieve them is a means of trust
building, as there is evidence that trust points are acquired through behaviors that demonstrate true interest in
and caring for the patient. It is also appropriate as a
time-management strategy to ask the patient to identify the
three or four most important questions for each visit, and to
even to send them ahead of time if possible if the physician
finds that answering a dozen such questions extends the
clinical encounter beyond the feasible limits.
The building of trust through openness to and respect for
a patients dedicated effort to understand his or her disease
as much as possible can be reinforced by comments that
recognize or praise the patients efforts. This trust can
become important for times when the patient may raise a
question for which the physician does not have an immediate answer or informed perspective. This is especially true
for the rare cases in which a patient introduces such questions as a kind of test of the physicians competence or
knowledge.
The Misunderstander
Mrs. D is a 72-year-old woman with newly diagnosed

metastatic pancreatic adenocarcinoma. She has had a difficult time accepting her diagnosis and its associated prognosis, as has her husband and four grown sons. She encounters
another patient on a message board who had exactly the
same thing as her but who had intensive chemotherapy and
underwent surgery to remove the primary tumor and liver
metastases. She becomes convinced that she needs to undergo
e91
PAUL R. HELFT
exactly the same treatment, even though her oncologist has

assessed her performance status and associated problems
and found her not to be a candidate for intensive therapy. At
each clinical encounter, she raises the same questions over
and over, driven by continued Internet searches and interactions with other patients online.
In many ways, the archetype that Mrs. D represents is the
most challenging of the four oversimplified patient types
presented. She is challenging because such patients appear
to place more credence in information they find online than
they do in their oncologist, even if they feel they have a good
relationship with him or her. This is plausibly driven by
psychological distress and the need to be reassured by
seeking out positive information preferentially in the face of
such a devastating diagnosis.
Several important issues should shape the approach to
such a patient. The first is to recognize that the primary
driver of this patients reaction to information (much of
which she may not fully understand or accept) is distress
about her illness. In this respect, it would be valuable to
pursue efforts at helping to alleviate her psychological
distress (e.g., counseling, cognitive behavior therapy, medication, supportive interventions). The second is to take into
account that such patients apparent lack of understanding
or acceptance of information from the physician that contradicts online information is most likely due to distress and
not to cognitive impairment, low health literacy, or low level
of education (although all of these issues could also have a
role). Thus, as a management strategy, continued explanations and attempts to correct misunderstandings are likely
not to be productive in the long term. Stated differently,
after an initial attempt to correct misinformation or misunderstandings, further long, cognitive explanations are not
likely to change the patients mind and, indeed, might even
appear adversarial. A different strategy I would recommend
is to implement a supportive listening approach, in which
the physician listens supportively and actively to the patients stories about information acquired on the Internet
and acknowledges the patients efforts but does not attempt
to correct the perceptions on a repeated basis, except when
specifically asked to render an opinion. This approach recognizes the cognitive block that may be affecting the
patients perceptions and the fact that further logic-driven
discussion is not likely to lead to greater understanding. At
the same time, this approach allows the physician to engage
in relationship-building behavior (active listening). This
issue is difficult, however, driven as we are to logic and
explanation as a means of effective communication and so
requires some letting go of those goals of communication.
In my own experience with patients who do not appear
emotionally prepared to have their misunderstandings and
misperceptions corrected. I have found that shifting the
focus of our interactions to supportive listening and abandoning attempts at contradictory argument is a reasonably
productive approach.
Conclusion
The Internet has changed all of our lives forever and has
certainly changed the way in which patients acquire information, share their stories, find others in similar circumstances, and analyze their medical situations. It is clear that
patients have widely adopted the use of online resources in
the face of illness. Access to unfiltered information online
clearly has positive and negative potential effects, and the
introduction of Internet information into the physicianpatient encounter may be managed in more or less productive ways. The means of managing such introductions of
information should vary based on the physicians analysis of
a patients information preferences and styles and his or her
apparent reactions to the information. Managed well,
knowledegable patients can offer important opportunities of
informed and shared decision making.
Author
Paul R. Helft
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Genentech; LillyAmylin (I);
Merck (I)
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Helft PR, Eckles RE, Johnson-Calley CS, Daugherty CK. Use of the
internet to obtain cancer information among cancer patients at an urban
county hospital. J Clin Oncol. 2005;23:4954-4962.
2. Castleton K, Fong T, Wang-Gillam A, et al. A survey of internet
utilization among patients with cancer. Support Care Center. 2011;19:11831190.
3. Pew Internet and American Life Project. Demographics of internet
users. http://pewinternet.org/Static-Pages/Trend-Data/Whos-Online.aspx. Accessed Feb 13, 2012.
4. Eysenbach G, Powell J, Kuss O, Sa ER. Empirical studies assessing the
quality of health information for consumers on the world wide web: a
systematic review. JAMA. 2002;287:2691-2700.
5. Lawrentschuk N, Sasges D, Tasevski R, et al. Oncology ealth information auality on the Internet: a multilingual evaluation. Ann Surg Oncol. Epub
2011 Dec 7.
6. Murray E, Lo B, Pollack L, et al. The impact of health information on the
internet on the physician-patient relationship: patient perceptions. Arch
Intern Med. 2003;163:1727-1734.
7. Wald HS, Dube CE, Anthony DC. Untangling the Webthe impact of
e92
Internet use on health care and the physician-patient relationship. Patient

Educ Couns. 2007;68:218-224.
8. Hesse BW, Nelson DE, Kreps GL, et al. Trust and sources of health
information: the impact of the Internet and its implications for health care
providers: findings from the first Health Information National Trends Survey.
Arch Intern Med. 2005;165:2618-2624.
9. Hesse BW, Moser RP, Rutten LJ. Surveys of physicians and electronic
health information. N Engl J Med. 2010;362:859-860.
10. Chung DS, Kim S. Blogging activity among cancer patients and their
companions: Uses, gratifications, and predictors of outcomes. J Am Soc
Information Sci Technol. 2008;59:297-306.
11. Helft PR, Hlubocky F, Daugherty CK. American oncologists views of
internet use by cancer patients: a mail survey of American Society of Clinical
Oncology members. J Clin Oncol. 2003;21:942-947.
12. Miller SM, Combs C, Stoddard E. Information coping and control in
patients undergoing surgery and stressful medical procedures. In: Steptoe A,
Appels A (Eds.) Stress, Personal Control, and Health. Chichester, England;
John Wiley & Sons, 1989;107-130.
13. Miller SM. Cognitive informational styles in the process of coping with
threat and frustration. Adv Behav Res Ther. 1989;11:223-234.
GLOBAL VARIATIONS IN ACCESS TO NEW

THERAPIES FOR SARCOMAS AND
GASTROINTESTINAL STROMAL TUMOR
CHAIR
Dana-Farber Cancer Institute/Harvard Medical School
Boston, MA
SPEAKERS
Stefan Sleijfer, MD, PhD
Erasmus University Medical Center, Daniel den Hoed Cancer Center
Rotterdam, Netherlands
Ian Judson, MD
Royal Marsden Hospital and NHS Foundation Trust
How to Decide Whether to Offer and Use

Nonstandard Therapies in Patients with
Advanced Sarcomas and Gastrointestinal
Stromal Tumors: Global Variations in Clinical
Practice, Assessment, and Access to
Therapies in Diseases with Limited
Incidence and Data
By Stefan Sleijfer, MD, PhD, Ian Judson, MD, and George D. Demetri, MD
Overview: As cancer is more generally recognized as a

collection of various rare diseases rather than a homogeneous
illness, sarcomas have become a model for the manner in
which data can and cannot be used to drive clinical decision
making. In this article, we explore the limitations of data
generated in rare diseases such as sarcomas to provide an
evidence base for clinical practice. How should patients be
treated if there is no standard that offers proof of clinical
benefit? By asking this question, we also raise the issue of
ANCER IS increasingly recognized as a complex group

of hundreds if not thousands of different and very
diverse diseases. Even the most common forms of cancer are
evolving into subsets of rare diseases. As a paradigmatic
example of this fragmentation of neoplastic diseases based
on improved clinicopathologic, molecular, and genomic subclassifications, the rare diagnostic group of soft-tissue sarcomas (STS) comprises tumors in which more than 50
different histologic subtypes can be distinguished. Despite
substantial differences between the diverse subtypes in
genetic aberrations, natural course, and sensitivity to systemic drugs, patients with advanced STS were until recently
treated similarly, with doxorubicin and ifosfamide being the
most commonly applied compounds.
This changed after the importance of subtyping in
STS was underlined by the finding that receptor tyrosine
kinase inhibitors, such as imatinib or sunitinib, dramatically improved the prognosis of patients with advanced
gastrointestinal stromal tumors (GISTs) and dermatofibrosarcoma protuberans based on a rational understanding
of the fundamental pathophysiologic abnormalities in
these diseases. Importantly, the level of antitumor activity
in other STS subtypes with imatinib was quite low, and
anecdotal evidence of activity with the more broadly
acting sunitinib has never been fully validated in large
studies. Certain studies have been performed in specific
STS subtypes (such as tenosynovial giant cell tumor or
alveolar soft part sarcoma), but the data have remained
quite limited.
Performing very selective, subtype-specific trials is the
most rigorous and logical way to proceed in STS clinical
research. However, an obvious consequence of doing studies
in very rare subgroups of a rare disease like STS is the
practical logistical and technical difficulty to perform clinical
trials that are adequately powered by conventional statistical standards. As a result, many recent publications describe relatively small, single-arm phase II studies in
specific tumor types from which it is difficult, if not impossible, to draw definitive conclusions to guide evidence-based
decision making for daily clinical practice.
what constitutes clinical benefitand how to measure that

for patients with sarcomas and other rare diseases. As physicians become more accountable for decisionsand yet are
always accountable to the patients and families who rely on
them to provide the best and most appropriate care oncologists must be cognizant of the limitations of data in rare
diseases and be ready to justify actions that are in the best
medical and social interests of patients.
Although this will perhaps become a problem for all forms

of cancer, as biologic understanding fragments even the
most common cancers into groupings of increasingly rare
subtypes (witness the 3% of nonsmall cell lung carcinoma
with ALK genomic aberrancies, or the 10% to 20% incidence
of HER2 overexpression in a subset of gastric carcinomas),
the international sarcoma community may be in the unenviable position to manage this challenge now. There is a
tension between wanting to provide each individual patient with the best possible and most appropriate care
options and the lack of proof on which to base such
judgments about what is truly best possible or even appropriate. One first important step is to agree internationally on parameters and thresholds for activity to define
clinical benefit and to implement these in clinical trials,
taking into account the rarity of STS subgroups. Another
important issue is how to treat patients for whom no
standard treatment is available. Are outcomes of singlearm studies with drugs enough to justify off-label use of
these drugs with all the potential financial, regulatory,
legal, and practical consequences? And what should physicians do with a drug that is considered to yield clinical
benefit for a particular patient population but which is not
considered cost-effective by administrative bodies (such as
governmental agencies that decide public policy or private
payers)?
This article addresses how physicians in different parts of
the world involved in the treatment of STS patients struggle
with, and manage, these issues.
From the Erasmus MC Daniel Den Hoed Cancer Center, Rotterdam, Netherlands; CR UK
Centre for Cancer Therapeutics, Sutton, United Kingdom; Ludwig Center and Sarcoma
Center, Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA.
Address reprint requests to Stefan Sleijfer, MD, PhD, Erasmus MC Daniel Den Hoed
Cancer Center, Medical Oncology, Groene Hilledijk 301, Rotterdam 3075 EA, Netherlands;
email: s.sleijfer@erasmusmc.nl.
1092-9118/10/1-10
645
SLEIJFER, JUDSON, AND DEMETRI

What Metrics Should Be Used to Define Clinical Benefit
in Sarcomas? Is There Consensus Internationally?
The concept of clinical benefit appears on the surfaceas intuitively clear and simple: that which helps a
patient feel, function, or survive better. The challenge is that
measuring anything other than survival is impossibly complex and very insensitive to variations that may be highly
relevant and desirable to patients. Even survival is a
complex and composite measure dependent on many variables; as such, survival does not directly measure a single
intervention such as an individual new drug or therapy. One
could keep a patient alive longer than the patient might
wish, for example, using heroic measures that do not provide
a desirable or reasonable quality of life; alternatively, some
patients who lose hope might withdraw and avoid potentially life-prolonging standard therapy options. The challenge for physicians caring for patients as well as for
physicians serving as regulators (or advisors to regulatory
bodies) is to focus on the needs of the patients in a given
clinical context.
This critically important element of clinical context is
often glossed over or ignored entirely in discussions of
clinical benefit. The fact that people with different types of
illnesses and with different risks of morbidity or mortality
may choose to make different decisions represents one clear
and compelling piece of evidence that clinical context truly
matters. The risk of life-altering symptoms, such as pain or
severe dyspnea from massive lesions pressing on nerves or
major bronchi, is a major concern to many patients with
advanced sarcomas, given the fact that sarcomas can occur
in young, otherwise healthy individuals and can grow to
KEY POINTS
646
It is impossible to generate statistical proof of beneficial activity in every rare disease for every possible
therapy.
Physicians must learn to read between the lines and
make clinical decisions on the basis of best available
evidence and incomplete information.
All drugs have side effects and risks, but professional
judgment can and must be used in discussions with
patients about management options once standard
proven therapy has failed.
Clinical context is very important for weighing risks
and benefits, and patient preferences must also be
integrated into such clinical context for decision making.
Best supportive care may sometimes be the most
appropriate option for patients with incurable diseases and with no available therapy of proven value,
but such decisions also should take into account other
lines of evidence as well as patient preferences.
Assessments of cost-effectiveness require clinical expert opinions, and clinical experts have an important
role in advocacy for patient preferences and patient
needs, as well as rigorously interpreting limited data
by conventional methods of statistical proof.
significant bulk before interfering with function in these

patients. However, once that tipping point has been reached,
the patients may have a limited ability to move from
adequate function to decreased function with symptoms
to severe dysfunctional status or even death. This contextual distance is highly relevant for patients with STS, who
risk living with the burden of very bulky disease, often far
more bulky than patients with other cancers.
This is directly applicable to a controversy that is perhaps
one of the most contentious in clinical oncology research
today: what is the value to patients of maintaining patients
in a progression-free state? This is particularly vexing in
rare-disease research, where it may be difficult or impossible to prove any effect on overall survival (OS). This is
exemplified by two large phase III clinical trials that enrolled patients with a variety of STS subtypes in different
clinical settings. PALETTE was a phase III international
trial in which a limited spectrum of patients with STS
were randomly selected to receive either placebo or a multikinase inhibitor, pazopanib, in the setting of metastatic
disease that was progressing despite prior standard chemotherapy.1 In contrast, SUCCEED was a phase III international trial in which patients with a broader array of
sarcomas were randomly selected to receive either placebo
or a mammalian target of rapamycin inhibitor, ridaforolimus, in the setting of metastatic disease that was stable or
in objective remission (partial or complete) following previous standard chemotherapy.2 These trials both met statistically predefined endpoints that demonstrated the beneficial
effects of the two drugs under study (pazopanib or ridaforolimus, respectively) compared with placebo. However,
neither study was able to prove a statistically significant
effect of either drug on OS. Nonetheless, these trials highlight the challenge: in patients with metastatic sarcomas,
what is the metric of clinical benefit? Is this metric constant
across clinical settings? In both studies, patients with metastatic, incurable sarcomas were enrolled and studied. However, in PALETTE, these patients had progressive disease
that had already progressed despite prior standard systemic
therapy; in SUCCEED, these patients entered the trial with
evidence of disease control from prior standard therapy.
These differences in clinical context may or may not be
important in weighing the clinical benefit of a new treatment, especially in the palliative setting such as metastatic
sarcomas.
The relative risks and benefits of any new treatment are
clearly important, but the relative risks of uncontrolled
progressive sarcoma are also important to keep in mind.
This is especially true in diseases such as sarcomas that
tend to strike a somewhat younger and overall healthier
population than many cancers. Finally, the lack of tools to
dissect out subtle variations and differences in patient
preferences as well as patient experiences with disease and
with treatments also complicates the problems of assessing
clinical benefit in this field.
The rarity of sarcomas only adds to the complexities of
interpreting such data and making clinical extrapolations to
interpret true clinical benefit for patients. Ultimately, clinical experience, expert judgment, and the totality of available evidence must be used to assess the clinical benefits to
patients with advanced sarcomas.
NONSTANDARD THERAPIES FOR ADVANCED SARCOMAS AND GISTS

Decision Making in Patients with Sarcomas Following
Failure of Conventional TherapiesOff-Label Drug
Use or Hospice?
At a certain point in time, virtually all patients with

advanced STS will eventually be confronted with the fact
that no standard treatment options with solid evidence of
clinical benefit are available. This is always a difficult
situation for both patients and physicians, with the most
important question being how to proceed. (Although we
focus on drug use here, we note that the same logic can also
apply to decisions regarding application of surgical or radiotherapeutic interventions, perhaps with even greater risk of
morbidity and financial implications than that of drug use.)
Many patients may be in a poor clinical condition after
their sarcomas fail to react to standard treatments. Importantly, the strongest prognostic factor for early mortality of
patients with advanced STS is performance status. A recent
analysis of the European Organisation for Research and
Treatment of Cancer Soft Tissue and Bone Sarcoma Group
(EORTC-STBSG) database of patients with advanced STS
treated with systemic treatments in the context of clinical
trials showed that the mortality rate within 90 days after
treatment start was 3.3%, 9.4%, and 25.5% for patients with
a World Health Organization (WHO) performance of 0, 1,
and, 2 or more, respectively.3 This study concerned only
patients treated in the first-line setting, and it is likely that
the outcomes would be worse for patients following failure of
prior systemic regimens. This observation, together with the
risk of toxicities accompanying systemic treatments, justify
best supportive care for patients with advanced STS following failure of prior therapies and who have a frail clinical
status (certainly, those with a WHO performance status of
3 or 4).
For pretreated patients who still have a good performance
status, have good organ functions, and who strongly wish to
be treated, the situation is more complex. Roughly, in the
absence of available studies, there are two options: best
supportive care or off-label use of drugs. Off-label use of
drugs refers to those situations in which drugs are used
outside their official labeling. This applies to the use of drugs
in diseases other than mentioned in the labeling of that
particular drug, but also to those situations when a drug is
given on a different schedule than mentioned in the officially
approved language of the labeling text. Examples of the
latter include other doses, administration as combination
therapy if the drug was only approved as a single agent, or
an alternative duration or route of administration.
There are relatively few academic studies on the frequency of off-label drug use in sarcoma care, or even in
oncology overall. Some studies estimate the frequency of
off-label drug use in oncology at approximately 30% to 50%
of the prescribed medications.4 Although there are no specific studies on the exact frequency of off-label use in
advanced STS, it is clear from our collective experience that
off-label use in this setting is exceedingly common. Only a
very limited number of drugs have been approved for advanced STS by regulatory agencies. For example, the drugs
that are approved by the U.S. Food and Drug Administration (FDA) comprise doxorubicin for adult and childhood
sarcomas, imatinib and sunitinib for GISTs, and vincristine
for rhabdomyosarcomas.5 That means that a drug such as
ifosfamide, which is commonly applied in advanced STS, is
technically used off-label when used in STS. It is important,

however, to recognize the totality of evidence used in making
therapeutic decisions, and that is why inclusion in compendia such as the National Comprehensive Cancer Network
practice guidelines for sarcomas is based on criteria other
than formal and full regulatory approval.
These are several important issues that influence off-label
use. These include access to off-label drugs, legal responsibilities of physicians and pharmacists, and costs. In some
countries, off-label use of drugs is even illegal.4 All these
aspects may create major difficulties for patients and physicians, which should be taken into account when considering prescribing a drug off-label. Since legislation and
regulations concerning off-label use vary largely by country,
these will not be discussed here in detail.
Another important issue is the evidence underlying offlabel use, which greatly differs per drug. Whereas large
randomized studies have established the clinical value of
drugs such as ifosfamide in adult STS6 and etoposide in
small blue round cell sarcomas,7 and therefore have been
incorporated in guidelines, for many other drugs used offlabel the evidence is far less robust. Given the lack of
publications on this topic specifically in advanced STS, it is
difficult to discern exactly which drugs are frequently used
off-label in advanced STS and for which strong evidence is
lacking. One example is sorafenib, a drug approved by
regulatory agencies for hepatocellular carcinoma and renal
cell carcinoma, but which has been used off-label in GISTs.
A recent study by Italiano and colleagues investigated the
patterns of care in patients with metastatic GISTs after
failure of the only two approved standard treatments imatinib and sunitinib.8 Out of the 223 patients with GISTs,
24.5% received sorafenib as third-line therapy. Other treatments administered to these 223 patients were (1) best
supportive care, (2) rechallenge with imatinib, (3) imatinib
combined with other drugs, (4) nilotinib, or (5) clinical trials
with investigational products. In the group treated with
sorafenib, 19% were reported to have exhibited a response;
median progression-free survival (PFS) and OS were 4.9 and
10.7 months, respectively.8 Apart from this study, at the
time of writing this article, there were no other full publications on the outcomes of patients with GISTs treated with
sorafenib in this setting, except for two very small uncontrolled series of 38 and 32 patients, respectively, both
roughly showing similar results as the study by Italiano and
colleagues but as yet only presented in abstract form.9,10
Are these results sufficient to justify off-label use of
sorafenib in GISTs? Single-arm studies or retrospective
series are highly prone to all kinds of bias, as a consequence
of which many treatments considered promising on the basis
of such studies failed to demonstrate their clinical value in
subsequent studies. Undoubtedly, sorafenib induces clinical
benefit in a proportion of patients. But as long as we cannot
identify the patients who are likely to benefit from sorafenib,
some might question whether the benefits of sorafenib in a
proportion of the patients outweigh the toxicities to which
all patients are exposed. Sorafenib is a drug with wellknown toxicity necessitating a dose modification (e.g., interruption or reduction) in the majority of patients.10 In patients
with renal cell carcinoma, termination of sorafenib treatment
because of toxicities or progressive disease substantially
improved quality-of-life scores,11 stressing the effect that
these toxicities have on patient-reported quality of life.
647
Altogether, there are large differences in the level of

evidence underlying the application of off-label use in patients with advanced STS. It should be this evidence that
determines whether a certain drug should be given off-label
to a particular patient category. Some might argue that
physicians should be very cautious not to spoil the precious
time of patients suffering from diseases with a short and
dismal prognosis by exposing them to drugs with uncertain
efficacy and certain toxicity. Others might note that for some
patients, being denied a chance to benefit from uncertain
drugs, even with the certainty of toxicity, could be unacceptably grim and could further affect negatively the patients
own perceived quality of life. There are many individualspecific variables in these equations, and the lack of data
makes evidence-based decision making challenging if not
impossible.
Apart from some rare exceptions such as imatinib as
first-line treatment in advanced GISTs yielding unprecedented outcomes, whether a new regimen outperforms best
supportive care in terms of OS and quality of life (the
ultimate goals we should strive for in our patients) can only
be determined in the context of adequately designed and
rigorously conducted randomized studies. Additionally, offlabel use of drugs can be accompanied with issues of legal
responsibilities and costs. If there is a reason to treat a
patient with an off-label drug and it is ensured that an
individual patient can get access to the drug, a patient
should be informed about these potential consequences. Also
in health care systems where costs of drugs are covered by
society and not by an individual patient, such as in the
Netherlands, costs of off-label drugs, in particular expensive
ones, can be a major issue. In such situations, prescribing
off-label medications will eventually be at the expense of the
whole system and subsequently will have an effect on the
health care provided to other patients. But again, legislation
on off-label use greatly differs between countries.
For patients for whom no standard therapies are available
and who are candidates for further treatments, participation
in a clinical trial rather than off-label use should be the next
step. Many physicians are convinced that sorafenib is potentially effective in patients with advanced GIST, yet no
randomized trials have been developed to test this agent,
which was already approved by the FDA in 2005. Clearly,
there are currently too many barriers to set up such studies.
At a global scale, patient advocacy groups, physicians, regulatory agencies, health insurance companies, and politicians should join forces to facilitate the efficient design,
development, and deployment of such clinical trials. In this
way, our field might determine more swiftly whether a
promising drug should be implemented as an accepted
standard in daily clinical practice. And if so, labeling of such
a drug should be adjusted, thereby taking away all potential
financial and legal burdens associated with the use of an
off-label drug.
The Role of Physicians in Advocating for Patients
When Cost-effectiveness Is Questioned by
Administrative Bodies Such as NICE
The Physicians Duty as a Clinical Expert
Physicians in certain countries may be called on to participate in the regulatory process by serving as clinical experts.
The role of such clinical experts varies depending on the
648
regulatory setting: for example, the role of an expert physician on the Oncology Drugs Advisory Committee (ODAC) in
the U.S. FDA process is significantly different from the role
of an expert physician advising a governmental regulatory
body such as the National Institute for Health and Clinical
Excellence (NICE) in the United Kingdom. With NICE, the
focus is often on evaluating the cost-effectiveness of a new
treatment, rather than the risk-benefit, which is often the
subject of other regulatory bodies such as the FDA or the
European Medicines Agency. The first thing to recognize
when accepting the task of clinical expert in relation to
determining the cost-effectiveness of a new treatment is that
such physicians have a similar duty to that of an expert
witness in court. In order to be credible, one needs to be
(1) well-briefed, (2) prepared for robust cross-examination,
(3) demonstrably independent of the pharmaceutical company whose product is the subject of discussion, and (4) able
to argue effectively on behalf of the group of patients to
whom the treatment relates.
The United Kingdoms NICE as an Example of a Societal Process
in Drug
NICE was set up to oversee the evaluation of new treatments. Its deliberations and standard operating procedures
are being followed by other countries, which is why we chose
to scrutinize the way that it operates. One of the stated aims
of NICE was to reduce the inequalities of access to novel
therapies according to the geographical location of the patient and the local financial arrangements. This was to be
done by providing a robust method of determining costeffectiveness, and if a drug was deemed to be cost-effective,
it would be made available everywhere in the country. In
fact, within the United Kingdom this only applies to England and Wales; Scotland has a different system called the
Scottish Medicines Consortium, which appears to have a
more streamlined approach and usually produces a decision
soon after licensing. Unfortunately, NICE has not succeeded
in abolishing the postcode lottery for a variety of reasons,
including not yet carrying out appraisals on certain drugs
for rare cancers, and because it has always been possible to
appeal to local funding bodies and more recently a regional
Cancer Drugs Fund, in order to try and circumvent a
negative NICE appraisal decision. Unfortunately, the application of these mechanisms is inconsistent across the
country, hence the persistence of regional variations. Not
surprisingly, this gives rise to considerable distress among
patients. This pattern would most likely be applicable to
other countries trying to introduce a similar mechanism of
health care rationing.
The biggest source of controversy concerns the methods
used to generate data on cost-effectiveness. The process of
model generation is open to tender and groups of health
economists, generally based in academic institutions, bid for
the job of evaluating the new agent. The models used are not
placed in the public domain and hence cannot be directly
challenged, and it is openly acknowledged in certain technology appraisal documents that many assumptions have
been made owing to the lack of data on quality of life,
survival of patients in disease subsets, etc. In some cases the
results produced by different models for the same drug in
different appraisals vary significantly.
NONSTANDARD THERAPIES FOR ADVANCED SARCOMAS AND GISTS

Preparing a Professional Organization Statement
First, it is valuable to work closely with patient advocacy

groups in order to understand the needs and concerns of the
patients themselves. A carefully argued case in favor of a
new treatment will be enhanced by the input of patients in
the process of preparing the scientific and clinical case. It
may even be appropriate on occasion to submit a joint report.
You need to be aware of who you represent as a clinical
expert and ensure that you have consulted widely before
submission. This also lends the submission more weight.
You are required to state that you are an expert in the
treatment of patients who have the condition for which
NICE is considering the new treatment, that you understand the scientific and clinical basis for the treatment, and
that you are an employee of an organization representing
clinicians who treat the condition in question. The framework for the submission takes into account a number of
specific issues that include the current therapy for the
disease; any important variations in treatment, including
the merits of alternative treatment, if any; whether there
are any patient subgroups that might benefit particularly;
how the treatment would be used; and the existence of any
guidelines for the disease in question.
You will need to explain the scientific basis for the treatment, especially the number and quality of clinical trials
and how they might relate to everyday practice. Particular
emphasis might be placed on side effects and toxicities, that
is, the balance between benefit in terms of disease control
and the potentially adverse effects of treatment on quality of
life.
In practice this means having an in-depth knowledge of
the disease and its treatment, being familiar with the
clinical trial evidence for the new treatment, and in particular being able to argue effectively that the clinical benefit
outweighs the common adverse side effects.
What Are the Key Issues?
Cost-effectiveness in cancer treatment is determined by

the cost per quality adjusted life year (QALY) gained. The
calculations take into account the total number of patients
treated relative to the number who might live longer. Hence
if the percentage of patients who benefit is small and the cost
is high, a drug could be deemed not to be cost-effective even
if the benefit is prolonged for the few for whom it works. For
example, in Technology Assessment T209 on high-dose imatinib, it is stated that the estimated median increase in life
expectancy for imatinib 800 mg compared with best supportive care (which included imatinib 400 mg) was 4.2 months,
but cost-effectiveness for high-dose imatinib was not demonstrated compared with sunitinib.12 This was despite the
fact that no comparison of sunitinib with imatinib 800 mg
has been performed. In addition, in this particular assessment, no data on the benefit of imatinib 800 mg in patients
with KIT exon 9 mutant GISTs were regarded as permissible for evaluation because the assessment focused exclusively on patients whose cancers had progressed on imatinib
400 mg. The data on the PFS for patients with KIT exon 9
came from randomized studies (62005 and S0033), in which
patients received either 400 mg or 800 mg imatinib from day
1. These data were of course evaluated in a meta-analysis
demonstrating a significant benefit in terms of PFS for 800
mg for the exon 9 mutant group (p 0.012),13 but because
they did not address the limited scope of this appraisal they
were not considered.
Quality adjustment is often contentious, largely because
there are few data available and the tools to assess it are
somewhat crude. Essentially, the assumption is that patients with advanced cancer already have an impaired
quality of life and hence any prolongation of life is weighted
according to the disease itself and its likely effect on quality
of life and further weighted by the side effects of treatment.
In Technology Assessment 209, the utility of patients with
GISTs on best supportive care is given as 0.577, based on a
very small study.12 In the technology assessment for trabectedin in the treatment of advanced STS, the utility
values for progression-free and progressive disease health
states are stated as 0.653 and 0.473, respectively.14
It must be said at this stage that utility in this context
means the mean quality of life of patients relative to 1.00
being perfect, at this stage of life, and hence a presumed
reduction in the perceived value of any extension of life. This
is an assumption that patients tend not to accept. However,
it must be acknowledged that we do need better data on the
effect of cancer on quality of life at different stages of disease
in order to help make such assessments more realistic and
meaningful.
Incremental Improvements and the Comparator Approach
In determining the value of a new therapy, the standard

approach is to compare the new treatment against the
current best available therapy and if there is no significant
improvement in PFS or OS, then it is deemed ineffective, or
if the improvement is small and it is expensive, then it is
deemed cost-ineffective. However, no system has yet been
devised to evaluate the cost-effectiveness of sequential effective therapies, each with an individual small incremental
benefit but a significant combined effect. It is apparent that
survival of patients with many cancer types is increasing
because of earlier diagnosis, more effective primary treatment, and more effective therapy for advanced disease.
What is less easy to quantify is the effects of new targeted
therapies on survival and whether sequential therapy with
effective agents, each with a limited duration of benefit, is
truly beneficial or not. The current methods used by regulators and health care utility bodies such as NICE do not
address this issue.
The Question of Balance
There is no publicly funded health care system in the

world with unlimited resources, and methods need to be
developed that are transparent and respected to restrict
health care costs while ensuring value for money. No one
questions the value of imatinib for the treatment of GISTs or
chronic myeloid leukemia. However, a number of high-cost
new drugs have been licensed for the treatment of common
cancers, which have a statistically significant, but nevertheless limited effect on PFS or OS, or indeed have yet to
demonstrate any effect on survival. Notwithstanding the
issue of cumulative incremental benefit, this needs to be
kept in mind when preparing evidence to funding bodies or
health care evaluation organizations such as NICE. In
addition, it is clearly easier to demonstrate statistically
significant differences in PFS or OS in common cancers,
where large trials can be performed, compared with rare
cancers, even if the benefits are greater.
649
Author
Stefan Sleijfer
Ian Judson
George D. Demetri
Employment or
Leadership
Positions
Consultant or
Advisory Role
GlaxoSmithKline
(U); Novartis
GlaxoSmithKline;
Morphotek;
Novartis; Pfizer;
PharmaMar;
Threshold
Amgen; ARIAD;
Daiichi Sankyo;
Genentech;
GlaxoSmithKline;
Idera
Pharmaceuticals;
Infinity; Johnson
& Johnson;
Kolltan
Pharmaceuticals;
Merck Serono;
Momenta
Pharmaceuticals;
Novartis; Pfizer;
Plexxikon;
Ziopharm
Oncology
Stock
Ownership
Honoraria
Novartis; Pfizer
Champions
Biotechnology;
EmergingMed;
Kolltan
Pharmaceuticals;
Plexxikon
Novartis; Pfizer
Research
Funding
GlaxoSmithKline;
Johnson &
Johnson;
Novartis; Pfizer
GlaxoSmithKline;
Novartis; Pfizer
Expert
Testimony
Amgen; ARIAD;
Bristol-Myers
Squibb; Daiichi
Sankyo;
Genentech;
Infinity; Johnson
& Johnson;
Novartis; Pfizer;
PharmaMar
ARIAD (U);
Infinity (U);
Johnson &
Johnson (U);
Novartis (U);
Pfizer (U);
PharmaMar (U)
Other
Remuneration
REFERENCES
1. Van Der Graaf WT, Blay J, Chawla SP, et al. PALETTE: A randomized,
double-blind, phase III trial of pazopanib versus placebo in patients (pts) with
soft-tissue sarcoma (STS) whose disease has progressed during or following
prior chemotherapyan EORTC STBSG Global Network Study (EORTC
62072). J Clin Oncol. 2011;29:605s (suppl; abstr LBA10002).
2. Chawla SP, Blay J, Ray-Coquard IL, et al. Results of the phase III,
placebo-controlled trial (SUCCEED) evaluating the mTOR inhibitor ridaforolimus (R) as maintenance therapy in advanced sarcoma patients (pts)
following clinical benefit from prior standard cytotoxic chemotherapy (CT).
3. Penel N, Glabbeke MV, Mathoulin-Pelissier S, et al. Performance status
is the most powerful risk factor for early death among patients with advanced
soft tissue sarcoma: the European Organisation for Research and Treatment
of Cancer-Soft Tissue and Bone Sarcoma Group (STBSG) and French Sarcoma Group (FSG) study. Br J Cancer. 2011;104:1544-1550.
4. Leveque D. Off-label use of anticancer drugs. Lancet Oncol. 2008;9:11021107.
5. National Cancer Institute. Drugs Approved for Soft Tissue Sarcoma.
http://cancer.gov/cancertopics/druginfo/soft-tissue-sarcoma. Accessed February 6, 2012.
6. Lorigan P, Verweij J, Papai Z, et al. Phase III trial of two investigational
schedules of ifosfamide compared with standard-dose doxorubicin in advanced or metastatic soft tissue sarcoma: a European Organisation for
Research and Treatment of Cancer Soft Tissue and Bone Sarcoma Group
Study. J Clin Oncol. 2007;25:3144-3150.
7. Grier HE, Krailo MD, Tarbell NJ, et al. Addition of ifosfamide and
etoposide to standard chemotherapy for Ewings sarcoma and primitive
neuroectodermal tumor of the bone. N Engl J Med. 2003;348:694-701.
650
8. Italiano A, Cioffi A, Maki RG, et al. Patterns of care, prognosis, and

survival in patients with metastatic gastrointestinal stromal tumors (GIST)
refractory to first-line imatinib and second-line sunitinib. J Clin Oncol.
2011;19:615s (suppl; abstr 10044).
9. Reichardt P, Montemurro H, Gelderblom H, et al. Sorafenib fourth-line
treatment in imatinib-, sunitinib-, and nilotinib-resistant metastatic GIST: a
retrospective analysis. J Clin Oncol. 2009;27:15s (suppl; abstr 10564).
10. Kindler HL, Campbell K, Wroblewski R, et al. Sorafenib (SOR) in
patients (pts) with imatinib (IM) and sunitinib (SU)-resistant (RES) gastrointestinal stromal tumors (GIST): final results of a University of Chicago
phase II consortium trial. J Clin Oncol. 2011;29:607s (suppl; abstr 10009).
11. Cella D, Escudier B, Rini BI, et al. Patient-reported outcomes (PROs) in
a phase III AXIS trial of axitinib versus sorafenib as second-line therapy for
metastatic renal cell carcinoma (mRCC). J Clin Oncol. 2011;29:290s (suppl;
abstr 4504).
12. National Institute for Health and Clinical Excellence. Imatinib for the
Treatment of Unresectable and/or Metastatic Gastrointestinal Stromal Tumours. NICE Technology Appraisal 209. http://nice.org.uk/guidance/TA209.
Accessed February 5, 2012.
13. Gastrointestinal Stromal Tumor Meta-Analysis Group (MetaGIST).
Comparison of two doses of imatinib for the treatment of unresectable or
metastatic gastrointestinal stromal tumors: a meta-analysis of 1,640 patients.
J Clin Oncol. 2010;28:1247-1253.
14. National Institute for Health and Clinical Excellence. Trabectedin for
the Treatment of Advanced Soft Tissue Sarcoma. NICE Technology Appraisal
185. http://nice.org.uk/guidance/TA185. Accessed February 5, 2012.
TARGETED THERAPIES IN TARGETED OR

UNTARGETED SARCOMAS
CHAIR
Duke Cancer Institute, Duke University Medical Center
Durham, NC
SPEAKERS
Andrew J. Wagner, MD, PhD
Boston, MA
Robert G. Maki, MD, PhD
New York, NY
Targeted Therapy in Sarcoma: Should We Be

Lumpers or Splitters?
By Richard F. Riedel, MD, Robert G. Maki, MD, PhD, and Andrew J. Wagner, MD, PhD
Overview: The identification of KIT as a critical driver in the

pathogenesis of GI stromal tumor (GIST), and its subsequent
inhibition with imatinib, have resulted in tremendous efforts to
identify other potential therapeutic targets for the heterogeneous group of diseases known as sarcomas. Because of the
rarity of sarcoma and the often limited number of patients per
individual sarcoma subtype, clinical trials to date have often
utilized unselected patient populations including multiple subtypes. Although this strategy increases the ease with which a
particular trial may accrue patients, statistically significant
therapeutic responses across an unselected patient popula-
ARCOMAS REPRESENT a diverse group of mesenchymal neoplasm affecting approximately 13,000 individuals each year.1 The treatment of advanced disease is
particularly challenging because of the limited number of
effective systemic therapies. As a result of the success of
tyrosine kinase inhibitors (TKIs), such as imatinib in the
management of GIST, investigation of other targeted therapies for this rare group of diseases has become an active
and ongoing area of research. Some of the most active
investigations have centered on inhibiting angiogenesis as
well as the phosphoinositide 3-kinase (PI3K), Akt/mammalian target of rapamycin (mTOR), and insulin-like growth
factor 1 receptor (IGF1R) oncogenic pathways. Although a
number of multitargeted TKIs (MTKIs) and pathwayspecific therapies have been explored, response rates have
generally been disappointing and few have been investigated in the phase III setting.
Whereas a mesenchymal origin is shared between histologic subtypes of sarcoma, the molecular drivers of tumor
initiation, growth, and maintenance are more varied and
complex. Critical to our success in further understanding
the key signals and pathways that drive pathogenesis is
active communication between basic scientists and clinical
investigators. Although the traditional knowledge flow has
been from the laboratory to the clinic, observations noted
from clinical trials have, in fact, created fertile ground for
scientific exploration and discovery in the laboratory setting.
This review will highlight recent data from both the
laboratory and clinical settings, including recently reported
phase III studies, to support the ongoing investigation of
targeted therapies in the of patients with sarcoma. In
addition, the rationale behind the use of specific therapies in
more targeted populations of patients, with specific attention to unique sarcoma subtypes, will be provided.
Targeted Therapy in Unselected Subtypes
Because of the rarity of sarcoma and the desire to accrue

patients in a timely fashion, prospective clinical trials have
traditionally been more encompassing rather than more
restrictive with respect to the inclusion of histologic subtypes. The result is often a mixture of histologies i n different proportions from one study to the next, including
translocation-associated (e.g., synovial sarcoma) and aneuploid subtypes (e.g., leiomyosarcoma and undifferentiated
pleomorphic sarcoma), with heterogeneous biologic mecha-
652
tion are often limited. Furthermore, in the absence of biologic

correlatives, the identification of significant activity and subsequent interpretation of clinical trial results utilizing targeted
therapies for this patient population have been challenging.
However, hints have emerged, on the basis of preclinical and
clinical observations, to suggest that certain targeted therapeutic approaches are appropriate in select histologic subtypes. This brief review will highlight data supporting the use
of targeted therapy in both unselected and selected sarcoma
patient populations.
nisms driving their formation and growth. To date, the most

robust areas of investigation have included therapies targeting mTOR, angiogenesis, and IGF1R.
mTOR Inhibition and Sarcomas
The mTOR pathway plays an important role in cell growth

and proliferation and is an active target for developing
cancer therapeutics. Activation of the mTOR pathway has
been identified in a number of human cancers, including
sarcoma.2 A variety of mTOR inhibitors have been developed
and are in active investigation.
A phase II study of ridaforolimus, a nonprodrug rapamycin analog, in 212 heavily pretreated patients with advanced
soft tissue and bone sarcomas was reported.3 Patients received ridaforolimus 12.5 mg intravenously once daily for 5
days every 2 weeks. A benefit rate of 29% (complete
response [CR], partial response [PR], or stable disease [SD]
lasting 16 weeks or longer) with five PRs was observed. The
median progression-free survival (PFS) and overall survival
(OS) were 15.3 and 40 weeks, respectively.
An oral formulation of ridaforolimus was also developed
and subsequently moved into the clinical trial setting as part
a maintenance strategy in sarcoma. Initial results from the
randomized, multicenter, phase III SUCCEED (Sarcoma
Multi-Center Clinical Evaluation of the Efficacy of Ridaforolimus) trial were reported at ASCOs 47th Annual Meeting
(June 4 8, 2011, Chicago, IL).4 Patients with advanced soft
tissue or bone sarcomas were randomly assigned 1:1 to
receive oral ridaforolimus (40 mg daily for 5 days each week)
or placebo as a maintenance approach in the setting of SD or
responsive disease after at least four cycles but no more than
12 months of chemotherapy. This was one of the largest
randomized controlled trials performed in sarcoma to date
and included over 700 patients. Median PFS, the primary
outcome of interest, was 17.7 weeks with ridaforolimus
compared with 14.6 weeks with placebo (hazard ratio [HR]
0.72, p 0.0001), as assessed by independent review. PFS
From the Duke Cancer Institute, Duke University Medical Center, Durham, NC; Mount
Sinai School of Medicine, New York, NY; Dana Farber Cancer Institute, Boston, MA.
Address reprint requests to Richard F. Riedel MD, Box 3198 DUMC, Durham, NC 27710;
e-mail: richard.riedel@duke.edu.
1092-9118/10/1-10
TARGETED THERAPY IN SARCOMA
rates at 3- and 6-months were 70% and 34%, respectively,

with ridaforolimus compared with 54% and 23%, respectively, with placebo. The benefit rate (CR, PR, and SD
lasting 4 months or longer) was 41% with ridaforolimus
versus 29% with placebo (p 0.0009), and best target
response was a mean decrease of 1.3% with ridaforolimus
compared with a mean increase of 10.3% with placebo (p
0.0001). A nonsignificant trend in OS was seen favoring
ridaforolimus (21.4 vs. 19.2 months, HR 0.88, p 0.23).
Although the results from this study suggest that mTOR
maintenance therapy may be considered an alternative to
the more traditional watch and wait approach, it is important to consider associated adverse events (AEs) with this
medication. In the data presented, nearly two-thirds of
patients experienced at least 1 grade 3 or worse AE, including stomatitis (9%), infection (6%), thrombocytopenia (10%),
anemia (7%), and hyperglycemia (7%). When considering all
grades of AEs in enrolled patients, 61% experienced stomatitis; 52% infections; approximately one-third of patients
experienced fatigue, thrombocytopenia, diarrhea, cough,
and rash; and approximately one-quarter of patients experienced anemia, hypertriglyceridemia, and hypercholesterolemia. Even though these numbers represent events
occurring in a minority of patients, the potential impact of
lesser grade 1 to 2 events occurring in the context of the
maintenance setting should not be overlooked and results
from quality-of-life questionnaires obtained as part of exploratory analyses are anticipated.
In an effort to identify predictors of response to ridaforolimus, additional exploratory analyses were performed and
recently reported.5 Post hoc analyses revealed greater median PFS in those patients who experienced grade 2 or worse
stomatitis compared with those who did not, or those who
received placebo. In addition, an analysis of the cohort of
patients with SD at study entry suggested an increased
benefit in those patients with truly stable or early growth
compared with those with minor responses.
The role of mTOR inhibition in the treatment of metastatic sarcoma continues to be developed. Combination therapy with conventional cytotoxics, as well as combination
with other targeted therapies, is being considered. Assessing
alternative pathway activation in mTOR inhibitorresistant
disease will be important, and will likely guide future trials.
KEY POINTS
Sarcomas are a rare, heterogeneous cancer in need of

new therapies.
Clinical trials to date have focused largely on unselected patient populations including multiple subtypes.
A better understanding of the molecular mechanisms
driving the pathogenesis of sarcoma is needed.
Clinical trials with integrated biologic correlatives
are more likely to identify subtypes likely to benefit
from targeted therapy.
Collaboration between basic scientists and clinical
investigators is critical.
Angiogenesis Inhibition and Sarcomas
The use of antiangiogenic agents in sarcoma is not a new

concept. The mixed clinical results in cancer studies suggest
that the road to understanding the key elements of angiogenesis and where one may safely intervene in the patient
with cancer is far from straightforward.
Recent focus has been on therapies directed against vascular endothelial growth factor (VEGF) and its receptors
(VEGFRs). The combination of anti-VEGF monoclonal antibody bevacizumab with doxorubicin was explored in a small
phase II study.6 The combination was not obviously more
active than doxorubicin itself, and was associated with
toxicity. A more recent phase I study of gemcitabine and
docetaxel with bevacizumab showed a 31% RECIST (Response Evaluation Criteria in Solid Tumors) response rate.7
It is not clear whether this response rate reflects the activity
of the gemcitabine-docetaxel combination alone, or synergy
with bevacizumab. We hope this question will be answered
for leiomyosarcoma in a randomized phase III clinical trial
from the Gynecologic Oncology Group (clinicaltrials.gov
identifier NCT01012297) comparing gemcitabine and docetaxel with or without bevacizumab. For angiosarcoma, the
situation may be slightly different, with greater RECIST
response rates (12%) observed for bevacizumab.8
MTKIs have shown modest benefit in specific non-GIST
sarcomas as single agents, but their multitargeted nature
makes ascribing the activity of one kinase or another to the
growth of a specific sarcoma difficult. There are few data to
suggest that any MTKI has substantial activity in sarcomas
such as leiomyosarcoma or unclassified pleomorphic sarcoma/malignant fibrous histiocytoma (UPS/MFH). The phase
II studies of sunitinib, sorafenib, and pazopanib yielded very
low RECIST response rates across an unselected patient
population.9-12 Similar to bevacizumab, a higher response
rate (14%) for sorafenib was observed in angiosarcomas.
Nonetheless, MTKIs may have measurable activity, as
demonstrated in a randomized clinical trial of pazopanib
compared with placebo in advanced soft tissue sarcoma
subtypes.13 The PALETTE (Pazopanib Explored in SoftTissue Sarcoma) study explored 369 patients randomly
assigned to receive pazopanib 800 mg daily or placebo until
disease progression. Median PFS was 20 weeks for pazopanib
compared with 7 weeks for placebo (HR 0.31, p 0.0001). An
interim analysis showed a nonsignificant improvement in
median OS for pazopanib (11.9 vs. 10.4 months, HR 0.83,
p 0.18), in this study with a cross-over design.
Similarly, specific sarcoma subtypes may demonstrate
sensitivity to antiangiogenic agents. In addition to angiosarcoma and VEGF inhibitors, alveolar soft-part sarcoma can
respond to VEGFR inhibitors cediranib14 or sunitinib,15 and
solitary fibrous tumors demonstrate at least hints of activity
to sunitinib16 or the combination of bevacizumab and temozolomide.17 The precise mechanism of antiangiogenic agents
in these sarcoma subtypes remains unknown.
If mTOR can be construed to be a proangiogenic kinase,18
then the results of phase II and phase III studies are
less-than-convincing evidence for the use of first-generation
mTOR inhibitors as antiangiogenic agents.19 These agents
do not inhibit the reflex activation of Akt by mTOR complex
TORC2 that is common to all presently available allosteric
TORC1 complex inhibitors. As a result, newer direct TORC1
653
RIEDEL, MAKI, AND WAGNER
and TORC2 inhibitors could potentially demonstrate greater

activity as antiangiogenic compounds.
Other antiangiogenic agents have shown hints of activity
in sarcomas and other cancers, without convincing survival
improvement. The thrombospondin mimetic ABT-510 demonstrated only one RECIST PR in 82 patients receiving
treatment, although time to progression suggested some
activity of the compound.20 Vascular-disrupting agent combretastatin A4, a tubulin antagonist and antiangiogenic
agent, showed only modest clinical activity as a single agent
and in combination with cytotoxic chemotherapy agents in
phase I studies, although it appeared to have biologic effects
by dynamic contrast-enhanced magnetic resonance imaging
and other correlative studies.21,22 New derivatives of combretastatin are being examined preclinically as well as in
phase I studies. A newer approach involves targeting of
other kinases involved in angiogenesis, including TIE1 and
TIE2, with monoclonal antibodies, peptide-antibody conjugates, or small molecules, some of which have reached phase
I clinical trials.
Targeted Therapy in Selected Subtypes
Although the clinical classifications of sarcomas can provide prognostic and predictive information, they rarely include the emerging molecular biologic data that may lead to
advances in tumor-specific treatments. Several findings
from laboratory-based studies have recently been brought to
clinical investigation in specific sarcoma subtypes for which
there are few known effective therapies. Although the true
efficacy of these novel approaches is still unknown, the
pairing of genetic and biochemical alterations with targeted
therapies has generated proof-of-concept evidence of activity
in laboratory and early clinical settings.
IGF1R Inhibition in Ewing Sarcoma
IGF1R has been recognized as a viable target for treatment of pediatric sarcomas for 20 years.23 For example, the
role of IGF1R signaling to maintain activity of the Ewing
sarcoma EWSR1-FLI1 translocation product indicated that
blockade of this pathway would consistently block Ewing
sarcoma growth. It was not until there was availability of
agents more potent than somatostatin analogs at blocking
the IGF1R signaling axis that there has been an opportunity
to examine this hypothesis. In what is a common story
regarding the sophistication of cancer, IGF1R blockade,
which appeared to be nearly a perfect tumor-specific target
in Ewing sarcoma, has become a more complicated story.
The avalanche of data regarding Ewing sarcoma and
IGF1R inhibitors has come from the observation of responses from limited number of patients in the phase I
development of the monoclonal antibodies targeting the
receptor. Despite differences in specific antibody characteristics, a consistent RECIST response rate has been observed.
For example, AMG479 IV every 2 weeks yielded a long-term
RECIST CR and one PR in Ewing sarcoma and a PR in a
patient with a neuroendocrine tumor, although 10 other
Ewing sarcoma patients did not experience response.24 Two
of 13 assessable patients with Ewing sarcoma experienced
RECIST PR while receiving intravenous figitumumab in a
phase I expanded cohort study; others showed disease
shrinking but did not experience PR.25 A phase I/II study of
cixutumumab in a pediatric population showed two (10%) of
654
20 patients with a RECIST PR at the highest dose studied (9

mg/kg/week), but the median PFS for this cohort was
slightly longer than 6 weeks.26
Further study of figitumumab in a pediatric Ewing sarcoma population (10 to 18 years) in the phase II portion of a
phase I/II study demonstrated a 14% RECIST response rate
(15 PR in 106 patients).27 In this study, a higher baseline
total or free IGF1 level was associated with longer survival
than was low baseline free or total IGF1. A population of
patients with very high baseline IGF1 fared as poorly as
those with the lowest levels of IGF1, however. A similar
phase II study performed with IGF1R monoclonal antibody
yielded a RECIST response rate of 10% (11 of 115 patients,
10 of whom had had a bony primary site).28 In addition,
among patients with IGF1 samples available at week 6 of
the study, total serum IGF1 greater than 110 ng/mL at 6
weeks of therapy and a greater proportional increase in total
IGF1 from baseline to week 6 both predicted for improved
survival.
The rapid development of progressive disease observed in
some patients, after evidence of early treatment benefit,29,30
speaks against the development of secondary mutations in
IGF1R. Other mechanisms of resistance include heterodimerization with other receptor tyrosine kinases31 and
nuclear translocation of adapter proteins such as IRS1,32,33
and IGF1R itself, which is mediated by SUMOylation, a
post-translational modification.34 The low but consistent
response rate seen for IGF1R inhibitors increases our appreciation for the multiple levels of regulation that exist to
ensure cellular homeostasis. At the same time, new mechanisms complementary to kinase blockade are needed to yield
better combinations of anticancer therapeutics.
ALK Inhibition in Inflammatory Myofibroblastic Tumors
Inflammatory myofibroblastic tumors (IMTs) are a rare

subset of sarcomas of intermediate biologic potential with a
predilection for local recurrence and, much less commonly,
metastasis.35 Recurrent cytogenetic alterations involving
the short arm of chromosome 2 and immunohistochemical
overexpression of the ALK gene that resides in this chromosome36 has led to the identification of translocations of ALK
in approximately 50% of IMTs.37 In IMTs, ALK encodes a
receptor tyrosine kinase that can be fused to one of at least five
other gene partners. ALK expression in IMT reliably predicts ALK gene rearrangement. Tumors without ALK rearrangement appear to have a higher propensity for distant
metastases without an increased risk of local recurrence.38
Chromosomal translocations of ALK have been identified
in a subset of nonsmall cell lung cancer (NSCLC),39 and
point mutations in ALK have been reported in neuroblastoma.40 The discovery of activating structural rearrangements and point mutations among different lineages
strongly supported the oncogenic role of ALK, which was
corroborated in a variety of in vitro models as well. Concurrent with many of these discoveries was the early-phase
clinical development of crizotinib (formerly PF-02341066),
an orally administered inhibitor of the MET and ALK
tyrosine kinases. Dramatic clinical activity was observed in
patients with ALK-rearranged NSCLC41 ultimately leading
to regulatory approval for this indication.
Two patients with IMTs were enrolled in a phase I study
of crizotinib.42 A sustained PR was observed in a patient
with an ALK-rearranged tumor, whereas no activity was
seen in the patient without the translocation. Of note, a

F1174L point mutation was observed in the first patient,
who developed secondary resistance to crizotinib. This point
mutation was previously identified in neuroblastoma specimens and demonstrated in vitro resistance to crizotinib but
sensitivity to TAE684, a chemically distinct ALK inhibitor
that has not yet been clinically tested in patients with
IMT.43
Together, these genetic, pathologic, and limited clinical
findings demonstrate that a subset of IMT have activating
rearrangements of ALK, and provide preliminary evidence
that IMT patients with ALK rearrangements may derive
clinical benefit from ALK inhibitors. Further clinical experience with ALK inhibitors in this disease will strengthen
the translation of these observations.
mTOR Activation in Perivascular Epithelioid
Cell Tumor
The perivascular epithelioid cell tumor (PEComa) family

consists of mesenchymal neoplasms with epithelioid morphology and myomelanocytic differentiation.44 These tumors present as small pulmonary nodules with cystic lung
destruction in lymphangioleiomyomatosis (LAM), benign
renal masses with vascular, myogenic, and adipocytic differentiation in angiomyolipoma (AML), or PEComa, an epithelioid tumor with nests and sheets of cells with clear-togranular eosinophilic cytoplasm arising most commonly in
the uterus, retroperitoneum, and somatic soft tissues. Although LAM, AML, and PEComa may arise sporadically,
LAM and AML in particular are also found in patients with
tuberous sclerosis, an autosomal dominant syndrome caused
by inherited mutations in the TSC1 or TSC2 tumor suppressor
genes. Inactivation of the TSC1/2 complex leads to hyperactivation of mTOR and dysregulated cellular proliferation.
An evaluation of the mTOR inhibitor sirolimus in patients
AML or LAM45 showed modest reduction in size of AML and
improvement in lung function, supporting the mechanistic
role of mTOR activity in these diseases. On the basis of
the morphologic similarity of PEComa to AML and LAM,
biochemical evidence of enhanced mTOR activity in
PEComa,46,47 and chromosomal abnormalities at the TSC1
or TSC2 loci, the effects of sirolimus was explored in three
consecutive patients with malignant PEComa.48 All three
patients experienced clinical benefit, but with duration varying from several months to several years. Retrospective
analysis of tumor specimens demonstrated loss of expression
of TSC2 in all tumors, but not in healthy tissues, and loss of
both copies of TSC1 in one tumor. These findings support
the use of mTOR inhibitors in this disease. Similar clinical
responses to mTOR inhibitors have been reported in other
case reports,49,50 but not all patients have derived such
benefit. Further studies identifying predictive factors and

mediators of resistance are needed.
Hedgehog Signaling in Chondrosarcoma
Chondrosarcomas are malignant neoplasms that most

commonly arise in bone and have cartilaginous differentiation. No effective systemic therapies have been described for
unresectable or metastatic disease. The hedgehog (HH)
signaling pathway is implicated in normal chondrocyte development as well as in malignant transformation. Oncogenic activation of the HH pathway can result from multiple
insults, including loss-of-function mutations in Patched (ligand) that lead to constitutive activation of the receptor,
Smoothened (SMO) as well as dysregulated expression of
HH ligands leading to paracrine or autocrine tumor cell
stimulation.
Constitutively activated HH signaling has been identified
in chondrosarcoma tumors.51 IPI-926, an orally available
SMO antagonist, has been evaluated on primary chondrosarcoma xenograft model.52 Treatment of established xenografts with IPI-926 resulted in substantially smaller tumor
sizes compared with chemotherapy- or control-treated animals. Consistent with HH pathway modulation, reduction in
expression of target genes was also observed in both the
tumor cells and the surrounding tumor-associated stroma.
These laboratory findings formed the basis of an ongoing
phase II study of IPI-926 compared with placebo in patients
with unresectable or metastatic chondrosarcoma.
Recently, mutations in the Krebs cycle enzymes isocitrate
dehydrogenase 1 (IDH1) and isocitrate dehydrogenase 2
(IDH2) have been identified in more than 50% of cartilaginous tumors.53 Similar mutations have been suggested to
confer an oncogenic neoactivity in gliomas and acute myelogenous leukemia. Whereas the precise role of IDH1 and IDH2
mutations has yet to be described in chondrosarcoma behavior, these, too, may present an opportunity for targeted
therapeutic approaches that can be explored when suitable
drugs are available for clinical study.
Conclusion
The examples described in this article validate the concept

that deeper understanding of the molecular mechanisms
associated with the development of sarcoma positively affects patient outcomes. Until our understanding of individual sarcoma subtypes evolves further, it is essential to
conduct clinical trials with biologic correlatives designed to
gain a better understanding of potential subtypes likely to
benefit from novel targeted therapies. The rarity and heterogeneity of soft tissue and bone sarcoma necessitates that
our basic scientists and clinical colleagues continue collaboration to bring observations from both preclinical and clinical realms to the forefront to more rapidly provide effective
therapies to our patients.
655
RIEDEL, MAKI, AND WAGNER
Author
Richard F. Riedel
Robert G. Maki
Andrew J. Wagner
Employment or
Leadership
Positions
Consultant or
Advisory Role
Merck
Stock
Ownership
Bayer; Eisai;
Lilly; Novartis
Honoraria
ARIAD; Novartis;
Pfizer
Novartis;
Ziopharm
Oncology
Astex
Therapeutics (U);
EMD Serono;
Genentech;
Infinity;
Morphotek;
Novartis; Pfizer;
Sanofi
Research
Funding
ARIAD; Novartis;
Ziopharm
Oncology
Pfizer; Roche
Expert
Testimony
Other
Remuneration
ArQule;
Genentech;
Prolexys
REFERENCES
2010;60:277-300.
2. Baird K, Davis S, Antonescu CR, et al. Gene expression profiling of
human sarcomas: insights into sarcoma biology. Cancer Res. 2005;65:92269235.
3. Chawla SP, Staddon AP, Baker LH, et al. Phase II study of the
mammalian target of rapamycin inhibitor ridaforolimus in patients with
advanced bone and soft tissue sarcomas. J Clin Oncol. 2012;30:78-84.
4. Chawla SP, Blay J, Ray-Coquard IL, et al. Results of the phase III,
placebo-controlled trial (SUCCEED) evaluating the mTOR inhibitor ridaforolimus (R) as maintenance therapy in advanced sarcoma patients (pts)
following clinical benefit from prior standard cytotoxic chemotherapy (CT).
J Clin Oncol. 2011;29(suppl; abstr 10005).
5. Demetri GD, Chawla SP, Ray-Coquard I, et al. Exploratory analysis of
potential predictive markers to identify sensitive/responder sarcoma patients
with ridaforolimus in the phase 3 randomized placebo-controlled trial (SUCCEED). Poster presented at: Combined Meeting of the Connective Tissue
Oncology Society/Musculoskeletal Tumor Society; October 2011; Chicago, IL.
6. DAdamo DR, Anderson SE, Albritton K, et al. Phase II study of
doxorubicin and bevacizumab for patients with metastatic soft-tissue sarcomas. J Clin Oncol. 2005;23:7135-7142.
7. Verschraegen CF, Arias-Pulido H, Lee SJ, et al. Phase IB study of the
combination of docetaxel, gemcitabine, and bevacizumab in patients with
advanced or recurrent soft tissue sarcoma: the Axtell regimen. Ann Oncol.
Epub 2011 Jul 11.
8. Agulnik M, Okuno SH, Von Mehren M, et al. An open-label multicenter
phase II study of bevacizumab for the treatment of angiosarcoma. J Clin
Oncol. 2009;27:XXX (suppl; abstr 10522).
9. Maki RG, DAdamo DR, Keohan ML, et al. Phase II study of sorafenib in
patients with metastatic or recurrent sarcomas. J Clin Oncol. 2009;27:31333140.
10. George S, Merriam P, Maki RG, et al. Multicenter phase II trial of
sunitinib in the treatment of nongastrointestinal stromal tumor sarcomas.
J Clin Oncol. 2009;27:3154-3160.
11. Sleijfer S, Ray-Coquard I, Papai Z, et al. Pazopanib, a multikinase
angiogenesis inhibitor, in patients with relapsed or refractory advanced soft
tissue sarcoma: a phase II study from the European organisation for research
and treatment of cancer-soft tissue and bone sarcoma group (EORTC study
62043). J Clin Oncol. 2009;27:3126-3132.
12. Grignani G, Palmerini E, Dileo P, et al. A phase II trial of sorafenib in
relapsed and unresectable high-grade osteosarcoma after failure of standard
multimodal therapy: an Italian Sarcoma Group study. Ann Oncol. 2012;23:
508-516.
13. Van Der Graaf WT, Blay J, Chawla SP, et al. PALETTE: a randomized,
double-blind, phase III trial of pazopanib versus placebo in patients with
soft-tissue sarcoma whose disease has progressed during or following prior
chemotherapyan EORTC STBSG Global Network Study (EORTC 62072).
J Clin Oncol. 2011;29 (suppl; abstr LBA002).
14. Gardner K, Judson I, Leahy M, et al. Activity of cediranib, a highly
potent and selective VEGF signaling inhibitor, in alveolar soft part sarcoma.
J Clin Oncol. 2009;27:15s (suppl; abstr 10523)
15. Stacchiotti S, Negri T, Zaffaroni N, et al. Sunitinib in advanced alveolar
soft part sarcoma; evidence of a direct antitumor effect. Ann Oncol. 2011;22:
1682-1690.
16. Stacchiotti S, Negri T, Palassini E, et al. Sunitinib malate and figitu-
656
mumab in solitary fibrous tumor; patterns and molecular bases of tumor

response. Mol Cancer Ther. 2010;9:1286-1297.
17. Park MS, Patel SR, Ludwig JA, et al. Activity of temozolomide and
bevacizumab in the treatment of locally advanced, recurrent, and metastatic
hemangiopericytoma and malignant solitary fibrous tumor. Cancer. 2011;117:
4939-4947.
18. Miriuka SG, Rao V, Peterson M, et al. mTOR inhibition induces
endothelial progenitor cell death. Am J Transplant. 2006;6:2069-2079.
19. Dancey JE, Monzon J. Ridaforolimus: a promising drug in the treatment of soft-tissue sarcoma and other malignancies. Future Oncol. 2011;7:
827-839.
20. Baker LH, Rowinsky EK, Mendelson D, et al. Randomized, phase II
study of the thrombospondin-1-mimetic angiogenesis inhibitor ABT-510 in
patients with advanced soft tissue sarcoma. J Clin Oncol. 2008;26:5583-5588.
21. Dowlati A, Robertson K, Cooney M, et al. A phase I pharmacokinetic
and translational study of the novel vascular targeting agent combretastatin
a-4 phosphate on a single-dose intravenous schedule in patients with advanced cancer. Cancer Res. 2002;62:3408-3416.
22. Bilenker JH, Flaherty KT, Rosen M, et al. Phase I trial of combretastatin a-4 phosphate with carboplatin. Clin Cancer Res. 2005;11:1527-1533.
23. van Valen F, Winkelmann W, Jurgens H. Type I and type II insulin-like
growth factor receptors and their function in human Ewings sarcoma cells. J
Cancer Res Clin Oncol. 1992;118:269-275.
24. Tolcher AW, Sarantopoulos J, Patnaik A, et al. Phase I, pharmacokinetic, and pharmacodynamic study of AMG 479, a fully human monoclonal
antibody to insulin-like growth factor receptor 1. J Clin Oncol. 2009;27:58005807.
25. Olmos D, Postel-Vinay S, Molife LR, et al. Safety, pharmacokinetics,
and preliminary activity of the anti-IGF-1R antibody figitumumab (CP751,871) in patients with sarcoma and Ewings sarcoma: a phase 1 expansion
cohort study. Lancet Oncol. 2010;11:129-135.
26. Malempati S, Weigel B, Ingle AM, et al. Phase I/II trial and pharmacokinetic study of cixutumumab in pediatric patients with refractory solid
tumors and Ewing sarcoma: a report from the Childrens Oncology Group.
J Clin Oncol. 2012;30:256-262.
27. Juergens H, Daw NC, Geoerger B, et al. Preliminary efficacy of
the anti-insulin-like growth factor type 1 receptor antibody figitumumab
in patients with refractory Ewing sarcoma. J Clin Oncol. 2011;29:45344540.
28. Pappo AS, Patel SR, Crowley J, et al. R1507, a monoclonal antibody to
the insulin-like growth factor 1 receptor, in patients with recurrent or refractory
Ewing sarcoma family of tumors: Results of a phase II Sarcoma Alliance for
Research through Collaboration study. J Clin Oncol. 2011;29:4541-4547.
29. Huang F, Greer A, Hurlburt W, et al. The mechanisms of differential
sensitivity to an insulin-like growth factor-1 receptor inhibitor (BMS-536924)
and rationale for combining with EGFR/HER2 inhibitors. Cancer Res. 2009;
69:161-170.
30. Potratz JC, Saunders DN, Wai DH, et al. Synthetic lethality screens
reveal RPS6 and MST1R as modifiers of insulin-like growth factor-1 receptor
inhibitor activity in childhood sarcomas. Cancer Res. 2010;70:8770-8781.
31. Nahta R, Yuan LX, Zhang B, et al. Insulin-like growth factor-I receptor/
human epidermal growth factor receptor 2 heterodimerization contributes to
trastuzumab resistance of breast cancer cells. Cancer Res. 2005;65:1111811128.
32. Del Valle L, Wang JY, Lassak A, et al. Insulin-like growth factor I

receptor signaling system in JC virus T antigen-induced primitive neuroectodermal tumors-medulloblastomas. J Neurovirol. 2002;2:138-147 (suppl).
33. Lassak A, Del Valle L, Peruzzi F, et al. Insulin receptor substrate 1
translocation to the nucleus by the human JC virus T-antigen. J Biol Chem.
2002;277:17231-17238.
34. Sehat B, Tofigh A, Lin Y, et al: SUMOylation mediates the nuclear
translocation and signaling of the IGF-1 receptor. Sci Signal. 2010;3:ra10
35. Gleason BC, Hornick JL. Inflammatory myofibroblastic tumours:
where are we now? J Clin Pathol. 2008;61:428-437.
36. Griffin CA, Hawkins AL, Dvorak C, et al. Recurrent involvement of 2p23
in inflammatory myofibroblastic tumors. Cancer Res. 1999;59:2776-2780.
37. Lawrence B, Perez-Atayde A, Hibbard MK, et al. TPM3-ALK and
TPM4-ALK oncogenes in inflammatory myofibroblastic tumors. Am J Pathol.
2000;157:377-384.
38. Coffin CM, Hornick JL, Fletcher CD. Inflammatory myofibroblastic
tumor: Comparison of clinicopathologic, histologic, and immunohistochemical
features including ALK expression in atypical and aggressive cases. Am J
Surg Pathol. 2007;31:509-520.
40. Janoueix-Lerosey I, Lequin D, Brugieres L, et al. Somatic and germline
activating mutations of the ALK kinase receptor in neuroblastoma. Nature.
2008;455:967-970.
41. Shaw AT, Yeap BY, Mino-Kenudson M, et al. Clinical features and
outcome of patients with non-small-cell lung cancer who harbor EML4-ALK.
J Clin Oncol. 2009;27:4247-4253.
42. Butrynski JE, DAdamo DR, Hornick JL, et al. Crizotinib in ALKrearranged inflammatory myofibroblastic tumor. N Engl J Med. 2010;363:
1727-1733.
43. Sasaki T, Okuda K, Zheng W, et al. The neuroblastoma-associated
F1174L ALK mutation causes resistance to an ALK kinase inhibitor in
ALK-translocated cancers. Cancer Res. 2010;70:10038-10043.
44. Hornick JL, Fletcher CD. PEComa: What do we know so far? Histopathology. 2006;48:75-82.
45. Bissler JJ, McCormack FX, Young LR, et al. Sirolimus for angiomyolipoma in tuberous sclerosis complex or lymphangioleiomyomatosis. N Engl
J Med. 2008;358:140-151.
46 Kenerson H, Folpe AL, Takayama TK, et al. Activation of the mTOR
pathway in sporadic angiomyolipomas and other perivascular epithelioid cell
neoplasms. Hum Pathol. 2007;38:1361-1371.
47. Pan CC, Chung MY, Ng KF, et al. Constant allelic alteration on
chromosome 16p (TSC2 gene) in perivascular epithelioid cell tumour
(PEComa): genetic evidence for the relationship of PEComa with angiomyolipoma. J Pathol. 2008;214:387-393.
48. Wagner AJ, Malinowska-Kolodziej I, Morgan JA, et al. Clinical activity
of mTOR inhibition with sirolimus in malignant perivascular epithelioid cell
tumors: targeting the pathogenic activation of mTORC1 in tumors. J Clin
Oncol. 2010;28:835-840.
49. Italiano A, Delcambre C, Hostein I, et al. Treatment with the mTOR
inhibitor temsirolimus in patients with malignant PEComa. Ann Oncol.
2010;21:1135-1137.
50. Wolff N, Kabbani W, Bradley T, et al. Sirolimus and temsirolimus for
epithelioid angiomyolipoma. J Clin Oncol. 2010;28:e65-e68.
51. Tiet TD, Hopyan S, Nadesan P, et al. Constitutive hedgehog signaling
in chondrosarcoma up-regulates tumor cell proliferation. Am J Pathol. 2006;
168:321-330.
52. Campbell VT, Nadeson PP, Wang Y, et al. Direct targeting of the
Hedgehog pathway in primary chondrosarcoma xenografts with the Smoothened inhibitor IPI-926. Presented at: Paper presented at: 102nd Annual
Meeting of the American Association for Cancer Research; April 2011;
Orlando, FL.
53. Amary MF, Bacsi K, Maggiani F, et al. IDH1 and IDH2 mutations
are frequent events in central chondrosarcoma and central and periosteal
chondromas but not in other mesenchymal tumours. J Pathol. 2011;224:334-343.
657
WHAT THE BUSY ONCOLOGIST NEEDS TO KNOW

ABOUT GASTROINTESTINAL STROMAL TUMOR
CHAIR
Jaap Verweij, MD, PhD
Erasmus University Medical Center
Rotterdam, Netherlands
SPEAKERS
William D. Tap, MD
Santa Monica, CA
Michael C. Heinrich, MD
Oregon Health & Science University Knight Cancer Institute
Portland, OR
Adjuvant Treatment of Gastrointestinal

Stromal Tumor: The Proof, The Pro, and
the Practice
By Jaap Verweij, MD, PhD
Overview: Gastrointestinal stromal tumors (GIST) are rare

tumors, but they are the most common mesenchymal tumor of
the gastrointestinal tract, driven by mutations in KIT and
PDGF. The KIT and PDGF inhibiting agent imatinib has been
tested as adjuvant postsurgery in GIST patients with an
intermediate or high risk of relapse. Two of three prospective,
randomized controlled studies have meanwhile been reported.
The American College of Surgical Oncology Group (ACOSOG)
in 713 patients reported a relapse-free survival benefit for
adjuvant imatinib given for 1 year, but not an overall survival
benefit. The Scandinavian Sarcoma Group (SSG) performed a
study comparing 1 year of imatinib to 3 years of imatinib. At 3
ASTROINTESTINAL STROMAL tumors are rare tumors, but they are the most common mesenchymal
tumor of the gastrointestinal tract.1 Activating mutations in
the KIT gene and, to a lesser extent, the platelet-derived
growth factor receptor (PDGFR)-alpha gene, are considered
the key drivers of tumor growth in GIST,2 and the immunohistochemical assessment of overexpression of KIT has become common practice for an appropriate diagnosis.
The gold standard treatment for a localized primary GIST
is radical surgical resection. However, with the possible
exception of very small tumors (1 cm) that are usually
incidental findings, all GISTs have the potential to relapse
even after radical surgery.2,3 Most relapses occur within the
first 5 years after surgery, but late relapses after 20 years
and beyond have been reported.4
The estimation of the risk of recurrence in an individual
patient has been difficult. Initial prognostic scores were
based on tumor size, mitotic activity, and tumor site,5-8 but
none were validated or provided a more quantifiable risk.
More recently two formally validated risk-assessment systems after radical surgery were published.4,9 Gold and
colleagues9 used data on 127 patients who underwent complete resection of a localized primary GIST without adjuvant
therapy to develop a nomogram to predict RFS, which was
subsequently validated on two independent data sets of 212
and 148 patients, respectively. Within each data set, an
expert pathologist measured tumor size either before or
after formalin fixation, confirmed the diagnosis of GIST, and
calculated the mitotic index (number of mitoses per 50
randomly selected high-power microscopic fields [HPFs]).
The nomogram is based on tumor size (cm), primary tumor
site (stomach, small intestine, colon/rectum, or other), and
mitotic index (5 or 5 mitoses per 50 HPFs) (Table 1).9 The
relapse rate for high-risk GIST in older criteria roughly
corresponds to a greater than 50% relapse chance at 5 years
according to Golds nomogram that thus adds the tumor
primary site to the equation. Joensuu and colleagues4 pooled
data on 2,560 patients from 10 multicenter and international series, adding presence of tumor rupture and male
gender to the model. They subsequently validated their
model in an independent set of 920 patients. The estimated
15-year recurrence-free survival (RFS) after surgery was
59.9%. This model seems to provide better differentiation for
years the overall survival (OS) in patients with 3 years of

imatinib therapy was similar to the OS in those with 1 year of
imatinib 96% and 94% respectively, while at 5 years these
numbers were 92% and 82% (HR: 0.45; 95% CI [0.22 0.89]; p
0.019). Data from the largest study, conducted by the European Organisation for Research and Treatment of Cancer
(EORTC) in 908 patients randomly assigned to receive either 2
years of adjuvant imatinib or no imatinib, have not yet been
reported. Based on the current evidence, 3 years of imatinib at
a daily dose of 400 mg should be considered in patients with a
50% or higher risk of relapse within 5 years after surgery. The
evidence and the remaining caveats are discussed.
low to intermediate risks, but has the disadvantage of being

more complex than the Gold nomogram.
The relevance of these risk-stratification models is that
they enable distinction between patients who are likely to be
cured by surgery alone and those that in theory may benefit
from adjuvant treatments. The randomized studies on adjuvant imatinib have all included patients with intermediate
and high-risk tumors, although the criteria as used in these
studies were based on nonvalidated models.
Adjuvant Use of Imatinib: Aims, History, and
Nonrandomized Studies
To improve the outcome rates after surgery, adjuvant

treatments are applied. It is important to try to define the
main aim of such treatments. Since early application of
drugs in theory could also lead to early acquired tumor cell
resistance, and thus have a detrimental effect, overall survival (and thus cure) seems to be the most favored endpoint.
However, interestingly, a standard and commonly accepted
endpoint seems to be lacking (Table 2).
Imatinib, a tyrosine kinase inhibitor of the KIT and
PDGFR receptors, which drive the GIST phenotype, has
become the backbone of treatment for metastatic GIST.10-12
Given the magnitude of effect in that setting, there was an
obvious rationale to explore imatinib in the adjuvant setting.
Before proper studies were done, physicians in the United
States started to use adjuvant imatinib. In a retrospective
analysis of the National Cancer Data Base spanning 2001 to
2007, Bilimoria and colleagues13 reported that from 2001 to
2003 and from 2006 to 2007, the adjuvant use of imatinib
increased from 29% to 47%, without any formal evidence
supporting this use and before U.S. Food and Drug Administration approval for this indication.
Nonrandomized studies on adjuvant treatments cannot be
From the Department of Medical Oncology, Erasmus University Medical Center-Daniel

den Hoed Oncologic Center, Rotterdam, Netherlands.
Address reprint requests to Jaap Verweij, MD, PhD, Department of Medical Oncology,
Erasmus University Medical Center-Daniel den Hoed Oncologic Center, s-Gravendijkwal
230, 3015 CE Rotterdam, Netherlands; email: j.verweij@erasmusmc.nl.
1092-9118/10/1-10
659
JAAP VERWEIJ
Table 1. Risk Criteria for Relapse after Primary Surgery for GIST
Author
No. Patients
Initial
No. Patients
Validation
Criteria
Gold9
Joensuu4
127
2,560
360
920
Size, site, mitotic rate

Size, site, mitotic rate, rupture, gender
used for decision making since this design is unable to

exclude major selection bias. This is best exemplified by the
study of Li and colleagues14 who compared a group of 56
patients self-selected for 3 years of treatment with imatinib
postsurgery to a group of self-selected patients that rejected
the offer of postoperative imatinib. The latter control
group14 performed much worse than the control group in
either of the two prospectively randomized studies that will
be discussed. Here, I will therefore not discuss the other
performed nonrandomized studies.
Adjuvant Imatinib: Randomized Studies
Three randomized phase III studies on the adjuvant use of

imatinib have now been performed (Table 3), two of which
have been published, either as a full paper15 or abstract.16
DeMatteo and colleagues, through ACOSOG (trial Z9001)15
performed a randomized phase III, double-blind, placebocontrolled trial in patients after complete resection of a
primary GIST at least 3 cm in diameter and positive for the
KIT protein by immunohistochemistry. Seven hundred
seventy-eight patients were registered, but because of randomization problems only 713 were formally randomly assigned to either 400 mg imatinib daily (359 patients) or
placebo (354 patients) daily for 1 year. Patients and investigators were blinded to the treatment group. Cross-over to
or re-treatment with imatinib was allowed in case of tumor
recurrence. The primary endpoint was RFS. Accrual in this
study was terminated early on recommendation of the Independent Data Monitoring Committee (IDMC) because the
trial results crossed the interim analysis efficacy boundary
for RFS. The intention to treat analysis is taken for the
purpose of this review.
At a median follow-up of 19.7 months, 70 patients (20%) in
the placebo group as compared with 30 (8%) in the imatinib
group had tumor recurrence or had died. Imatinib significantly improved 1-year RFS compared with placebo (98% vs.
83%; HR: 0.35, p 0.0001). Overall survival, however, was
fully identical in both arms up to 48 months.15
A later analyses of the trial by mutation subtype held up
KEY POINTS
660
Validated models for postsurgery risk assessment are

now available.
Adjuvant therapy aims to improve survival.
Adjuvant imatinib given for 3 years at a daily dose of
400 mg improves overall survival in patients at high
risk of relapse.
Evidence is currently based on relatively limited
numbers.
Data from future studies should be used when available, to re-evaluate the evidence for adjuvant imatinib use.
Table 2. Suggested Aims for Adjuvant Treatment

Internet lay sites: Adjuvant treatment is aimed to improve overall treatment
outcome, and together with the initial treatment forms the curative therapy
Oncoline: Adjuvant systemic drug therapy is given after primary local treatment
with the aim to eradicate possible occult metastases and increase cure rates
ASCO guidelines: Adjuvant systemic therapy is aimed to improve clinically
meaningful outcomes (i.e., disease-free survival, overall survival, quality of life,
and toxicity) when compared with and/or added to other therapies
NCCN guidelines: No definition of the aim of adjuvant therapies given
ESMO guidelines: Aim differs per tumor type. Mostly disease-free survival
Abbreviations: ASCO, American Society of Clinical Oncology; NCCN, National
Comprehensive Cancer Network; ESMO, European Society of Medical Oncology.
the statistically significant difference in RFS after 2 years

for patients with KIT exon 11 and PDGFRA mutations
(excluding PDGFRA D842V), but not for KIT wild-type and
exon 9 mutations.17 A further update is expected this summer.
The SSG and the German Working Group on Medical
Oncology (Arbeitsgemeinschaft Internistische Onkologie)
have recently reported results of a smaller (400 patients)
randomized controlled trial comparing adjuvant imatinib for
3 years with adjuvant imatinib for 1 year, both at a daily
dose of 400 mg, in high-risk GIST.16 At a median follow-up
of 54 months, 42% of patients in the 1-year treatment group
had GIST recurrence, compared with 25% of patients in the
3-year treatment group.16 In the intention to treat analysis
at 3 years, the RFS for the 3 years of imatinib therapy was
87% compared with 60% for those given the drug for 1 year,
which translated into 66% and 48%, respectively, at 5 years
(HR: 0.46; 95% CI [0.32, 0.65]; p 0.0001).16 At 3 years the
OS in patients with 3 years of imatinib therapy was similar
to the OS in those with 1 year of imatinib (96% and 94%,
respectively), while at 5 years these numbers were 92% and
82% (HR: 0.45; 95% CI [0.22 0.89]; p 0.019).16 Although
statistically the analysis was robust, we still should realize
that because of the relatively small size of the study the
difference in OS is based on seven more patients dying in the
1-year treatment group. Death due to GIST occurred in 14
patients (7%) in the 1-year group and in 7 (4%) in the 3-year
group. The numbers at 5 years were even smaller and at
that point the difference is based on three events.
The third study, which has not yet been reported, is also
the largest study, accruing 908 patients and coordinated by
EORTC, in an intergroup setting with the Austral-Asian
Gastro-Intestinal Tumor Group, the French Sarcoma Group
(FSG), the Italian Sarcoma Group, and the Grupo Espanola
Investigaciones Sarcomas. This study randomly selected
patients at intermediate or high risk of relapse to either
receive no further adjuvant therapy or 2 years of daily
imatinib at 400 mg. The primary endpoint was overall
survival. The study was closed to accrual in October 2008
Table 3. Randomized Studies on Adjuvant Imatinib in GIST
Group
Randomization
No. of
Patients
Ineligible
(%)
ASCOSOG
SSG
EORTC
1 yr 400 mg dd versus control

1 yr 400 mg dd versus 3 yr 400 mg dd
2 yr 400 mg dd versus control
713
400
908
9.1
9.8
NR
Abbreviations: ACOSOG, American College of Surgical Oncology Group; dd,

daily dose; SSG, Scandinavian Sarcoma Group; EORTC, European Organisation
for Research and Treatment of Cancer; NR, not yet reported.
ADJUVANT THERAPY IN HIGH-RISK GASTROINTESTINAL STROMAL TUMORS
and has thereafter been reviewed by the EORTC IDMC on a

yearly basis. Based on IDMC recommendations and the
recent awareness that late imatinib given at progression of
disease could in theory make up for early adjuvant imatinib
without the theoretical disadvantage of inducing early resistance and that in the overall assessment, retreatment with
imatinib also has to be taken into account, it was decided to
change the primary endpoint of this study to time to imatinib failure: the start of any new systemic therapy other
than imatinib or death due to any cause. The first analysis
based on this new endpoint is currently expected in the first
half of 2012.
What Do These Studies Tell Us?
Importantly, the data details of the ACOSOG and SSG

studies have not been publically presented in the same way.
This renders it difficult to assess potential effects of differences in populations in prognostic parameters. Yet the
majority of patients in the studies had tumors with high risk
of relapse so the observations likely hold for this population.
Both studies indicate that adjuvant imatinib given for either
1 or 3 years is able to improve RFS. But as indicated this is
not the primary aim of adjuvant therapy, and in an editorial
Hohenberger has even described the use of RFS in this
setting as a self-fulfilling prophecy.18 Bearing in mind that
there may be differences in the study populations, it is
interesting to analyze the RFS curves. It is remarkable that
the RFS in the patients given 1 year of imatinib in the SSG
study is similar to the RFS of the control group in the
ACOSOG study, while the curves for the respective 3-year
and 1-year groups also seem to overlap. Of concern is the fact
that at the time of reporting both studies showed bananashaped curves, suggesting that we can yet not exclude that
the beneficial effects, if any, may be temporary.
Importantly, if we also project in the curves (both for RFS
and OS) of the nonrandomized comparative study14 it shows
that the control group in that study performed worst of all
control groups, while the treated group outperformed all
treatment groups. This in itself supports the notion that
nonrandomized studies can be heavily biased by patient
selection and should be ignored in deciding on practice
recommendations.
Plotting the data of the two published studies from the
time that patients went off drug, there again are remarkable
differences but also some similarities. The SSG study arms
seem to perform worse than those of the ACOSOG study, but
all arms suggest a steady drop-off rate, and the curves run in
parallel. This seems to suggest that stopping imatinib in
some patients leads to rapid relapse. This would be in line
with the observations in the FSG study in metastatic disease19 that has shown that patients in whom imatinib was
withheld while disease was not progressing rapidly experienced relapse. Yet, while in metastatic disease these data
support the continuation of imatinib as long as patients
seem to benefit, the absence of indisputable overall survival
benefit in the adjuvant setting should lead to a conservative
approach to the duration of adjuvant imatinib treatment.
For OS a limitation of all studies is that they could not
exclude cross-over to or retreatment with imatinib at recurrence, which may affect OS results. This is why the EORTC
decided to change its primary endpoint after the study had
fully accrued. The ACOSOG study15 has only presented OS
up to 4 years, while the SSG study16 has presented longer
follow-up. Again, remarkably, at 4 years the 1-year treated

group in the SSG study seems to have worse outcome than
any of the ACOSOG study arms. This may point to differences in the study populations. The ACOSOG study has not
yet shown any benefit in overall survival. The SSG study, as
indicated, shows a statistically significant difference in favor
of the 3-year treatment. If the ACOSOG study does not show
a difference between no treatment and 1-year treatment, in
theory this 1-year treatment is an adequate control in the
SSG study. However, if the above-mentioned differences
between the 1-year groups in the two studies cannot be
explained on population differences, the observed difference
in the SSG study might solely be based on a worse performance of the 1-year group. Yet, since the statistics support
this, for the time being we can consider the outcome of the
SSG study as evidence supporting the use of 3 years of
adjuvant imatinib in the high-risk population. Still, caution
needs to be taken in interpreting the outcome of the study
since it is relatively small and differences are based on small
numbers. For the same reason of small patient numbers in
individual subgroups, further investigations are still warranted. One could wonder why the IDMC of the EORTC
study has not decided to unblind the data of that study now
that the SSG study shows a survival benefit. In theory, this
could mean the study does not show a survival benefit and
data need to further mature. It will be very interesting to see
the report of that study later this year or early next year.
Conclusion and Recommendation
With all of the caveats mentioned, for the time being

adjuvant imatinib can be considered the standard of care
in patients with a high risk of GIST relapse. The most
important support for this statement comes from the overall
survival benefit suggested in the SSG trial. The current
consensus seems to be to consider patients with a greater
than 50% chance of relapse within 5 years according to the
Gold nomogram as candidates for such treatment. Although
not fully based on evidence, because of their rare occurrence,
patients with neurofibromatosis type 1associated GIST,
which strongly expresses wild-type KIT and has a very
indolent behavior, and those with PDGFR-alpha D842V
mutations, which are known to be insensitive to imatinib,17,20 may be excluded from this recommendation. Patients with tumor rupture may best be considered as having
metastatic disease and be treated accordingly.
Longer-term follow-up data as well as publication of the
results of the EORTC study are eagerly awaited and should
lead to a critical reconsideration of the above-mentioned
recommendation.
Based on the studies, adjuvant imatinib should be initiated within 3 months after definitive surgery and preferably
be handled by an expert team after multidisciplinary discussion. Since all studies have applied a dose of 400 mg daily
regardless of the mutation status of the primary tumor, this
is also the dose recommended for standard application of
adjuvant imatinib. Based on the data of the SSG study, this
dose should be given for 3 years. In GIST cases where
preoperative adjuvant imatinib therapy is used for localized
disease for the purpose of organ-sparing tumor shrinkage or
patient refusal of surgery, the duration of neoadjuvant
therapy should be considered as part of the whole adjuvant
treatment duration of 3 years.
Based on the performed studies, during adjuvant treat-
661
JAAP VERWEIJ
ment, patients should be assessed for blood cell counts and

blood chemistry analysis at 3 monthly intervals. Abdominal
imaging is recommended every 3 to 6 months during adjuvant imatinib therapy and after cessation of imatinib ther-
apy every 3 months, for a 2-year period, every 6 months for

the 3 years thereafter, and annually subsequently up to year
10. Recurrence thereafter is rare4 and thus the benefits of
imaging no longer outweigh the risks.
Author
Jaap Verweij
Employment or
Leadership
Positions
Consultant or
Advisory Role
Abbott
Laboratories;
Boehringer
Ingelheim;
Enzon;
GlaxoSmithKline;
InteRNA;
Johnson &
Johnson;
Novartis; Otsuka;
Pfizer; Sanofi;
Teva
Stock
Ownership
Honoraria
Abbott
Laboratories;
Boehringer
Ingelheim; Eisai;
Enzon;
GlaxoSmithKline;
InteRNA;
Johnson &
Johnson; Merck
Serono; MSD
Oncology;
Novartis; Otsuka;
Pfizer; Sanofi;
Sun Pharma;
Synthon; Teva
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
1. Miettinen M, Lasota J. Gastrointestinal stromal tumors: Review on
morphology, molecular pathology, prognosis, and differential diagnosis. Arch
Pathol Lab Med. 2006;130:1466-1478.
2. Rubin BP, Heinrich MC, Corless CL. Gastrointestinal stromal tumour.
Lancet. 2007;369:1731-1741.
3. DeMatteo RP, Lewis JJ, Leung D, et al. Two hundred gastrointestinal
stromal tumors: recurrence patterns and prognostic factors for survival. Ann
Surg. 2000;231:51-58.
4. Joensuu H, Vehtari A, Riihimaki J, et al. Risk of gastrointestinal
stromal tumour recurrence after surgery: an analysis of pooled populationbased cohorts. Lancet Oncology. Epub 2011 December 7.
5. Miettinen M, El-Rifai W, L HLS, et al. Evaluation of malignancy and
prognosis of gastrointestinal stromal tumors: a review. Hum Pathol. 2002;33:
478-483.
6. Nilsson B, Bumming P, Meis-Kindblom JM, et al. Gastrointestinal
stromal tumors: the incidence, prevalence, clinical course, and prognostication in the preimatinib mesylate eraa population-based study in western
Sweden. Cancer. 2005;103:821-829.
7. Huang HY, Li CF, Huang WW, et al. A modification of NIH consensus
criteria to better distinguish the highly lethal subset of primary localized
gastrointestinal stromal tumors: a subdivision of the original high-risk group
on the basis of outcome. Surgery. 2007;141:748-756.
8. Goh BK, Chow PK, Yap WM, et al. Which is the optimal risk stratification system for surgically treated localized primary GIST? Comparison of
three contemporary prognostic criteria in 171 tumors and a proposal for a
modified Armed Forces Institute of Pathology risk criteria. Ann Surg Oncol.
2008;15:2153-2163.
9. Gold JS, Gonen M, Gutierrez A, et al. Development and validation of a
prognostic nomogram for recurrence-free survival after complete surgical
resection of localised primary gastrointestinal stromal tumour: a retrospective analysis. Lancet Oncol. 2009;10:1045-1052.
10. Verweij J, Casali P, Zalcberg J, et al. Progression-free survival in
gastrointestinal stromal tumours with high-dose imatinib: randomised trial.
Lancet. 2004;364:1127-1134.
11. Blanke CD, Rankin C, Demetri GD, et al. Phase III randomized,
intergroup trial assessing imatinib mesylate at two dose levels in patients
662
with unresectable or metastatic gastrointestinal stromal tumors expressing

the kit receptor tyrosine kinase: S0033. J Clin Oncol. 2008;26:626-632.
12. Blanke CD, Demetri GD, von Mehren M, et al. Long-term results from
a randomized phase II trial of standard- versus higher-dose imatinib mesylate
for patients with unresectable or metastatic gastrointestinal stromal tumors
expressing KIT. J Clin Oncol. 2008;26:620-625.
13. Bilimoria KY, Wayne JD, Merkow RP, et al. Incorporation of adjuvant
therapy into the multimodality management of gastrointestinal stromal
tumors of the stomach in the United States. Ann Surg Oncol. 2012;19:184191.
14. Li J, Gong JF, Wu AW, Shen L. Post-operative imatinib in patients with
intermediate or high risk gastrointestinal stromal tumor. Eur J Surg Oncol.
2011;37:319-324.
15. Dematteo RP, Ballman KV, Antonescu CR, et al. Adjuvant imatinib
mesylate after resection of localised, primary gastrointestinal stromal tumour: a randomised, double-blind, placebo-controlled trial. Lancet. 2009;373:
1097-1104.
16. Joensuu H, Eriksson M, Hartmann J, et al. Twelve versus 36 months of
adjuvant imatinib (IM) as treatment of operable GIST with a high risk of
recurrence: final results of a randomized trial (SSGXVIII/AIO). J Clin Oncol.
2011;29 (suppl, abstr LBA1).
17. Corless CL, Ballman KV, Antonescu C, et al. Relation of tumor
pathologic and molecular features to outcome after surgical resection of
localized primary gastrointestinal stromal tumor (GIST): Results of the
intergroup phase III trial ACOSOG Z9001. J Clin Oncol. 2010;28 (suppl, abstr
10006).
18. Hohenberger P. Adjuvant imatinib in GIST: a self-fulfilling prophecy,
or more? Lancet. 2009;373:1058-1060.
19. Le Cesne A, Ray-Coquard I, Bui B, et al. Continuous versus interruption of imatinib (IM) in responding patients with advanced GIST after three
years of treatment: a prospective randomized phase III trial of the French
Sarcoma Group. J Clin Oncol. 2007;25 (suppl, abstr 10005).
20. Biron P, Cassier Ph.A, Fumagalli E, et al. Outcome of patients with
PDGFRA D842V mutant gastrointestinal stromal tumor treated with imatinib for advanced disease. J Clin Oncol. 2010;28:710s (suppl, abstr 10051).
Management of Tyrosine Kinase

InhibitorResistant Gastrointestinal
Stromal Tumors
By Christine M. Barnett, MD, and Michael C. Heinrich, MD
Overview: The treatment of gastrointestinal stromal tumors

(GISTs) has been a model for targeted cancer therapy. The
discovery of driver somatic mutations in the KIT and PDGFRA
receptor tyrosine kinases led to a shift of therapy from
conventional cytotoxic chemotherapy to inhibitors of these
receptors. Targeted molecular therapy of GIST has markedly
increased the overall survival of patients with advanced dis-
ISTS ARE the most common sarcoma arising in the

abdominal cavity. The annual incidence in the United
States is approximately 4,000 to 6,000 new cases per year.
The stomach is the most common site of origin, but GISTs
can arise anywhere in the gastrointestinal tract.1 Localized
GIST is primarily managed with surgical resection, but
these tumors can recur locally or metastasize, typically to
the liver or peritoneum. Increasingly, GIST is managed with
systemic therapy utilizing tyrosine kinase inhibitor (TKI)
therapy, even in the case of localized disease.
GIST Biology
Only in the last 15 years has GIST been recognized as a

distinct pathologic entity. Histologically, there is an overlap
of smooth muscle and neural elements, which historically
made classification of GIST difficult. GISTs are now recognized as a malignancy of the interstitial cells of Cajal (ICC)
which serve as pacemaker cells of the gastrointestinal
tract. KIT activity is required for development of certain
types of ICCs,1 and in addition, 95% of GISTs are positive for
KIT expression by immunohistochemistry.2 Notably, the
types of ICCs that are developmentally dependent on KIT
are also the types that can give rise to GIST.
KIT (CD117) is a type III receptor tyrosine kinase. This
family also includes platelet-derived growth factor receptors
A and B (PDGFRA/B), CSF-1R, and FLT3. Binding of stem
cell factor results in homodimerization of the KIT receptor
with subsequent activation of the intracellular tyrosine
kinase domain.3 Tyrosine phosphorylation of KITinteracting proteins leads to activation of multiple signal
transduction pathways that stimulate cell proliferation and
resistance to apoptosis. Under normal circumstances, the
KIT receptor is autoinhibited by the juxtamembrane domain, which prevents the protein activation loop from assuming the necessary conformation for kinase activity.
Oncogenic Mutations in GIST
Mutations in KIT and PDGFRA result in ligandindependent kinase activation. 70% to 80% of GISTs have
mutations in KIT,1 making this the single most important
therapeutic target for this tumor. The majority of KIT
mutations affect the juxtamembrane domain encoded by
exon 11, and up to two-thirds of GISTs harbor a mutation
of this exon. Exon 11 mutations disrupt the autoinhibitory
properties of the juxtamembrane domain, thus allowing
kinase activation.3
ease. However, the ability of kinase therapy to control metastatic disease is ultimately limited by the ability of these
agents to overcome intrinsic or acquired resistance mechanisms. Ongoing basic and clinical research is focusing on
identifying new agents to inhibit KIT/PDGFRA kinase activity
and/or other novel molecular targets in GIST.
7% to 10% of GISTs have mutations of exon 9, which

encodes the membrane-proximal most portion of the extracellular domain. These mutations lead to an extracellular
conformational change that is similar to that created by
ligand binding, resulting in kinase activation.3 Exon 9
mutations are more commonly seen in GISTs of the small
and large bowel origin and are rarely seen in gastric GISTs.
There are also less common primary mutations in the
activation loop (exon 17), which stabilize the active conformation of the receptor, and the ATP binding region (exon
13), which may affect the inhibitory function of the juxtamembrane domain.3
Approximately 10% of patients with GIST have tumors
with activating mutations of PDGFRA,1 which is closely
related to KIT. KIT and PDGFRA mutations are mutually
exclusive. The downstream pathways that are activated by
PDGFRA are very similar to those activate by KIT. Most
PDGFRA mutations occur in exon 18, which encodes the
activation loop of the kinase. In addition, approximately 1%
of GISTs have mutations of the PDGFRA juxtamembrane or
ATP-binding domains.
The remaining 10% to 15% of GISTs lack mutations in
PDGFRA or KIT and have been traditionally designated
wild type because of a lack of mutated kinase targets for
therapy. Interestingly, KIT is still strongly phosphorylated
in these tumors, indicating the presence of activated KIT.4
The mechanism responsible for this activation remains
undefined. Increasingly, alternative oncogenic drivers have
been identified in tumors lacking KIT or PDGFRA mutations. These include BRAF V600E mutations or, less commonly, mutations of RAS family members.3 These make up
an estimated 10% of wild-type GISTs, or 1% to 2% of GISTs
overall. Rare loss-of-function mutations of the succinate
dehydrogenase (SDH) complex have been identified in some
familial syndromes associated with GIST (e.g., CarneyStratakis).3 SDH is important in the regulation of HIF1alpha, and loss of SDH complex function has similar biologic
consequences as loss of VHL protein (the primary molecular
From the Division of Hematology and Medical Oncology, Portland VA Medical Center
and the Knight Cancer Institute, Oregon Health & Science University, Portland, OR.
Address reprint requests to Michael C. Heinrich, MD, Division of Hematology/Oncology,
Departments of Medicine and Cell and Developmental Biology, Portland VA Medical Center
and OHSU Knight Cancer Institute, Oregon Health & Science University, R&D-19 3710
U.S. Veterans Hospital Road, Portland, OR 97239; email: heinrich@ohsu.edu.
1092-9118/10/1-10
663
BARNETT AND HEINRICH
defect in most cases of renal cell carcinoma). In addition, loss

of SDHB protein expression is seen in pediatric GISTs and
some GISTs with pediatric features that arise in adult
patients. The underlying molecular defect in these wild-type
GISTs likely has significant influence on the response (or
lack thereof) to TKI therapy.
Initial Treatment with TKIs
Before 2000, patients were treated with conventional

chemotherapy, resulting in very poor response rates and no
effect on overall survival. In 2000, this situation was dramatically altered by the first trials of TKI therapy for
advanced or unresectable GIST.
Imatinib was originally developed as a treatment for
chronic myelogenous leukemia (CML). The fusion protein
BCR-ABL is the target for imatinib in CML,5 and the
knowledge that ABL is structurally similar to KIT led to
preclinical investigation of the ability of imatinib to inhibit
KIT signaling.6 In these studies, imatinib was shown to
inhibit both wild-type and mutant KIT protein, including
KIT exon 11 or 13 mutations that are found in GIST.6,7 In
addition, imatinib was also found to inhibit PDGFRA/B
kinase activity.
Based in part on these preclinical studies, imatinib was
used to treat a single patient with widely metastatic disease.
Based on the remarkable tumor response in this patient,
imatinib was further tested in a phase I study of patients
with GIST or other soft tissue sarcoma. In the European
(EORTC) phase I study, 36 patients with GIST were treated
with imatinib doses ranging from 400 to 800 mg.8 Notably,
only four patients demonstrated progressive disease; all
others demonstrated response or stable disease (by RECIST
criteria).
This very successful phase I trial lead to two phase II
trials, one by the EORTC and another by a U.S.-Finnish
group.9 Both trials compared doses of 400 mg compared with
600 mg of imatinib. Overall, 89% of patients in the EORTC
trial and 87% of patients in the U.S.-Finnish trial had stable
KEY POINTS
664
Imatinib is the front-line treatment for patients with

unresectable or advanced GIST.
Primary imatinib resistance is usually related to the
pretreatment biology of the GIST and is more common in patients whose tumor has one of the following
genotypes: wild-type, KIT exon 9-mutant, or PDGFRA D842V.
Secondary imatinib resistance is typically due to the
expansion of tumor clones with acquired kinase mutations that confer drug resistance.
Sunitinib treatment is indicated for patients with
imatinib intolerance and/or imatinib-resistant tumors.
The optimal treatment for patients with sunitinibresistant tumors is unknown. Treatment options include participation in a clinical study, off-label use
of other KIT tyrosine kinase inhibitors, or palliative
surgery.
disease or partial response as their best response to therapy.

One notable difference between the trials was that patients
in the U.S.-Finnish trial were allowed to cross over to the
600 mg arm if they progressed on 400 mg. 30% of patients
who progressed on the 400 mg dose who crossed over to the
600 mg dose obtained a partial response or stable disease
status.
Following these pivotal trials, two phase III trials were
performed, both similarly designed with the intent to combine them into a meta-analysis.10 These trials compared 400
mg daily with 800 mg daily of imatinib. Both trials allowed
cross-over to 800 mg on progression during treatment with
400 mg per day of imatinib. In EORTC 62005, only 18% of
patients had progressive disease, while in the S0033 study
26% of patients had progressive disease as the best response
to treatment. In the Meta-GIST analysis of these two studies, there was a small progression-free survival (PFS) seen
with the 800 mg dose of imatinib, but this was almost
exclusively due to the superior results seen with the higher
dose among patients with KIT exon 9-mutant GIST. In
contrast, there was no improvement in PFS for GISTs with
non exon-9 mutations treated with high-dose imatinib. Notably, there was no improvement in overall survival in
patients treated with high-dose imatinib.
Disease Persistence
Despite high rates of response, it appears that continuous

therapy is required for optimal disease control. A randomized study of patients who had disease control after 3 years
of imatinib demonstrated a high rate of progression following interruption of treatment.11 The 2-year PFS was only
16% in those who stopped imatinib compared with 80% in
patients who continued on imatinib.
So why doesnt TKI therapy completely eradicate (cure)
GIST? There is evidence that unlike mature GIST cells,
GIST stem and progenitor cells are immune to KIT inhibition.3 Therefore, KIT inhibitors such as imatinib can inhibit
and potentially kill mature cells, but a pool of quiescent stem
and progenitor cells will remain. With cessation of treatment, this pool can relatively quickly repopulate tumors
with differentiated GIST cells. This phenomenon is referred
to as disease persistence and is distinct from clinical resistance that is discussed below.
Imatinib Resistance
Based on the phase II studies, imatinib was approved by

the U.S. Food and Drug Administration (FDA) for first-line
treatment of advanced GIST. Despite these promising results, treatment with imatinib is not curative and GIST
lesions can develop drug resistance after varying intervals of
time. Resistance to imatinib can be divided into two types,
primary resistance for the tumors that progress within the
first 6 months of imatinib treatment, and secondary resistance in tumors that progress after this period. Notably,
distinct molecular mechanisms are associated with primary
versus secondary resistance.
The median PFS on front-line imatinib is in the range of
20 to 24 months. In the U.S.-Finnish phase II study cited
above, approximately 18% of patients remain on front-line
imatinib a decade later. Thus, the majority of imatinibtreated patients will eventually develop resistant disease.12
TKI-RESISTANT GIST
Primary Imatinib Resistance
Approximately 10% of treated GIST patients experience

primary tumor resistance to imatinib therapy. The most
common cause of early imatinib resistance is a primary
resistance mutation in PDGFRA or KIT.13 KIT exon 9
mutations have decreased in vitro sensitivity to imatinib,
and larger clinical doses of imatinib may be required to
achieve a response.10 Therefore, early progression on 400 mg
of imatinib is common for patients with KIT exon 9 mutant
GIST. In addition, the most common PDGFRA mutation,
D842V, is strongly resistant to imatinib in vitro. This
mutation changes the receptor to the active conformation to
which imatinib cannot bind. Recently, mutations of downstream signaling pathways (KRAS, BRAF, or PIK3CA) have
been reported to coexist with KIT mutations in a minority of
patients. These mutations could reduce or eliminate the
efficacy of upstream KIT inhibition by maintaining signaling
through KIT-dependent downstream signaling pathways.3
Wild-type GISTs (i.e., no KIT or PDGFRA mutation) also
have a higher rate of primary resistance to imatinib than
the typical KIT exon 11-mutant GIST. The rate of response
likely correlates with the underlying molecular defect, but
definitive genotype/outcome data are not available. For
instance, BRAF exon 15 mutations have been found in this
wild-type group, and these mutations are not sensitive to
imatinib. Additionally, those GISTs with abnormalities in
the SDH complex appear to be less responsive to imatinib.3
In addition to the effect of the underlying tumor biology,
there is controversy about whether a threshold drug level
of imatinib is needed to see a response in GIST. There is no
prospective data looking at the proper imatinib levels
needed to achieve a response in GIST. However, in a
retrospective subgroup analysis of one of the phase II
studies, a minimum trough level of 1,100 ng/mL of imatinib
was required for optimal results.14 However, the utility of
imatinib levels to adjust dosing remains unproven.
Secondary Imatinib Resistance
Eventual progression of disease is an issue for the vast

majority of patients treated with imatinib. This is likely due
to expansion of GIST clones with secondary kinase mutations.15 Secondary KIT or PDGFRA mutations have only
been found in patients who had a primary mutation, that is,
not in wild-type patients. In addition, the secondary mutations occur in the same gene and on the same allele (in cis)
as the primary mutation.15 Tumors with a primary exon 11
mutation are the most likely to develop a secondary KIT
mutation. In the above mentioned U.S.-Finnish phase II
trial of imatinib, two-thirds of patients with progression on
imatinib were found to have secondary mutations in one or
more post-treatment samples. These secondary mutations
occur primarily in the activation loop (exons 17 and 18) or
ATP binding pocket (exons 13 and 14). The situation is
further complicated by multiple studies demonstrating significant intra- and interlesional heterogeneity of secondary
mutations.16 In vitro profiling of alternative KIT/PDGFRA
TKIs (e.g., sunitinib) has shown that most inhibitors lack
activity against all relevant secondary mutations. Thus
heterogeneity of resistance mutations in imatinib-resistant
GIST can result in differential responses of lesions to
second- or third-line TKIs. In addition to complicating rou-
tine clinical care, this situation also can obscure the potential benefit of new agents tested in clinical studies.
With respect to the model of GIST stem and progenitor
cells discussed above, secondary resistance is due to the
appearance of clones with secondary mutations. These
clones could arise from either stem/progenitor cells or from
more differentiated cell populations. It is yet unclear if these
secondary, more resistant, kinase mutations are acquired
during drug therapy, or pre-exist and simply expand under
the selective pressure of TKI therapy.
Second-Line TKI Therapy
For patients with documented primary or secondary

imatinib-resistant GIST, the suggested initial intervention
is to increase the dose of imatinib to 800 mg if the patient is
on a lesser dose. Most patients will not benefit from imatinib
dose escalation, with a median PFS following dose increase
of only 3 months. However, 15% to 25% of patients may have
more prolonged responses that can last for many months or
years.9 This may be due to inadequate drug levels with the
original dose or in some cases may be due to overcoming
secondary mutations with the higher imatinib dose.
If there is progression on the 800 mg dose of imatinib or if
the patient cannot tolerate the 800 mg total daily dose, the
next FDA-approved treatment is to switch to sunitinib. In a
phase III study, sunitinib-treated patients had a median
time to progression of disease of 27.3 weeks compared with
6.4 weeks with placebo.17 These results were obtained using
the typical sunitinib dosing schedule of 50 mg daily for 4
weeks with 2 weeks off. Notably, a phase II trial of continuous daily dosing of sunitinib at 37.5 mg produced a similar
time to progression at 34 weeks.18 Therefore, given better
tolerability, many GIST experts will recommend starting
patients on continuous daily dosing of sunitinib.
Because sunitinib was the first medication to be tested in
the imatinib-resistant population, it was the first to get FDA
approval. However, with our increased understanding about
secondary KIT and PDGFRA mutations, it is clear that the
range of activity of sunitinib against these secondary mutations is suboptimal. Sunitinib has substantial activity
against secondary mutations in exons 13 and 14, but cannot
inhibit kinase activity in the presence of mutations in exons
17 or 18.19 Given that there is approximately equal frequency of these secondary mutations in most series, and the
aforementioned heterogeneity of secondary mutations in a
given patient, it is common to see mixed responses during
sunitinib therapy. Sunitinib, unlike imatinib, has activity
against VEGFR1/2.20 However, it appears that most of the
activity addition of this agent against imatinib-resistant
GIST is due to inhibition of KIT rather than inhibition of
angiogenesis.
Third-Line Therapy (And Beyond)
Following progression on sunitinib, the optimal next step

in therapy is unclear. Multiple TKIs have been studied in
this setting, but there has yet to be a clear best therapy for
imatinib- and sunitinib-resistant GIST. Therefore, patient
participation in a clinical study should be strongly considered in this setting.20
There are multiple drugs being tested in clinical trials
that directly or indirectly target KIT, KIT-dependent signaling (e.g., MEK or PI3K), or more remote downstream events
665

Table 1. Investigational Therapies in GIST
Drug Type
Tyrosine kinase
inhibitors
Drug
Target(s)
Imatinib
Sunitinib
Nilotinib
KIT, PDGFR
KIT, PDGFR, VEGFR
KIT, PDGFR
Dasatanib
KIT, PDGFR
Sorafenib
KIT, PDGFR, VEGFR
Regorafenib
KIT, PDGFR, VEGFR
Vatalanib
KIT, PDGFR, VEGFR
Masitinib
(AB1010)
Pazopanib
KIT, PDGFR
KIT, PDGFR, VEGFR
Crenolanib
PDGFR
Dovitinib
VEGFR, PDGFR, FGFR
STA-9090
HSP90
AT-13387
HSP90
AUY922
HSP90
IMC-3G3
(Olaratumab)
Bevacizumab
PDGFR
VEGFR
mTOR inhibitor
Everolimus
mTOR
Miscellaneous
Perifosine
AKT (PI3K pathway)
BKM120
PI3K pathway
TH-302
Tumor hypoxia
HSP90 inhibitors
Monoclonal
antibody
Trial Information
FDA approved
FDA approved
Phase III
NCT00785785
Phase II
NCT00568750
Phase II
NCT01091207
Phase III
NCT01271712
Phase II
NCT00117299
Phase III
NCT00812240
Phase II
NCT01323400
Phase II
NCT01243346
Phase II
NCT01478373
Phase II
NCT01039519
Phase II
NCT01294202
Phase II
NCT01404650
Phase II
NCT01316263
Phase III
NCT00324987
Phase II
NCT00510354
Phase II
NCT00455559
Phase I
NCT1468688
Phase I
NCT01381822
(histone deacetylase inhibitors). These medications are

sometimes used in combination with KIT inhibitors. For
example, multiple heat shock protein 90 (Hsp90) inhibitors
are in phase II trials in GIST. Hsp90 is an important
chaperone protein for oncoproteins such as KIT. Inhibition
of this target leads to preferential degradation of activated
forms of KIT (as well as other activated kinases). There is
strong preclinical data showing that Hsp90 inhibitors have
activity against imatinib-resistant GIST in vitro;21 however,
there is not yet strong clinical data for the use of these
medications in GIST. In addition, there is evidence to
suggest that the downstream PI3K-mTOR pathway is perhaps the most important signaling pathway in GIST3 and
inhibitors of this pathway including everolimus, BKM120,
and perifosine are currently in phase II clinical trials (Table
1). In addition to clinical strategies to inhibit downstream
pathways, multiple TKIs are currently in trials for TKIresistant GIST. Of note, a novel agent, crenolanib, is being
tested in a phase II study for treatment of PDGFRA D842Vmutant GIST. As noted above, this particular mutation is
strongly resistant to imatinib (and sunitinib as well).22
In terms of FDA-approved KIT/PDGFRA kinase inhibitors, a number of these have been tested in phase II studies
of drug-resistant GIST. For example, nilotinib has been
666
studied in a several clinical studies against drug-resistant

GIST.23 Nilotinib is structurally very similar to imatinib
and, like sunitinib, lacks the ability to overcome any of the
activation loop mutations in exons 17 or 18. In addition, like
imatinib, it also has limited potency against ATP binding
pocket mutations. Because of this, nilotinib has shown
limited activity in imatinib- and sunitinib-resistant GIST.
In a phase III trial, 248 patients were randomly assigned to
nilotinib versus best supportive care for third-line therapy.24 There was no statistical difference in overall survival
between nilotinib and best supportive care. Sorafenib has
more promising data for treatment of drug-resistant GIST,
with an approximate 20-week PFS in the third-line, and
even fourth-line setting.25 This may be due to its apparent
ability to overcome some exon 17 mutations in vitro. However, it appears that this agent has been supplanted in
clinical development by its closely related analog, regorafenib. In a recent phase II trial, regorafenib showed a
promising 10-month PFS in the third- or fourth-line setting.26 Based on these results, a randomized, double-blind,
placebo-controlled phase III study was initiated for patients
who were intolerant of or who had progression during prior
first- and second-line therapy. The study completed enrollment in mid-2011 and the final study analysis is planned for
sometime in early 2012.
The National Comprehensive Cancer Network guidelines
suggest consideration of a clinical study or off-label treatment with sorafenib, nilotinib, or dasatinib for physicians
with patients with imatinib- and sunitinib-resistant GIST.20
In addition, reintroduction of imatinib can be considered.
Fumagalli and colleagues have published data supporting
rechallenging patients with imatinib or sunitinib after failure of standard and investigational agents.27 Most GIST
experts believe that life-long continuation of TKI therapy
should be considered an essential component of best supportive care. As noted above, GIST patients often harbor
deposits of quiescent, drug-sensitive tumor cells that will
become clinically active following discontinuation of TKI
therapy. These lesions can add to overall disease burden and
patient symptoms. In addition, there is evidence to suggest
that some secondary drug resistance mutations result in
relative rather than absolute resistance to TKI therapy;
continued TKI treatment may slow (but not arrest) such
clones, potentially palliating symptoms.
Role of Surgery in Advanced GIST
Given that there can be significant heterogeneity in secondary mutations and therefore mixed responses with TKI
treatment of advanced GIST, many patients inquire about
surgery to control their disease. Raut and colleagues published a study about their series of 69 patients who underwent surgery while receiving TKI therapy, with the majority
receiving front-line imatinib (but a minority on second-line
sunitinib). Notably, only those patients with overall stable
disease (i.e., disease controlled by TKIs) strongly benefited
from surgery in terms of disease control (Table 2).28 In
another study, 50 patients receiving second-line sunitinib
underwent surgery for their disease and median PFS was a
mere 5.8 months, and was independent of the immediate
presurgical sunitinib clinical response status. That is,
sunitinib-responding patients did not have better PFS after
surgery than patients with frank progression on sunitinib
before surgery.29 In addition, over half of the patients had
TKI-RESISTANT GIST
Table 2. Clinical outcomes following surgery for TKI treated patients (pts) with advanced GIST. Median progression-free and
overall-survival data are tabulated based on the immediate pre-surgery response status of patients to TKI therapy. a
Progression-Free Survival
Overall Survival
Study
Pt #
Treatment Status
Stableb
Limitc
Gend
Stableb
Limitc
Gend
Raut (28)
Dematteo (30)
Raut (29)
69
40
50
IM 45 pts IM 3 SU 21 pts
IM 37 pts IM 3 SU 3 pts
IM 3 SU 50 pts
40 mo.
48 mo.
11 mo.
8 mo.
12 mo.
4 mo.
3 mo.
3 mo.
4 mo.
40 mo.
48 mo.
36 mo.
30 mo.
19 mo.
19 mo.
6 mo.
11 mo.
9 mo.
a
b
c
d
IM (surgery during front-line imatinib therapy). IM 3 SU (surgery during second-line sunitinib).

Stable or responsive disease prior to surgery.
Limited or focal disease progression prior to surgery.
Generalized disease progression prior to surgery.
complications from surgery, and it is important to keep in

mind that surgical complications usually lead to a delay in
restarting systemic therapy. Therefore, only for patients with
primary disease, or for symptom control in advanced disease,
is surgery generally recommended as a treatment strategy.
Conclusion
The knowledge of the biology of GIST has made this

cancer a model of targeted therapy with tyrosine kinase
inhibitors. Currently, the mainstay of treatment for this

disease is TKI-based therapy with enrollment in a clinical
trial if multiple TKIs fail to control disease. Understanding
the mechanisms of disease persistence and drug resistance
has helped identify new targets for therapy. Hopefully, our
increased understanding of GIST biology will not only help
us improve current treatment, but ultimately help create
new treatments that might cure patients with advanced
disease.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
MolecularMD;
Novartis; Pfizer
MolecularMD
Novartis
Research
Funding
Expert
Testimony
Other
Remuneration
Christine M. Barnett*
Michael C. Heinrich
ARIAD; Arog;
ImClone
Systems;
Novartis; Pfizer
REFERENCES
1. Corless CL, Fletcher JA, Heinrich MC. Biology of gastrointestinal
stromal tumors. J Clin Oncol. 2004;22:3813-3825.
2. Miettinen M, Sobin LH, Sarlomo-Rikala M. Immunohistochemical spectrum of GISTs at different sites and their differential diagnosis with a
reference to CD117 (KIT). Mod Pathol. 2000;13:1134-1142.
3. Corless CL, Barnett CM, Heinrich MC. Gastrointestinal stromal tumours: origin and molecular oncology. Nat Rev Cancer. 2011;11:865-878.
4. Duensing A, Joseph NE, Medeiros F, et al. Protein Kinase C theta
(PKCtheta) expression and constitutive activation in gastrointestinal stromal
tumors (GISTs). Cancer Res. 2004;64:5127-5131.
5. Druker BJ, Tamura S, Buchdunger E, et al. Effects of a selective
inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat
Med. 1996;2:561-566.
6. Heinrich MC, Griffith DJ, Druker BJ, et al. Inhibition of c-kit receptor
tyrosine kinase activity by STI 571, a selective tyrosine kinase inhibitor.
Blood. 2000;96:925-932.
7. Tuveson DA, Willis NA, Jacks T, et al. STI571 inactivation of the
gastrointestinal stromal tumor c-KIT oncoprotein: biological and clinical
implications. Oncogene. 2001;20:5054-5058.
8. van Oosterom AT, Judson I, Verweij J, et al. Safety and efficacy of
imatinib (STI571) in metastatic gastrointestinal stromal tumours: a phase I
study. Lancet. 2001;358:1421-1423.
9. Blanke CD, Demetri GD, von Mehren M, et al. Long-term results from a
randomized phase II trial of standard- versus higher-dose imatinib mesylate
for patients with unresectable or metastatic gastrointestinal stromal tumors
expressing KIT. J Clin Oncol. 2008;26:620-625.
10. Comparison of two doses of imatinib for the treatment of unresectable
or metastatic gastrointestinal stromal tumors: A meta-analysis of 1,640
patients. J Clin Oncol. 2010;28:1247-1253.
11. Le Cesne A, Ray-Coquard I, Bui BN, et al. Discontinuation of imatinib
in patients with advanced gastrointestinal stromal tumours after 3 years of
treatment: an open-label multicentre randomised phase 3 trial. Lancet Oncol.
2010;11:942-949.
12. von Mehren M, Heinrich MC, Joensuu H, et al. Follow-up results after
9 years (yrs) of the ongoing, phase II B2222 trial of imatinib mesylate (IM) in
patients (pts) with metastatic or unresectable KIT gastrointestinal stromal

tumors (GIST). J Clin Oncol. 2011;29 (suppl; abstr 10016).
13. Heinrich MC, Corless CL, Demetri GD, et al. Kinase mutations and
imatinib response in patients with metastatic gastrointestinal stromal tumor.
J Clin Oncol. 2003;21:4342-4349.
14. Demetri GD, Wang Y, Wehrle E, et al. Imatinib plasma levels are
correlated with clinical benefit in patients with unresectable/metastatic
gastrointestinal stromal tumors. J Clin Oncol. 2009;27:3141-3147.
15. Antonescu CR, Besmer P, Guo T, et al. Acquired resistance to imatinib
in gastrointestinal stromal tumor occurs through secondary gene mutation.
Clin Cancer Res. 2005;11:4182-4190.
16. Liegl B, Kepten I, Le C, et al. Heterogeneity of kinase inhibitor
resistance mechanisms in GIST. J Pathol. 2008;216:64-74.
17. Demetri GD, van Oosterom AT, Garrett CR, et al. Efficacy and safety of
sunitinib in patients with advanced gastrointestinal stromal tumour after
failure of imatinib: a randomised controlled trial. Lancet. 2006;368:13291338.
18. George S, Blay JY, Casali PG, et al. Clinical evaluation of continuous
daily dosing of sunitinib malate in patients with advanced gastrointestinal
stromal tumour after imatinib failure. Eur J Cancer. 2009;45:1959-1968.
19. Gramza AW, Corless CL, Heinrich MC. Resistance to tyrosine kinase
inhibitors in gastrointestinal stromal tumors. Clin Cancer Res. 2009;15:75107518.
20. Demetri GD, von Mehren M, Antonescu CR, et al. NCCN Task Force
report: update on the management of patients with gastrointestinal stromal
tumors. J Natl Compr Canc Netw. 2010;8 Suppl 2:S1-41; quiz S2-4.
21. Bauer S, Yu LK, Demetri GD, et al. Heat shock protein 90 inhibition in
imatinib-resistant gastrointestinal stromal tumor. Cancer Res. 2006;66:91539161.
22. Heinrich MC, Corless CL, Duensing A, et al. PDGFRA activating
mutations in gastrointestinal stromal tumors. Science. 2003;299:708-710.
23. Demetri GD, Casali PG, Blay JY, et al. A phase I study of single-agent
nilotinib or in combination with imatinib in patients with imatinib-resistant
gastrointestinal stromal tumors. Clin Cancer Res. 2009;15:5910-5916.
24. Reichardt P, Blay J, Gelderblom H, et al. Phase III trial of nilotinib in
667

patients with advanced gastrointestinal stromal tumor (GIST): First results
from ENEST g3. J Clin Oncol. 2010; 28 (suppl; abstr 10017).
25. Wiebe L, Kasza KE, Maki RG, et al. Activity of sorafenib (SOR) in
patients (pts) with imatinib (IM) and sunitinib (SU)-resistant (RES) gastrointestinal stromal tumors (GIST): a phase II trial of the University of Chicago
Phase II Consortium. J Clin Oncol. 2008;26 (suppl; abstr 10502).
26. George S, von Mehren M, Heinrich MC, et al. A multicenter phase II
study of regorafenib in patients (pts) with advanced gastrointestinal stromal
tumor (GIST), after therapy with imatinib (IM) and sunitinib (SU). J Clin
Oncol. 29 (suppl; abstr 10007).
27. Fumagalli E, Coco P, Morosi C, et al. Sunitinib rechallenge in two
668
advanced GIST patients after third-line anti-tyrosine kinase therapy. J Clin

Oncol. 2010;28 (suppl; abstr e20519).
28. Raut CP, Posner M, Desai J, et al. Surgical management of advanced
gastrointestinal stromal tumors after treatment with targeted systemic
therapy using kinase inhibitors. J Clin Oncol. 2006;24:2325-2331.
29. Raut CP, Wang Q, Manola J, et al. Cytoreductive surgery in patients
with metastatic gastrointestinal stromal tumor treated with sunitinib
malate. Ann Surg Oncol. 2010;17:407-415.
30. DeMatteo RP, Maki RG, Singer S, et al. Results of tyrosine kinase
inhibitor therapy followed by surgical resection for metastatic gastrointestinal stromal tumor. Ann Surg. 2007;245:347-352.
BIOLOGIC PRINCIPLES OF TARGETED

COMBINATION THERAPY
CHAIR
Johann S. de Bono, MB, ChB, MSc, PhD
Sutton, United Kingdom
SPEAKERS
Antoni Ribas, MD
Los Angeles, CA
Geoffrey Shapiro, MD, PhD
Boston, MA
Opportunities and Pitfalls of Targeted

Therapeutic Combinations in Solid Tumors
By Joaquin Mateo, MD, MSc, Michael Ong, BSc, MD,
Timothy A. Yap, BSc, MBBS, MRCP, and Johann S. de Bono, MB ChB, MSc, PhD
Overview: Recent advances in cancer biology have led to the

discovery and development of chemical compounds and drugs
that target specific cellular receptors, mediators, or effectors
that are central to oncogenic survival, growth, and invasion.
However, the complexity of tumor biology makes it is unrealistic to expect an antitumor therapeutic to be successful
based on the inhibition of a single target or even a lone
signaling pathway. Therefore, the potential success of such
targeted therapies is likely to require the development of
multiagent combinations. Combination strategies have a
greater likelihood of addressing issues with genetically complex tumors, potentially avoiding drug resistance mechanisms
OVEL DRUGS aim to selectively inhibit mechanisms

essential to cancer pathogenesis, while limiting offtarget effects and undesired toxicities. Despite the hype,
only a limited number of targeted agents have been successful; this includes the inhibition of ABL and c-KIT by imatinib for chronic myeloid leukemia and in gastrointestinal
stromal tumors (GIST).1 However, primary and secondary
resistance to therapy is common because of mechanisms
such as novel mutations that decrease imatinib binding, or
mutations in downstream proteins involved in the phosphatidylinositol 3-kinase (PI3K) pathway that are c-KIT independent.2,3 Ultimately, the application of a single targeted
agent even in these successful models has not abrogated
the development of resistance, nor has it led to increased
rates of cure.4,5
Opportunities with Targeted Combinations
Numerous well-established examples exist in oncology

where combination regimens of anticancer drugs are more
successful than single agents. An example of a successful
antitumor combination is 5-fluorouracil, leucovorin, and
irinotecan (FOLFIRI) with bevacizumab in colorectal cancer.6 In addition, it is well known that diseases such as
Hodgkin lymphoma require multiagent chemotherapy to
achieve cure. This and our increased understanding of
disease biology indicate that in order to increase the odds of
successful anticancer strategies, specific targeted agents
should be combined with either other selective inhibitors or
chemotherapy.
The development of combination regimens involving targeted therapies versus chemotherapy combinations may
differ in that chemotherapy was traditionally studied in
combination only once a single agent was approved and
demonstrated clinical activity benefit. Conversely, targeted
drugs may and probably should be tested in hypothesistesting rational combinations earlier in the drug development journey if tolerability data are available, assuming
there is a strong biologic rationale to combine two drugs.
Even if these agents demonstrate modest anticancer activity
as single agents, with substantial pharmacodynamic inhibition of target and pathway, satisfactory pharmacokinetics
and tolerability, combination studies may be warranted. We
envision that earlier testing of such combinations in early-
670
through the inhibition of escape signaling pathways and

slowing the development of newly resistant tumor cells. Combination regimes also have the potential of enhancing target
inhibition through synergistic antitumor effects and minimizing drug-related toxicities to patients. However, numerous
challenges to developing these combinations exist. This review will focus on the opportunities and pitfalls of developing
novel targeted drug combinations, with a particular focus on
early-phase drug development, where the greatest challenges
exist, analyzing key points for the design and development of
clinical trials for combinations of targeted agents.
phase trials to be increasingly desirable and an important

opportunity to accelerate drug development.
Before the clinical testing of a combination regimen, it is
critical to understand the underlying biologic rationale and
to have some preclinical evidence demonstrating synergistic
or additive antitumor effects for the anticancer drugs being
evaluated. Because targeted agents have been developed
as potent inhibitors of select biologic targets, combination
targeted therapy strategies can also seize the opportunity to
selectively titrate effects on multiple targets using potent
inhibitors. In this manner, combination dosages and schedules can be altered to optimize antitumor efficacy and
minimize the toxicity profile (Table 1).
Examples of Strategies to Combine Targeted Agents
Superinhibition of a single target. This strategy maximizes the inhibition of a single target, enhancing the on
target effect. Examples include strategies of dual HER2
blockade (e.g., trastuzumab with lapatinib or pertuzumab in
combination). Several studies have demonstrated these
strategies to be tolerable, with evidence of clinical benefit,
in various stages of breast cancer.7,8 This strategy may be
best utilized in tumors that are highly dependent on a single
gene or target, without significant resistance mechanisms
that bypass the target. This strategy also has the disadvantage of possibly enhancing overlapping drug-related
toxicities.
Inhibition of several targets within a single pathway. This
strategy takes an approach to inhibiting multiple substrates
to maximize pathway inhibition while minimizing resistance mechanisms that may occur because of regulatory
feedback loops. For example, there is evidence of upregulation of AKT phosphorylation after mTOR inhibition in several cell lines and human tumor types,9 suggesting a role for
From the Drug Development Unit, Royal Marsden NHS Foundation Trust, The Institute
of Cancer Research, Downs Road, Sutton, Surrey, United Kingdom.
Address reprint requests to Professor Johann S. de Bono, MB ChB, MSc, PhD, Division
of Cancer Therapeutics and Division of Clinical Sciences, The Institute of Cancer Research/
Royal Marsden NHS Foundation Trust, Downs Road, Sutton, Surrey SM2 5PT, United
Kingdom; email: johann.de-bono@icr.ac.uk
1092-9118/10/1-10
COMBINING TARGETED AGENTS IN CLINICAL TRIALS
the rational combination of mTOR inhibitors with either AKT

or PIK3CA inhibitors or the development of TOR kinase
inhibitors. This strategy may be best utilized in tumors driven
mainly by one pathway but which may fail if alternative
pathways are prominent as a mechanism of resistance. This
strategy also has the disadvantage of possible overlapping
drug-related toxicities via inhibiting the same pathway.
Targeting parallel or unrelated pathways. This strategy
maximizes the likelihood of avoiding resistance mechanisms, especially those caused by the upregulation of parallel pathways, and also minimizes overlapping toxicities. The
concomitant inhibition of two pathways may also result in
selective tumor cell cytotoxicity, whereas each agent alone
may not be active as a single agent.
An example of this approach is the dual targeting of the
PI3K and Ras/Raf/Mek/ERK pathways. Deregulation of the
Ras/Raf/Mek/ERK pathway has been identified as a determinant of resistance to PI3K-Akt inhibitors.10 There is also
strong preclinical evidence that ERK-independent mechanisms of resistance to B-RAF and MEK inhibitors are
mediated by the upregulation of the PIK3CA/Akt/mTOR
pathway.11 In murine models, double inhibition of PIK3CA
and MEK shows synergistic antitumor activity for K-RAS
mutant tumors,12 and PI3K-Akt activation leads to MEKinhibitor resistance in K-RAS mutants, which is reversible
when blocking both pathways.13 Moreover, mutations in
both pathways tend to coexist.14
Clinical trials are currently being conducted to evaluate
this strategy of targeting parallel pathways. Recently, a
first-in-human trial of the pan-AKT inhibitor MK-2206 has
been reported; although strong pharmacodynamics evidence
of AKT signaling blockade was demonstrated, minimal evidence of single-agent antitumor activity was observed.15 A
combination trial of MK-2206 with the MEK inhibitor
AZD6244 is currently being evaluated (NCT01021748) and
is demonstrating antitumor activity in KRAS-mutant cancers. Other similar early-phase trials of combinations based
on this approach include drugs targeting MEK and mTOR
(NCT01378377), or MEK and PIK3CA/mTOR inhibition
KEY POINTS
The biology of cancer cells involves complex signaling

networks, so anticancer strategies may not be successful if blocking single targets.
A combination of targeted agents will be necessary to
maximize antitumor activity, counteracting primary
and secondary resistance.
The rationale for evaluating combinations of drugs
must rely on biology-driven hypothesis and smart
screening strategies.
Different strategies include using two or more drugs
against the same target, targeting different points
upstream and downstream of one pathway, or targeting several pathways simultaneously on which the
cancer cell is dependent.
Design of clinical trials with combinations of targeted
agents should consider how to deal with the challenges pertaining to such studies.
Table 1. Opportunities and Pitfalls of Combining

Targeted Agents
Opportunities
1. Validate novel biologic hypotheses

2. Synergize antitumor effect without
synergizing toxicity: increasing the
therapeutic window
3. Further develop single agents that do
not have activity as monotherapy:
synthetic lethality
4. Counteract primary and secondary
resistance
5. Develop novel indications for existing
and/or approved drugs
Pitfalls
1. Unreliable preclinical models

2. Optimal selection of drugs and
targets to combine
3. Optimal sequence and dosage of
the combination
4. Risk of overlapping toxicities
5. Lack of standardized design for
phase I/II trials for combination
targeted therapies
6. Competing interests of different
researchers, corporations, and/or
institutions to combine therapies
(NCT01390818). These data allow us to envision the testing

of biology-driven hypotheses with multiple combinations;
these may eventually require more drugs for optimal blockade of cancer growth. Combinations for the treatment of
advanced prostate cancer that we propose include abiraterone acetate and MDV3100; either of these agents with
PI3K/AKT/TOR inhibitors; or possibly triple therapy with
abiraterone, MDV3100, and a PI3K/TOR inhibitor.
Choosing the Right Combination Therapy
When combining two (or more) targeted therapies for the

first time, each must be tested at various doses and dosing
schedules in a phase I study. However, the process of dose
finding, or defining the recommended phase II trial dose,
can take 1 year or longer to complete depending on the
complexity of the trial and compatibility of the agents, all at
significant cost. Therefore, it is critical to utilize strategies
that identify the best therapeutic combinations, and furthermore identify targets that cancer cells are reliant on so that
when blocked, there is a lethal or at least cytostatic effect.
This type of approach is based on concepts of oncogene
addition, nononcogene addiction, and synthetic lethality
that attack the hallmarks of cancer.16
Adopting systems-based approaches may help to address
this complexity, such as utilizing information from deep
transcriptomic, genomic, proteomic, and/or metabolic analyses.17 Although this personalized approach is attractive,
it may often yield information on targets for which there is
no active drug, for which there are only passenger mutations rather than drivers of oncogenesis, or for genes for
which we have incomplete functional information. It has
been suggested that systems-based approaches can be used
to select targeted combinations17; however, further development and application of bioinformatics is clearly necessary
to process the high amounts of data derived from massive
screening strategies.
Another strategy is to identify potential drug combinations using RNA interference (RNAi) to simulate models of
pharmacologic inhibition. In this method, wild-type cell lines
or cell lines with induced loss of function for a particular
target are exposed to RNAi libraries with or without exposure to a particular drug to look for antitumoral synergism
or drug sensitization. These RNAi screens can therefore
select potential combination of targets to be tested further18,19 and ideally identify potential synthetic lethality
effects when inhibiting two a priori nonlethal targets. How-
671
MATEO ET AL
Fig. 1. Schematic demonstrating the increased complexity of dose and schedule selection when combining targeted agents. Key factors that
must be considered include the pharmacokinetic (PK) and pharmacodynamic (PD) properties of each single agent, potential PK and PD
interactions between the agents, how exposure to the combination may result in antitumor synergism, and what toxicities may overlap.
ever, current limitations of this approach include incomplete

specificity of the RNAis used, the appearance of several
phenotypes after the inhibition by RNAi,20 and the issue
that subsequent preclinical models are not reliable in predicting the ultimate sensitivity of tumor tissue to combination therapy.21
Ultimately, however, combinations of targeted agents
must be selected for evaluation based on a strong biologic
rationale. Preclinical evidence for targeted combinations is
often lacking, but can be critical to take a proper combination forward. For example, Wam and colleagues observed an
upregulation of AKT phosphorylation in rhabdomyosarcoma
cell lines and xenografts that was dependent on insulin
growth factor-1 receptor (IGF-1R) activation.22 This codependence suggested a potential for combination targeted
therapy using mTOR and IGF-1R blockade. Results reported
from an early-phase trial combining ridaforolimus and an
anti-IGF-1R antibody showed signs of promising clinical
benefit.23 Nonetheless, initiatives of sequential biopsies
should be integral to clinical research for a better understanding of molecular mechanism of resistance to targeted
agents17 to better define whether combination treatments
can address these issues.
672
Design of Early-Phase Clinical Trials for Combinations
The design of trials evaluating combinations of targeted

agents can be significantly more complex than trials of
single targeted agents. Successful combination strategies
require not only validation of the biologic rationale but also
determination of the optimal dosing and scheduling. Combinations of drugs may also be significantly affected by
pharmacokinetic and pharmacodynamic interactions, and
phase I or II trials may be designed to evaluate multiple
schedules of interest to determine not only tolerable regimens but feasible dosing regimens (Fig. 1).
Early valid pharmacodynamic biomarkers should be required for proof of mechanism for the combination, ideally
from both tumor and normal tissue biopsies when feasible.
Design of early-phase clinical trials should be flexible
enough to facilitate real time reverse translation between
clinical and laboratory research.
Dose Escalation in Combination Trials
In single-agent phase I trials, dose escalation usually

occurs by standard rules such as the 33 design. However,
in combination trials, the maximally tolerated dose (MTD) of
each drug is likely to be well defined; therefore, evaluation of
COMBINING TARGETED AGENTS IN CLINICAL TRIALS
the doses for combination can occur in many permutations.

Dose-escalation rules for combination therapy must take into
consideration preclinical data for the single agent and the
combination. Depending on the stage of development of each
agent, their mechanism of action and the potential overlapping toxicities, different scenarios are envisioned: dose escalation is probably best pursued sequentially, increasing the
dose of one compound in every step; dose escalation of both
drugs at the same time is not recommended; and to optimize
antitumor cell kill there is also merit in fixing the drug
targeting the driver pathway (e.g., BRAF/MEK inhibitor for
BRAF mutation in melanoma) at the highest tolerable dose
while dose escalating the combination drug (e.g., PI3K/AKT/
TOR inhibitor). When combining two drugs with no pharmacokinetic interactions and nonoverlapping mechanisms
of action and toxicity, and when we already have extensive
experience regarding the tolerability profile of each drug,
fewer dose-escalation steps may be necessary to accelerate
combination development, as recently reported in a trial
exploring the combination of sunitinib and everolimus in
metastatic renal cell carcinoma (mRCC).24
Novel combinatory designs have also been suggested as
potential improvements over algorithmic approaches. These
strategies try to pursue fewer levels of dose escalation,
thereby reducing the rate of patients exposed to biologically
inactive doses without exposing them to more toxicities, in
contrast to more classical drug development designs.25
Novel models of multiple parallel cohorts permit escalating
two or more drugs in a sequential/parallel manner based on
mechanism-related toxicities.26 Bayesian model-based designs are based on continual reassessment of risk of toxicities and have been applied to dose escalation of combination
targeted agents.27
Flexible dose escalation and de-escalation rules and intense knowledge about the mechanisms of toxicity of each
drug in the combination are required to maximize outcomes
without compromising a potentially or already known active
dose of a single agent. Despite the availability of these novel
trial designs, they are not commonly used.
Furthermore, although the rationale for the combination
should consider which pathway is the main driver of the
disease and therefore which drug is administered at a dose
level closer to its single-agent MTD, investigators may have
to be prepared to decrease the dose of the main drug if the
biologic rationale of a trial suggests a likelihood that synergism and full-dose combination is not tolerable.
Biomarker-Guided Selection of Patients
Although the selection of patients in early-phase trials

based on biomarkers may make trial recruitment more difficult, combinations of targeted agents based on hypotheses of
synergistic biologic effect will ultimately require patient selection using biomarkers that identify populations that may
benefit. Unfortunately, often the development of valid biomarkers that predict for benefit are often slow to develop
in both clinical and preclinical models, making patient selection in early-phase trials difficult. The development of predictive biomarkers for patient selection is crucial to the success of
targeted agents as well as combinations of targeted agents.
Assessment of Tolerability in Combination Studies
Since chronic dosing is usually indicated for these drug

combinations, delayed and cumulative toxicities should be
carefully considered when selecting the recommended dose

for further development. As an example, a phase I trial of
the combination of bevacizumab and everolimus28 escalated
doses up to 10 mg/kg fortnightly and 10 mg/d po respectively, without observing dose-limiting toxicities during the
first cycles of treatment. In a phase II trial in mRCC,29 up to
a quarter of the patients had grade 3 or 4 proteinuria with
the combination, while grade 3 or 4 proteinuria with bevacizumab alone at the same dose in mRCC was below 8%.30
Trial designs that integrate information from late onset
toxicities for deciding regarding dose selection may be warranted.31 In addition, clinical trials should aim to assess
mechanistic data to define causal relationships with drug
and observed toxicities to make better informed decisions
regarding choices of combination targeted agents. Uncertainty regarding the mechanisms behind targeted agent
toxicity and off-target drug effects may negatively affect
promising combinations.
Regulatory Concerns and Legal Barriers
The evaluation of different combinations of several targeted agents from multiple companies may sometimes be
challenging because of competing interests, and may be
further complicated if multiple sponsors are not fully committed to the trial. A clear example may be combining an
established drug or chemotherapy regimen with a novel
agent, as the outcome of the trial may not be equally
profitable for both sponsors. Apart from intellectual property issues and the requirement of financial support, the
risk of novel toxicities and adverse outcomes with concomitant administration might be considered a risky commercial
decision for sponsors whose antitumor compound is already
at a more advanced stage of development. Academic institutions, patient lobbies, and regulatory authorities may
need to assume a championing role for some combinations
with a strong biologic rationale that are perhaps less commercially attractive.
Conclusion
The advent of targeted therapies has shifted the emphasis

in drug development from cytotoxic chemotherapy combinations to selective and potent molecular therapeutics. Such
specific inhibitors may not result in significant or durable
antitumor activity as single agents, and should therefore be
given in combination with one or more drugs to molecularly
defined populations of patients. We should now be focusing
more efforts on combinatorial drug development strategies,
perhaps even from the outset of first-in-human trials to
maximize the likelihood of benefit to patients dying of
advanced cancer. Such an approach may be able to further
accelerate the delivery of anticancer treatment to benefit
and improve the lives of our patients with cancer.
ACKNOWLEDGMENTS
The Drug Development Unit of the Royal Marsden NHS
Foundation Trust and The Institute of Cancer Research is
supported in part by a program grant from Cancer Research
United Kingdom. Support was also provided by the Experimental Cancer Medicine Centre (to The Institute of Cancer
Research) and the National Institute for Health Research Biomedical Research Centre (jointly to the Royal Marsden NHS
Foundation Trust and The Institute of Cancer Research).
673
MATEO ET AL

Employment or
Leadership
Positions
Author
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Joaquin Mateo*
Michael Ong*
Timothy A. Yap*
Johann S. de Bono
The Institute of
Cancer Research
(L)
AstraZeneca;
Boehringer
Ingelheim;
Genentech;
Johnson &
Johnson
AstraZeneca;
Boehringer
Ingelheim;
Genentech;
GlaxoSmithKline;
Johnson &
Johnson; Pfizer
AstraZeneca;
Immunicon
REFERENCES
1. Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety of
imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl
J Med. 2002;347:472-480.
Clin Cancer Res. 2005;11:4182-4190.
3. Yang J, Ikezoe T, Nishioka C, et al. Long-term exposure of gastrointestinal stromal tumor cells to sunitinib induces epigenetic silencing of the
PTEN gene. Int J Cancer. 2012;130:959-966.
4. Nilsson B, Sjolund K, Kindblom L-G, et al. Adjuvant imatinib treatment
improves recurrence-free survival in patients with high-risk gastrointestinal
stromal tumors (GIST). Br J Cancer. 2007;96:1656-1658.
5. Dematteo RP, Ballman KV, Antonescu CR, et al. Adjuvant imatinib
mesylate after resection of localized, primary gastrointestinal stromal tumor:
a randomised, double-blind, placebo-controlled trial. Lancet. 2009;373:10972104.
6. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus
irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer.
N Engl J Med. 2004;350:2335-2342.
7. Nahta R, Hung M-C, Esteva FJ. The HER-2-targeting antibodies trastuzumab and pertuzumab synergistically inhibit the survival of breast cancer
cells. Cancer Res. 2004;64:2343-2346.
8. Baselga J, Bradbury I, Eidtmann H, et al. Lapatinib with trastuzumab
for HER2-positive early breast cancer (NeoALTTO): a randomized, openlabel, multicenter, phase 3 trial. Lancet. 2012;379:633-640. Epub 2012 Jan 17.
Erratum in: Lancet. 2012;379:616.
9. Tabernero J, Rojo F, Calvo E, et al. Dose- and schedule-dependent
inhibition of the mammalian target of rapamycin pathway with everolimus: a
phase I tumor pharmacodynamic study in patients with advanced solid
tumors. J Clin Oncol. 2008;26:1603-1610.
10. Yu K, Toral-Barza L, Shi C, et al. Response and determinants of cancer
cell susceptibility to PI3K inhibitors: combined targeting of PI3K and Mek1 as
an effective anticancer strategy. Cancer Biol Ther. 2008;7:307-315.
12. Engelman JA, Chen L, Tan X, et al. Effective use of PI3K and MEK
inhibitors to treat mutant Kras G12D and PIK3CA H1047R murine lung
cancers. Nature Medicine. 2008;14:1351-1356.
13. Wee S, Jagani Z, Xiang KX, et al. PI3K pathway activation mediates
resistance to MEK inhibitors in KRAS mutant cancers. Cancer Res. 2009;69:
4286-4293.
14. Janku F, Lee JJ, Tsimberidou AM, et al. PIK3CA mutations frequently
coexist with RAS and BRAF mutations in patients with advanced cancers.
PLoS One. 2011;6e:22769.
15. Yap TA, Yan L, Patnaik A, et al. First-in-man clinical trial of the oral
674
pan-AKT inhibitor MK-2206 in patients with advanced solid tumors. J Clin

Oncol. 2011;29:4688-4695.
16. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000;100:5770.
17. Sequist LV, Waltman BA, Dias-Santagata D, et al. Genotypic and
histological evolution of lung cancers acquiring resistance to EGFR inhibitors.
Sci Transl Med. 2011;3:75ra26.
18. Whitehurst AW, Bodemann BO, Cardenas J, et al. Synthetic lethal
screen identification of chemosensitizer loci in cancer cells. Nature. 2007;
446:815-819.
19. OGrady M, Raha D, Hanson BJ, et al. Combining RNA interference
and kinase inhibitors against cell signaling components involved in cancer.
BMC Cancer. 2005;5:125.
20. Iorns E, Lord CJ, Turner N, et al. Utilizing RNA interference to
enhance cancer drug discovery. Nature Rev Drug Discov. 2007;6:556-568.
21. Becher OJ, Holland EC. Genetically engineered models have advantages over xenografts for preclinical studies. Cancer Res. 2006;66:3355-3358;
discussion 3358-3359.
22. Wan X, Harkavy B, Shen N, et al. Rapamycin induces feedback
activation of Akt signaling through an IGF-1R-dependent mechanism. Oncogene. 2007;26:1932-1940.
23. Di Cosimo S, Bendell J, Cervantes-Ruiperez A, et al. A phase I study of
the oral mTOR inhibitor ridaforolimus (RIDA) in combination with the
IGF-1R antibody dalotozumab (DALO) in patients (pts) with advanced solid
tumors. J Clin Oncol. 2010;28:15s (suppl; abstr 3008).
24. Molina AM, Feldman DR, Voss RJ, et al. Phase 1 trial of everolimus
plus sunitinib in patients with metastatic renal cell carcinoma. Cancer. Epub
2011 Sept 6.
25. Braun TM, Alonzo TA. Beyond the 33 method: Expanded algorithms
for dose-escalation in Phase I oncology trials of two agents. Clin Trials.
2011;8:247-259.
26. Rodon J, Perez J, Kurzrock R. Combining targeted therapies: practical
issues to consider at the bench and bedside. Oncologist. 2010;15:37-50.
27. Yin G, Li Y, Ji Y. Bayesian dose-finding in phase I/II clinical trials
using toxicity and efficacy odds ratios. Biometrics. 2006;62:777-784.
28. Zafar Y, Bendell J, Lager J, et al. Preliminary results of a phase I study
of bevacizumab (BV) in combination with everolimus (E) in patients with
advanced solid tumors. J Clin Oncol. 2006;24:145s (suppl; abstr 3097).
29. Hainsworth JD, Spigel DR, Burris HA, et al. Phase II trial of bevacizumab and everolimus in patients with advanced renal cell carcinoma. J Clin
Oncol. 2010;28:2131-2136.
30. Yang JC, Haworth L, Sherry RM, et al. A randomized trial of bevacizumab, an anti-vascular endothelial growth factor antibody, for metastatic
renal cancer. N Engl J Med. 2003;349:427-434.
31. Polley, MY. Practical modifications to the time-to-event continual
reassessment method for phase I cancer trials with fast patient accrual and
late-onset toxicities. Stat Med. 2011;30:2130-2143.
Combination Therapies Building on the

E f fi c a c y o f C T L A 4 a n d B R A F I n h i b i t o r s f o r
Metastatic Melanoma
By Antoni Ribas, MD
Overview: The demonstration of improved survival with the

anti-CTLA4 antibody ipilimumab and the BRAF inhibitor vemurafenib in patients with metastatic melanoma is arguably the
most significant advance in the treatment for these patients in
the last 30 years. However, the majority of patients will either
not experience response, or will experience response and
then progression, when receiving these therapies, so additional treatment options are required. Since these agents
URING 2011, two new single-agent therapies were

approved by the U.S. Food and Drug Administration
(FDA) for the treatment of patients with metastatic melanoma. The antitumor activity of these newly approved
agents, ipilimumab (formerly MDX010) and vemurafenib
(formerly PLX4032/RG7204), is based on improved means to
stimulate immune responses to melanoma and to block
oncogenic signaling in cancer cells, respectively. Ipilimumab
was approved on the basis of two randomized clinical trials,
one demonstrating improvement in survival over a peptide
vaccine in second-line or later therapy,1 and another in
combination with the chemotherapy agent dacarbazine compared with dacarbazine and placebo.2 Vemurafenib was
approved based on the results of a randomized clinical trial
demonstrating an early benefit in overall survival compared
with dacarbazine in patients with BRAFV600 mutant metastatic melanoma.3 Despite the unquestionable benefits of
these two therapies, the great majority of patients with
metastatic melanoma continue to require improved treatments because the majority will not experience response to
ipilimumab, and the majority with BRAFV600 mutations will
experience response to vemurafenib but will have progression within a matter of months.
These recent advances are based on targeted interventions that build on a detailed knowledge of what is being
modulated, either the CTLA4 coinhibitory molecule for ipilimumab or the BRAF driver oncogene for vemurafenib.
Their antitumor activity is based on a refined understanding
of their mechanisms of action, allowing the rational advancement of therapies building on these two agents. Other
single-agent therapies that have mechanisms of action of
immune modulatory effects, either inhibiting other negative
immune regulators such as PD1, or activating immune
coactivators such as CD40, CD137 or OX40, are being
developed in addition to the already approved immune activating cytokines interferon-alpha (IFN-a) and interleukin-2
(IL-2). Furthermore, other targeted driver oncogene inhibitors are in clinical testing blocking BRAF (such as dabrafenib, formerly GSK2118436), blocking ckit in a subset of
ckit mutant melanomas (primarily acral and mucosal melanomas) or the use of downstream MEK inhibitors for tumors
with oncogenic mitogen-activated protein kinase (MAPK)
signaling (Fig. 1).
All of these agents with proven antitumor activity in
melanoma are pointing the way of future research that
addresses current limitations while building on their benefits. Being able to combine several of these active agents in
have been developed with a refined understanding of their

mechanism of action and mechanisms leading to resistance
are being elucidated, then combination therapies building
on these single-agent therapies can be designed rationally.
Such combinations are being tested both preclinically and in
the clinic, and provide a strong promise to improve on the
current treatment approaches for patients with metastatic
melanoma.
a rational manner using in-depth understanding of the

underlying biology is likely to continue to provide advances
to the treatment options for advanced melanoma.
Combination of Immunotherapies
The immune system has multiple control mechanisms

aimed at preventing an overt immune activation to selfantigens, which would lead to high frequency of autoimmune diseases. Therefore, an activated immune response
to cancer requires fine tuning to balance between killing
cellular targets but not inducing autoimmunity. With the
improved understanding of the regulation of immune activation and the existence of multiple inhibitory immune
pathways, combinations of immune activators, and blockers
of immune inhibitors, are being tested in preclinical models
and in the clinic. Among them are combinations being tested
in the clinic building on anti-CTLA4 antibodies with the
addition of immune activating cytokines such as IFN-a4 or
IL-2,5 with dendritic cell vaccines,6 and also earlier studies
combining with other immune modulating antibodies such
as anti-PD1.
Combination of three or four immune modulating antibodies has been demonstrated to provide antitumor activity
over single-agent or double combinations in mouse models.7
For example, high antitumor activity has been demonstrated combining anti-CD40, anti-CD137, and anti-TRAIL,
with further enhancement with anti-CTLA4 antibodies.7,8
The testing of triplet immune-modulating monoclonal antibody therapy in humans is limited by the availability of
these antibodies and the unclear path for their combined
clinical development. Restrictions are imposed by the regulatory process, which requires demonstrating the individual
contribution to the antitumoral benefit while avoiding increased toxicities, as well as the anticipated high expense of
each one of these monoclonal antibodies. However, if the
benefit is as predicted in the mouse models, then a higher
rate of durable tumor responses than can be achieved
From the Department of Medicine, Division of Hematology-Oncology, Jonsson Comprehensive Cancer Center at the University of California, Los Angeles, Los Angeles, CA.
Address reprint requests to Antoni Ribas, MD, Department of Medicine, Division of
Hematology-Oncology, Jonsson Comprehensive Cancer Center at the University of California, Los Angeles, 11-934 Factor Building, 10833 Le Conte Avenue, Los Angeles, CA
90095-1782; email: aribas@mednet.ucla.edu.
1092-9118/10/1-10
675
ANTONI RIBAS
Fig. 1. Schematic of the mechanism of

action of new treatments demonstrating patient benefit in advanced melanoma. The
majority of melanomas demonstrate a constitutively active mitogen-activated protein
kinase (MAPK) pathway leading to uncontrolled cell proliferation and avoidance of
apoptosis, which is to the result of the presence of mutually exclusive activating mutations in the receptor tyrosine kinase cKit, in
NRAS or BRAF (red stars). Inhibitors of oncogenic driver mutations blocking cKit or
BRAF result in high response rates, and this
oncogenic pathway can also be blocked
with MEK inhibitors. Melanoma can also be
treated by activating an antitumor immune
response, either by turning it on against
cancer cells by the immune activating cytokines interleukin-2 (IL2) or interferon, or the
immune activating antibodies to CD40,
CD137 or OX40, or by administering antibodies blocking negative costimulatory signaling through CTLA4 or PD1.
otherwise would be compelling to develop clinically and

cost-effective by avoiding the need for further therapies.
Combination of Targeted Therapies
The description of molecular mechanisms of the development of acquired resistance to BRAF inhibitors, and of
toxicities with these agents, allow the rational development
of combination targeted therapies for the treatment of advanced melanoma. Mechanisms of resistance are diverse
and can be categorized between the ones that reactivate the
MAPK pathway and the mechanisms that lead to a MAPK
pathway-independent signaling that substitutes for the
blocked driver oncogenic signal. MAPK reactivating mechanisms reported to date in patient-derived samples include
truncations in the BRAF protein resulting in increased
KEY POINTS
676
Single-agent ipilimumab and vemurafenib have been

approved for clinical use after demonstration of improvement in the survival of patients with metastatic
melanoma.
Combinations of immunotherapy approaches may
increase the frequency of tumor responses while
maintaining their long durability.
Combinations based on BRAF inhibitors can prevent
or treat acquired resistance.
The combination of a BRAF and a MEK inhibitor can
increase the response rate, extend the response duration and prevent the major skin toxicity of these
agents.
Combinations of immunotherapy and BRAF-targeted
therapy may allow maintaining the high response
rate of the targeted therapy and increase its durability with immunotherapy.
kinase activity,9 amplifications of the mutant BRAF gene,10

secondary mutations in NRAS or MEK,11,12 or overexpression of COT.13 The mechanisms leading to MAPK-redundant
pathway activation are induced by overexpression or overactivation of receptor tyrosine kinases such as the plateletderived growth factor receptor beta (PDGFRb)11,14 or the
insulin-like growth factor receptor 1 (IGF-1R),15 leading
oncogenic signaling through the phosphoinositide 3-kinase
(PI3K)/AKT pathway that can be pharmacologically blocked
in preclinical models with specific inhibitors of this
pathway.14,16-18
Adding an MEK inhibitor to a BRAF inhibitor would be
able to treat or prevent most of the MAPK-reactivation
resistance mechanisms, and such combinations are already
in advanced clinical testing. Early evidence suggests that
some patients experiencing progression with single-agent
BRAF inhibitors can have secondary responses when adding
an MEK inhibitor to continued therapy with a BRAF inhibitor.19 On the contrary, and as suggested by preclinical
modeling,17 sequential therapy stopping a BRAF inhibitor
and changing the therapy to a MEK inhibitor provides no
benefit.20 An even greater benefit may be derived from the
initiation of therapy combining a BRAF and an MEK inhibitor as the first oncogene-targeted therapy for BRAF mutant
melanoma, as has been done with the combination of dabrafenib and the MEK inhibitor GSK1120212.19,21 This combination, compared with single-agent BRAF inhibitors, may
result in higher antitumor activity by more profound oncogenic MAPK inhibition, potentially leading to more durable
responses by preventing the MAPK-dependent acquired
resistance mechanisms. An additional benefit of the combination of BRAF and MEK inhibitors is the ability of the
MEK inhibitors to block paradoxical MAPK activation leading to the development of the main grade 3 toxicity with the
use of single-agent BRAF inhibitors, which is the development of secondary cutaneous squamous cell carcinomas.22
Therefore, this combination would be a rare example of two
COMBINATIONS FOR MELANOMA
agents that, when combined, have increased antitumor

activity with decreased toxicities.
Despite the wealth of supportive preliminary evidence,14,16-18 studies combining a BRAF inhibitor with a
PI3K or AKT inhibitor to treat or prevent MAPKindependent resistance are not in the clinic. However, there
are several experiences combining MEK inhibitors with
PI3K inhibitors that may provide the benefit of blocking
both pathways. But this potential benefit has to be weighed
against the expected increase in toxicities because neither
agent would be directly blocking a mutated oncogene in this
setting.
Combination of Immunotherapies and
Targeted Therapies
BRAF inhibitors are not detrimental to human or murine

lymphocyte function.25,26 Furthermore, given the evidence
that BRAF inhibitors can induce a paradoxical activation of
the MAPK pathway in cells that have strong RAS/GTP
activity,27-29 which would be expected to also occur in
activated lymphocytes, it is certainly possible that at certain
doses and concentrations BRAF inhibitors may actually
potentiate the function of lymphocytes. Besides this potential beneficial effect of BRAF inhibitors directly in immune
cells, these agents may benefit in combination by increasing
expression or cross-presentation of tumor antigens to the
immune system, or by releasing signals that lead to an
adverse tumor environment limiting immune cell function.
Conclusion
Combinations of immunotherapy agents and oncogene

driver inhibitors have the potential to merge the benefits of
both types of agents for patients with advanced melanoma.
Targeted oncogene pathway inhibitors induce high rates of
tumor responses that tend to last months, whereas immunemodulating agents tend to have infrequent but highly durable tumor responses. The promise of combining them is to
maintain the high rate of targeted therapy responses and
make them more frequently durable with immunotherapies.23,24
Key to the development of this type of combination is
demonstrating that both approaches are not negatively
interacting, although they can have synergistic effects by
potentiating each other. In vitro studies have shown that
Advances in the treatment of metastatic melanoma are

based on an improved understanding of the immunobiology
and oncogenic signaling in this disease. Single-agent therapies with ipilimumab and vemurafenib have demonstrated
improvement in patient survival in randomized clinical
trials.1-3 However, the majority of patients still require
additional treatment options. Understanding the mechanisms of action and resistance to the new single-agent
therapies allows an opportunity to rationally design combination therapies that may provide additional benefits to
patients. Such combinations are being developed preclinically and in the clinic, and are poised to change the current
treatment approaches for patients with metastatic melanoma.
Author
Antoni Ribas
Employment or
Leadership
Positions
Consultant or
Advisory Role
Celgene;
Genentech;
Merck;
Millennium;
Roche
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
REFERENCES
ipilimumab in patients with metastatic melanoma. New Engl J Med. 2010;
363:711-723.
for previously untreated metastatic melanoma. New Engl J Med. 2011;364:
2517-2526.
2011;364:2507-2716.
4. Tarhini AA, Kirkwood JM. Tremelimumab and interferon combination
in melanoma. Semin Oncol. In press.
5. Maker AV, Phan GQ, Attia P, et al. Tumor regression and autoimmunity in patients treated with cytotoxic T lymphocyte-associated antigen
4 blockade and interleukin 2: a phase I/II study. Ann Surg Oncol. 2005;12:
1005-1016.
6. Ribas A, Comin-Anduix B, Chmielowski B, et al. Dendritic cell vaccination combined with CTLA4 blockade in patients with metastatic melanoma.
Clin Cancer Res. 2009;15:6267-6276.
7. Uno T, Takeda K, Kojima Y, et al. Eradication of established tumors in
mice by a combination antibody-based therapy. Nat Med. 2006;12:693-698.
8. Takeda K, Kojima Y, Uno T, et al. Combination therapy of established
tumors by antibodies targeting immune activating and suppressing molecules. J Immunol. 2010;184:5493-5501.
9. Poulikakos PI, Persaud Y, Janakiraman M, et al. RAF inhibitor resistance is mediated by dimerization of aberrantly spliced BRAF(V600E).
Nature. 2011;480:387-390.
10. Shi H, Moriceau G, Kong X, et al. Melanoma V600E BRAF amplification drives acquired resistance to vemurafenib. Nat Commun. In press.
973-977.
2011;29:3085-3096.
968-672.
14. Shi H, Kong X, Ribas A, et al. Combinatorial treatments that overcome
PDGFR{beta}-driven resistance of melanoma cells to V600EB-RAF inhibition.
Cancer Res. 2011;71:5067-5074.
16. Jiang CC, Lai F, Thorne RF, et al. MEK-independent survival of
B-RAFV600E melanoma cells selected for resistance to apoptosis induced by
the RAF inhibitor PLX4720. Clin Cancer Res. 2011;17:721-730.
677
ANTONI RIBAS
17. Atefi M, von Euw E, Attar N, et al. Reversing melanoma crossresistance to BRAF and MEK inhibitors by co-targeting the AKT/mTOR
pathway. PloS One. 2011;6:e28973.
18. Paraiso KH, Xiang Y, Rebecca VW, et al. PTEN loss confers BRAF
inhibitor resistance to melanoma cells through the suppression of BIM
expression. Cancer Res. 2011;71:2750-2760.
19. Flaherty K, Infante JR, Falchook GS, et al. Phase I/II study of BRAFi
GSK2118436 MEKi GSK1120212 in patients with BRAF mutant metastatic melanoma who progressed on a prior BRAFi. Pigment Cell Melanoma
Res. 2011;24:LBA1.
20. Kim KB, Lewis KD, Pavlick AC, et al. A phase II study of the
MEK1/MEK2 inhibitor GSK1120212 in metastatic BRAF-V600E or K mutant
cutaneous melanoma patients previously treated with or without a BRAF
inhibitor. Pigment Cell Melanoma Res. 2011;24:1021.
21. Infante JR, Falchook GS, Lawrence DP, et al. Phase I/II study to assess
safety, pharmacokinetics, and efficacy of the oral MEK 1/2 inhibitor
GSK1120212 (GSK212) dosed in combination with the oral BRAF inhibitor
GSK2118436 (GSK436). J Clin Oncol 2011;29 (suppl; abstr CRA8503).
22. Su F, Viros A, Milagre C, et al. RAS mutations in cutaneous squamouscell carcinomas in patients treated with BRAF inhibitors. New Engl J Med.
2012;366:207-215.
678
23. Begley J, Ribas A. Targeted therapies to improve tumor immunotherapy. Clin Cancer Res. 2008;14:4385-4391.
24. Ribas A, Flaherty KT. BRAF targeted therapy changes the treatment
paradigm in melanoma. Nat Rev Clin Oncol. 2011;8:426-433.
25. Boni A, Cogdill AP, Dang P, et al. Selective BRAFV600E inhibition
enhances T-cell recognition of melanoma without affecting lymphocyte function. Cancer Res. 2010;70:5213-5219.
26. Comin-Anduix B, Chodon T, Sazegar H, et al. The oncogenic BRAF
kinase inhibitor PLX4032/RG7204 does not affect the viability or function of
human lymphocytes across a wide range of concentrations. Clin Cancer Res.
2010;16:6040-6048.
27. Heidorn SJ, Milagre C, Whittaker S, et al. Kinase-dead BRAF and
oncogenic RAS cooperate to drive tumor progression through CRAF. Cell.
2010;140:209-221.
28. Halaban R, Zhang W, Bacchiocchi A, et al. PLX4032, a selective
BRAF(V600E) kinase inhibitor, activates the ERK pathway and enhances cell
migration and proliferation of BRAF melanoma cells. Pigment Cell Melanoma
Res. 2010;23:190-200.
29. Poulikakos PI, Zhang C, Bollag G, et al. RAF inhibitors transactivate
RAF dimers and ERK signalling in cells with wild-type BRAF. Nature.
2010;464:427-430.
MECHANISMS OF RESISTANCE TO TARGETED

ANTICANCER AGENTS
CHAIR
Levi A. Garraway, MD, PhD
Boston, MA
SPEAKERS
Alberto Bardelli, PhD
Istituto Ricerca Cura Cancro and FIRC
Candiolo, Italy
Neil P. Shah, MD, PhD
San Francisco, CA
Alice Tsang Shaw, MD, PhD
Boston, MA
Mechanisms of Resistance to MitogenActivated Protein Kinase Pathway Inhibition

in BRAF-Mutant Melanoma
By Eva M. Goetz, PhD, and Levi A. Garraway, MD, PhD
Overview: Anticancer drug resistance remains a crucial impediment to the care of many patients with cancer. Although
the exact mechanisms of resistance may differ for each
therapy, common mechanisms of resistance predominate,
including drug inactivation or modification, mutation of the
target protein, reduced drug accumulation, or bypass of target
inhibition. With the discovery and use of targeted therapies
(such as small-molecule kinase inhibitors), resistance has
received renewed attention especially in light of the dramatic responses that may emerge from such therapeutics in
particular genetic or molecular contexts. Recently, the
mitogen-activated protein kinase (MAPK) pathway has become exemplary in this regard, since it is activated in many
different cancers. Drugs targeting RAF and MAPK kinase
(MEK) are currently in clinical trials for the treatment of
several types of cancer. Vemurafenib, a selective RAF kinase
ELANOMA IS among the most common cancers, and

with early detection and surgical resection, the 5-year
survival rate for melanoma is 98%.1 Unfortunately, the
5-year survival rate falls to 15% for individuals with metastatic disease. Even considering the recent spectacular therapeutic successes, relatively few treatment options exist for
metastatic melanoma. Those in current use include chemotherapy (dacarbazine), immunotherapy (high-dose interferon or ipilimumab), and the newly US Food and Drug
Administration (FDA)approved targeted therapy vemurafenib. Therefore, new, targeted therapies are needed to
improve these dire statistics.
Approximately 40% to 60% of cutaneous melanomas contain a V600E mutation in BRAF that results in constitutive
MAPK signaling, independent of RAS (Fig. 1).2-4 These
melanomas are dependent on MAPK, therefore, a series of
small molecule inhibitors targeting mutant BRAF have been
developed and remain a topic of intense clinical investigation. Vemurafenib (PLX4032) is an FDA-approved inhibitor
of BRAF(V600E), and although patients with BRAF mutation have enjoyed a clinical benefit with vemurafenib treatment, the fraction of patients who achieve a Response
Evaluation Criteria in Solid Tumors (RECIST) partial response is 48%.5 Furthermore, the average duration of response is only 7 months.6 Thus, the magnitude and duration
of clinical response is limited, and the development of
progressive disease after several months to approximately 1
year of treatment is nearly universal. Since mutant BRAF
melanoma is dependent on MAPK kinase/extracellular
signal-regulated kinase (MEK/ERK) signaling, MEK inhibitors were predicted to be useful for treatment of metastatic
melanoma. Early data from phase II clinical trials of MEK
inhibitors were disappointing, however, the addition of MEK
inhibitors to RAF inhibitors provides a promising investigational avenue, as described below.7 Clearly, overcoming
resistance to vemurafenib may enable additional clinical
benefits to patients with BRAF(V600E) melanoma.
The development of resistance to targeted anticancer
therapy is typical in solid tumors. Several overarching
680
inhibitor recently approved for the treatment of BRAF(V600E)

melanoma, shows strong efficacy initially; however, the
development of resistance is nearly ubiquitous. In vitro testing and analysis of patient samples have uncovered several
mechanisms of resistance to RAF inhibition. Surprisingly,
mutations in the drug-binding pocket have not thus far been
observed; however, other alterations at the level of RAF, as
well as downstream activation of MEK and bypass of MEK/
extracellular signal-regulated kinase (ERK) signaling altogether, confer resistance to vemurafenib. Looking forward,
combined RAF and MEK inhibitor treatments may improve
efficacyyet we must anticipate mechanisms of resistance
to this combination as well. Therefore, understanding and/or
determining the mechanism of resistance are paramount to
effective cancer treatment.
mechanisms have been described for many targeted therapies, including erlotinib and gefitinib epidermal growth
factor receptor (EGFR) inhibitors in nonsmall cell lung
cancerand imatinib, which is used for BCR-ABL inhibition
in leukemia and KIT blockade in gastrointestinal stromal
tumors.8-10 In studying resistance mechanisms, the overarching goal is the development and implementation of
therapeutic combinations that can anticipate and thwart
this process at the onset of treatment, thereby increasing the
magnitude and duration of clinical benefit. Taken to its
logical conclusion, such understanding could pave the way to
durable therapeutic control of cancers driven by druggable
oncoprotein dysregulation.
Mechanistic Categories of Resistance to RAF Inhibition
Since resistance is so common with targeted therapies,

studies of resistance to RAF inhibitors have been at the
forefront of translational research in this area, even during
preclinical development (Table 1). Given the importance of
MAPK signaling in BRAF-mutant melanoma, it stands to
reason that any resistance mechanism leading to sustained
or reactivated MEK/ERK signaling could in principle contribute to therapeutic resistance.
BRAF Bypass
In other genetically defined tumor subtypes, resistance

has been achieved by the acquisition of point mutations
within the oncoprotein target itself, such as the T315I
mutation in BCR-ABL resulting in resistance to imatinib,
nilotinib, and dasatinib and the T790M mutation in EGFR,
From the Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA;
The Broad Institute of Harvard and MIT, Cambridge, MA.
Address reprint requests to Levi A. Garraway, MD, PhD, Broad Institute of MIT
and Harvard, Dana Building, Room 1542, 44 Binney St., Boston, MA 02115; email:
levi_garraway@dfci.harvard.edu.
1092-9118/10/1-10
RESISTANCE MECHANISMS TO MAPK INHIBITION

Fig 1. Mechanisms of resistance to vemurafenib. BRAF(V600E) mutant tumors have
constitutive activation of MEK and ERK (MAPK)
signaling, which is inhibited by vemurafenib.
Activating mutations (*) in NRAS or KRAS (1) or
CRAF amplification (2) lead to activation of
CRAF/MEK/ERK and resistance to RAF inhibition. p61BRAF(V600E) (3) forms constitutive
dimers resulting in activation of MAPK signaling. COT amplification (4) leads to activation
of MEK/ERK, independent of BRAF or CRAF.
Mutations (*) in MEK (5) that activate ERK can
lead to resistance. Bypass of MAPK signaling
through RTKs, such as PDGFR and IGF1R (6)
have been shown to mediate resistance to
vemurafenib.
Abbreviations: ERK, extracellular signalregulated kinase; IGF1R, insulin-like growth
factor-1 receptor; MAPK, mitogen-activated
protein kinase; MEK, mitogen-activated protein kinase kinase; PDGFR. platelet-derived
growth factor-beta; RTK, receptor tyrosine
kinase.
which confer resistance to gefitinib and erlotinib.8,10,11 Such

point mutations lead to either diminished drug binding or
stabilization of ATP binding. In striking contrast, secondary
mutations in BRAF(V600E) have not been reported thus far
in patients with melanoma treated with vemurafenib. However, bypass of BRAF(V600E) was observed through other
RAF family members, initiating sustained MAPK signaling.
Preclinical studies have demonstrated that either intrinsic
CRAF overexpression or upstream alterations leading to
CRAF activation (e.g., mutant NRAS or KRAS, or RAP-1
expression) confer robust resistance to RAF inhibition.12-14
Mutations in RAS at hotspot codons 12, 13, and 61, found
in vemurafenib-resistant cell lines, lead to constitutive activation of RAS. In turn, these mutant oncoproteins engage
MAPK signaling through CRAF, thus obviating BRAF(V600E)
inhibition (Table 1). Also, an activating mutation in KRAS
at K117N is associated with resistance to vemurafenib
through CRAF-mediated MAPK signaling.15 Additionally,
KEY POINTS
Resistance to inhibitors of mutant BRAF inhibitors

in metastatic melanoma is nearly universal, even
though initial response is observed for many patients.
Secondary mutations in BRAF have not been found in
resistant patient samples.
Reactivation of the MAPK pathway by activating
mutations in NRAS or KRAS, CRAF amplification, or
COT/MAP3K8 amplification confer resistance to
BRAF inhibitors.
Bypass of MAPK through alternative receptor tyrosine kinases, such as platelet-derived growth factor
or insulin-like growth factor-1 receptor may also
confer resistance.
Clinical mutations in MEK are cross-resistant to both
BRAF and MEK inhibitors, suggesting a mechanism
of resistance to any RAF/MEK inhibitor combination
therapy.
CRAF overexpression itself can lead to activation of MAPK

signaling. These studies clearly show that CRAF dysregulation can lead to resistance to mutant BRAF inhibitors (Fig.
1).
In addition to upstream activation of CRAF, recent studies have uncovered an alternatively spliced BRAF(V600E)
isoform that mediates resistance. This alternative splicing
yields a truncated BRAF variant, termed p61BRAF(V600E).
The p61 protein lacks the RAS-binding domain and thus
confers constitutive dimerization and sustained MEK/ERK
signaling, even in the presence of vemurafenib.16 This interesting mechanism generally affirms the importance of
RAF dimerization in production of paradoxical MEK/ERK
activation in the RAS-activated setting and has been observed in several clinical specimens analyzed thus far.17-19
Thus, sustained RAF-dependent signaling yields a prominent resistance mechanism even though secondary
BRAF(V600E) point mutations have not yet been detected
clinically.
In addition to RAF-dependent mechanisms described
above, resistance to mutant BRAF inhibition may also arise
by dysregulation of signaling effectors capable of circumventing RAF altogether. One example of this is COT, which
is a MAP3 kinase known to direct MEK/ERK signaling in
certain hematologic and inflammatory contexts.12 COT was
identified in a systematic functional screen for kinases and
related proteins that promote resistance to PLX4720, a
preclinical compound related to vemurafenib. Exogenous
expression of COT conferred 100-fold resistance to
BRAF(V600E)-expressing cell lines, and COT overexpression was confirmed in de novo resistant cell lines and tumors
that were resistant to vemurafenib. Although MEK/ERK
signaling was restored with COT overexpression in the
presence of PLX4720, BRAF or CRAF was not required for
COT-mediated MAPK activity. This result accorded well
with the notion that COT acts downstream of RAF and
directly activates MEK/ERK signaling. More generally,
these studies suggest that reactivation of the MAPK pathway (or sustained MAPK signaling) in the setting of RAF
inhibition may compose a broadly important category of
resistance to vemurafenib.
681
GOETZ AND GARRAWAY

Table 1. Summary of Studies Probing Resistance to BRAF Inhibitors
Mechanism of Resistance
In Vitro Studies
NRAS mutation (Q61K, G12D)
KRAS mutation (K117N)
p61BRAF(V600E)
CRAF overexpression
CRAF overexpression
COT/MAP3K8 amplification
IGF1R activation
PDGFR activation
MEK mutation (many)
In Vivo Studies
MEK mutation (C121S)
Druga
Type of Study
Confirmed In Vivo?
Reference
PLX4032
PLX4032
PLX4032
PLX4720
AZ628
PLX4720
SB-590885
PLX4032
AZD6244, PLX4032
Drug-resistant clones
Stepwise selection
Systematic ORF screen
Systematic ORF screen
Random mutagenesis
Yes
No
Yes
No
No
Yes
Yes
Yes
No
14
15
16
12
13
12
20
14
25
PLX4032
Targeted massively parallel sequencing
N/A
24
Abbreviations: IGF1R, insulin-like growth factor-1 receptor; MEK, mitogen-activated protein kinase kinase; ORF, open reading frame; PDGFR, platelet-derived growth
factor.
a
PLX4032 is vemurafenib. PLX4720 is the preclinical equivalent of PLX4032. AZ628 inhibits BRAF(V600E) and wild-type CRAF. SB-590885 is a BRAF(V600E) kinase
inhibitor. AZD6244 is a MEK inhibitor. The generation of drug-resistant clones (continuous exposure at one dose) or stepwise selection (steadily increasing drug
concentration) is described in the text.
MAPK Bypass
In addition to persistent MAPK pathway activation, alternative signaling pathways may confer resistance, as evidenced by the discovery of several receptor tyrosine kinases
(RTKs) associated with this phenomenon. In particular,
the RTKs HER2, AXL, platelet-derived growth factor-beta
(PDGFR), and insulin-like growth factor-1 receptor
(IGF1R) have been linked to vemurafenib resistance in
preclinical studies.12,14,20 In addition, PDGFR and IGF1R
were activated and phosphorylated in some tumors that
developed resistance to vemurafenib.14,20 These RTKs seem
to operate independently of MAPK signaling (Fig. 1), since
ectopic expression of these RTKs tended not to show MAPK
pathway reactivation. The extent to which these RTK alterations are sufficient to confer clinical resistance remains
unclear, because overexpression of PDGFR and IGF1R did
not confer resistance in some preclinical studies, nor did the
administration of imatinib resensitize melanoma cells that
exhibited PDGFR upregulation.12,21 Thus, their resistance
effects may manifest in certain molecular contexts or in
cooperation with as-yet unmeasured microenvironmental
effects. The RTK activation phenomena may also relate to
the relief of feedback inhibition that may result from therapeutic inhibition of activated oncoproteins.22
Although the majority of robust resistance mechanisms
characterized thus far tend to occur through downstream
pathway reactivation, there is some evidence that ERKindependent mechanisms may also play a role. The ERKindependent effects of RTKs such as PDGFR and IGF1R
constitute a line of evidence in this regard. In addition,
reports of disease progression with persistent suppression
of MEK/ERK activity have begun to emerge.23 The mechanisms of ERK bypass remain poorly characterized and may
gain increased importance if MEK inhibitors gain a role in
the treatment of BRAF-mutant melanoma.
Anticipating Resistance to Combined
RAF/MEK Inhibition
The preponderance of MEK/ERK-dependent resistance

mechanisms provides a strong rationale for the use of RAF
and MEK inhibitors in combination. Indeed, clinical trials of
combined RAF/MEK inhibition are ongoing. Several resistance mechanisms described above should in principle exhibit sensitivity to MEK inhibition (e.g., NRAS mutation,
682
COT/MAP3K8 overexpression/copy gain, BRAF alternative

splicing). On the other hand, activating mutations in MEKs
might be predicted to confer resistance to combined inhibitor
exposure (Fig. 1). Indeed, clinical evidence has already
begun to emerge in support of this hypothesis. A C121S
mutation in MEK1 was discovered in a clinical specimen
resistant to vemurafenib, and expression of C121S MEK1
in a vemurafenib-sensitive cell line conferred resistance to
both RAF and MEK inhibition.24 Thus, mutations in MEK
that occur clinically can confer resistance to both RAF and
MEK inhibitionand additional preclinical studies have
indentified several other mutations in MEK that were crossresistant to RAF and MEK inhibitors.24,25 In principle,
mutations in MEK would still be vulnerable to ERK inhibition. Toward this end, selective ERK inhibitors have
recently entered clinical trials. The role (and safe administration) of such agents in either up-front or salvage therapy
for BRAF-mutant melanoma will undoubtedly emerge as an
area of active investigation.
Methods of Therapeutic Resistance Discovery
The most common method for examining resistance involves generation of drug-resistant subclones using cancer
cell lines in vitro. This is accomplished by either slowly
increasing drug concentrations over time (stepwise selection) or through continuous incubation in high doses of a
drug (Table 1). After several months, the clones are analyzed
for mechanisms of resistance. Toward this end, changes in
gene expression, protein modification, or point mutations
can all lead to development of resistance. These changes
may sometimes be difficult to disambiguate; alternatively,
multiple mechanisms may contribute. Moreover, stepwise
selection may not accurately model the acquisition of resistance, since tumors are not exposed to an increasing gradient of a drug over a period of weeks or months. Therefore, an
alternative method for in vitro selection of drug-resistant
subclones involves continuous exposure at high drug doses.
This approach has proved useful in studies of resistance to
vemurafenib in vitro.
Several other techniques, such as random mutagenesis
and systematic screens, are useful to probe resistance in
vitro. These methods can speed up development of resistance (weeks instead of months) and may inform distinct
spectra of resistance mechanisms. Random mutagenesis is
RESISTANCE MECHANISMS TO MAPK INHIBITION
beneficial for defining mutations within the drug target that

may cause resistance. Here, a library of complementary
DNA (cDNA) is generated using error-prone polymerase
chain reaction (PCR) or by growth in mutagenic bacteria.
The mutated library is introduced into cells, followed by
incubation of these cells in the presence of high concentrations of the targeted agent. After resistance cells emerge,
large-scale DNA sequencing is used to identify candidate
resistance mutations, which are then validated experimentally. This technique has been used to successfully determine mutations in MEK1 that confer resistance to the MEK
inhibitor AZD6244 (Table 1).25 Chemical-based mutagenesis
using alkylating agents has also been used to examine
resistance to kinase inhibitors in Ba/F3 cells, a murine pro-B
cell line. The growth of Ba/F3 cells is dependent on
interleukin-3 (IL3), however, these cells can be engineered
to be independent of IL3 by constitutive expression of a
dominant oncogene such as BCR-ABL.26 Chemical-based
mutagenesis of Ba/F3 cells expressing BCR-ABL and grown
in the presence of imatinib recapitulated mutations found in
patients whose cancers confer resistance to imatinib.27 Additionally, genome-wide short hairpin RNA or open reading
frame (ORF) overexpression screens may be used to find
synthetic lethal and resistance alleles in a systematic way.
These methods may help define the spectrum of mechanisms
through which a cancer cell may manifest resistance. A
caveat to these methods is that they may not exactly
recapitulate in vivo resistance, given the intrinsic artificial
nature of cell culture experiments. Nonetheless, many of the
mechanisms determined in vitro have been confirmed in vivo
(Table 1).
The use of animal models, in addition to cell culture
models, to study drug resistance may also identify clinically
relevant resistance mechanisms. In addition to confirmation
of in vitro resistance mechanisms in animal models, the use
of animals can also inform novel mechanisms of resistance.

Animals bearing tumor xenografts can be treated with
targeted therapies, and, over time, drug-resistant tumors
analyzed for DNA or protein changes. Additionally, genetically engineered mouse models have been used to probe
resistance; a study examining the inhibition of vascular
endothelial growth factor receptor in a model of pancreatic
islet carcinogenesis led to the discovery that fibroblast
growth factor induction conferred resistance.28 Since these
experiments are performed in vivo, they may be more likely
to represent actual alterations found in patients, but they
can also be costly and difficult to perform.
After a drug has been used in clinical trials, analysis of
pretreatment and postrelapse patient biopsies will be effective means of determining resistance. Deep sequencing of
exomes or whole genomes will provide information on the
alterations in DNA sequence. In addition, analysis of protein
expression patterns or protein phosphorylation may also
help elucidate mechanisms of resistance. Examining preand post-treatment patient samples will be the most accurate way to study mechanisms of resistance; however, determining the mechanisms of resistance during preclinical
testing may be just as important as examining postrelapse
tumors.
Conclusion
With every medical therapy used for cancer treatment, de

novo or acquired resistance causes treatment failure in a
subset of patients. Therefore, studying resistance during
preclinical drug investigation in addition to after clinical
testing has failed is paramount to developing effective therapies. In the future, rationally designed drug combinations
will hopefully minimize the appearance of resistance tumors
and enhance clinical response.
Author
Eva M. Goetz*
Levi A. Garraway
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Daiichi Sankyo;
Foundation
Medicine;
Novartis
Foundation
Medicine
Honoraria
Boehringer
Ingelheim;
Constellation
Pharmaceuticals;
Novartis
Research
Funding
Expert
Testimony
Other
Remuneration
Novartis
REFERENCES
1. American Cancer Society. Cancer Facts & Figures. Atlanta, GA: American Cancer Society; 2012.
2. Wellbrock C, Ogilvie L, Hedley D, et al. V599EB-RAF is an oncogene in
melanocytes. Cancer Res. 2004;64:2338-2342.
3. Wan PT, Garnett MJ, Roe SM, et al. Mechanism of activation of the
RAF-ERK signaling pathway by oncogenic mutations of B-RAF. Cell. 2004;
116:855-867.
4. Davies H, Bignell GR, Cox C, et al. Mutations of the BRAF gene in
human cancer. Nature. 2002;417:949-954.
2011;364:2507-2516.
7. Dummer R, Robert C, Chapman PB, et al. AZD6244 (ARRY-142886) vs
temozolomide (TMZ) in patients (pts) with advanced melanoma: an openlabel, randomized, multicenter, phase II study. J Clin Oncol. 2008;26:491s
8. Kobayashi S, Boggon TJ, Dayaram T, et al. EGFR mutation and

resistance of non-small-cell lung cancer to gefitinib. N Engl J Med. 2005;352:
786-792.
Clin Cancer Res. 2005;11:4182-4190.
Science. 2001;293:876-880.
11. OHare T, Walters DK, Stoffregen EP, et al. In vitro activity of Bcr-Abl
inhibitors AMN107 and BMS-354825 against clinically relevant imatinibresistant Abl kinase domain mutants. Cancer Res. 2005;65:4500-4505.
968-972.
13. Montagut C, Sharma SV, Shioda T, et al. Elevated CRAF as a potential
mechanism of acquired resistance to BRAF inhibition in melanoma. Cancer
Res. 2008;68:4853-4861.
683
GOETZ AND GARRAWAY

973-977.
15. Su F, Bradley WD, Wang Q, et al. Resistance to selective BRAF
inhibition can be mediated by modest upstream pathway activation. Cancer
Res. 2012;72:969-978. Epub 2011 Dec 28.
16. Poulikakos PI, Persaud Y, Janakiraman M, et al. RAF inhibitor
resistance is mediated by dimerization of aberrantly spliced BRAF(V600E).
Nature. 2011;480:387-390.
17. Poulikakos PI, Zhang C, Bollag G, et al. RAF inhibitors transactivate
RAF dimers and ERK signalling in cells with wild-type BRAF. Nature.
2010;464:427-430.
18. Su F, Viros A, Milagre C, et al. RAS mutations in cutaneous squamouscell carcinomas in patients treated with BRAF inhibitors. N Engl J Med.
2012;366:207-215.
19. Oberholzer PA, Kee D, Dziunycz P, et al. RAS mutations are associated
with the development of cutaneous squamous cell tumors in patients treated
with RAF inhibitors. J Clin Oncol. 2012;30:316-321.
21. Shi H, Kong X, Ribas A, et al. Combinatorial treatments that overcome
PDGFRbeta-driven resistance of melanoma cells to V600EB-RAF inhibition.
Cancer Res. 2011;71:5067-5074.
684
22. Pratilas CA, Taylor BS, Ye Q, et al. (V600E)BRAF is associated with

disabled feedback inhibition of RAF-MEK signaling and elevated transcriptional output of the pathway. Proc Natl Acad Sci U S A. 2009;106:45194524.
23. McArthur GA, Ribas A, Chapman PB, et al. Molecular analyses from a
phase I trial of vemurafenib to study mechanism of action (MOA) and
resistance in repeated biopsies from BRAF mutation-positive metastatic
melanoma patients (pts). J Clin Oncol. 2011;29:526s (suppl; abstr 8502).
2011;29:3085-3096.
25. Emery CM, Vijayendran KG, Zipser MC, et al. MEK1 mutations confer
resistance to MEK and B-RAF inhibition. Proc Natl Acad Sci U S A.
2009;106:20411-20416.
26. Warmuth M, Kim S, Gu XJ, et al. Ba/F3 cells and their use in kinase
drug discovery. Curr Opin Oncol. 2007;19:55-60.
27. Bradeen HA, Eide CA, OHare T, et al. Comparison of imatinib
mesylate, dasatinib (BMS-354825), and nilotinib (AMN107) in an N-ethyl-Nnitrosourea (ENU)-based mutagenesis screen: high efficacy of drug combinations. Blood. 2006;108:2332-2338.
28. Casanovas O, Hicklin DJ, Bergers G, et al. Drug resistance by evasion
of antiangiogenic targeting of VEGF signaling in late-stage pancreatic islet
tumors. Cancer Cell. 2005;8:299-309.
Mechanisms of Resistance to Targeted

Therapies in Acute Myeloid Leukemia and
Chronic Myeloid Leukemia
By Catherine C. Smith, MD, and Neil P. Shah, MD, PhD
Overview: Small molecule kinase inhibitors of BCR-ABL in

chronic myeloid leukemia (CML) and of FMS-like tyrosine
kinase 3 internal tandem duplication (FLT3-ITD) in acute
myeloid leukemia (AML) have been successful at achieving
remissions in these diseases as monotherapy, but these
leukemias do not initially respond in a subset of patients
(primary resistance) and they progress in an additional group
LINICALLY EFFECTIVE small molecule inhibitors

that target pathologically activated tyrosine kinases
have become increasingly prevalent as cancer therapeutics.
Identification of resistance mechanisms in patients treated
with clinically active targeted therapeutics can yield critical
insights into the importance of specific pathways in human
cancers. Though the first highly successful molecularly targeted therapeutic in leukemia was all-trans retinoic acid,
which promotes the differentiation of acute promyelocytic
leukemic blasts driven by the fusion protein PMLRARalpha, mechanisms of resistance responsible for clinical
resistance to this agent remain poorly defined. Therefore,
this review will focus on emerging data surrounding clinical
resistance mechanisms to tyrosine kinase inhibitors (TKIs)
in CML and AML, which target oncogenic tyrosine kinases.
CML: A Model of Resistance to Targeted Therapy
The remarkable success of the small molecule inhibitor

imatinib, which achieves remissions in CML in all phases of
disease, ushered in a new era of molecularly targeted therapeutics and created a new paradigm for the treatment of
cancer. Though CML in a small percentage of patients with
chronic phase CML (CP-CML) does not respond to imatinib
initially (primary resistance), the overall hematologic and
major cytogenetic response (MCyR) rates (95.3% and 85.2%,
respectively) are exceedingly high in newly diagnosed patients.1 Response rates in the advanced phases of CML
(accelerated and blast phases; AP-CML) are substantially
lower, and remissions in AP-CML are typically shortlived.2,3 Additionally, a proportion of patients with CP-CML
also experiences relapse despite continuation of treatment
(secondary resistance). Elucidation of the molecular mechanisms responsible for clinical resistance to effective targeted
therapeutics can lead to strategies to improve clinical outcomes and also provide important insights into the role of
specific molecular targets in disease pathogenesis. To this
end, studies that interrogate primary patient samples are
most likely to both affect clinical management and be most
relevant to disease. Illustrative of this point, paradigmatic
studies of mechanisms of relapse on imatinib in patients
with CML confirmed the importance of BCR-ABL as critical
for CML pathogenesis; facilitated the development of the
effective second-generation inhibitors dasatinib and nilotinib, which are effective against most imatinib-resistant
mutations4,5; and informed studies of secondary resistance
mechanisms to active targeted therapies in a variety of
malignancies. Resistance to targeted therapies can be conceptualized in two major categories, which can inform efforts
of patients after an initial response (secondary resistance).

Resistance to these agents can be divided into mechanisms
that allow reactivation kinase activity and those that bypass
reliance on oncogenic signaling mediated by the target kinase.
Elucidation of clinical resistance mechanisms to targeted
therapies for patients can provide important insights into
disease pathogenesis and signaling.
to override resistance: (1) on-target resistance, whereby

the disease remains dependent on aberrant signaling mediated by the target oncogene, and (2) off-target resistance,
whereby activation of downstream or parallel signaling
pathways bypasses tumor dependence on the target oncogene (Fig. 1).
On-target Resistance: BCR-ABLDependent
Mechanisms of Resistance
In a seminal report, Gorre and colleagues evaluated 11

patients with advanced phase Philadelphia (Ph) chromosomepositive CML or acute lymphoblastic leukemia (ALL)
whose leukemias relapsed after initially responding to imatinib and found reactivation of BCR-ABL kinase activity at
the time of relapse in all cases, most commonly by gene
amplification (three patients) or point mutation in the BCRABL kinase domain (six patients), specifically a C3 T substitution coding for an isoleucine substitution at Thr315 (the
gatekeeper residue) predicted to impair imatinib binding6
by blocking access to a deeper hydrophobic pocket at the
ATP binding site in the kinase domain of BCR-ABL. This
mutation has been subsequently shown to mediate resistance to dasatinib and nilotinib as well.5,7 The findings of
Gorre and colleagues confirmed the continued central dependence of CML on BCR-ABL kinase activity. A second mechanism of clinical resistance provided by Gorre and
colleagues was genomic amplification of the BCR-ABL
genomic locus, with presumed overexpression of BCR-ABL
mRNA and protein. Though others have reported clinical
resistance to be associated with BCR-ABL overexpression or
the acquisition of additional Ph chromosomes,8 numerous
studies have confirmed that BCR-ABL kinase domain mutations represent the most common cause of relapse in all
phases of disease,9-11 both for imatinib and the secondgeneration BCR-ABL inhibitors dasatinib and nilotinib.
Although clinical resistance to imatinib has been associated with more than 90 different resistance-causing kinase
domain mutations to date,12 the second-generation inhibitors dasatinib and nilotinib are vulnerable to far fewer
resistance-conferring mutations. Dasatinib has activity
From the Division of Hematology/Oncology, University of California, San Francisco, CA.

Address reprint requests to Neil P. Shah, MD, PhD, Division of Hematology/Oncology,
UCSF, 505 Parnassus Avenue, Suite M1286, Box 1270, San Francisco, CA 94143; email:
nshah@medicine.ucsf.edu.
1092-9118/10/1-10
685
SMITH AND SHAH
against most imatinib-resistant BCR-ABL mutants in

vitro4,13 and clinically. Notable exceptions are the highly
resistant T315I mutation and the moderately resistant
F317L mutation,4 which is frequently associated with hematologic response, rarely with cytogenetic response, and occasionally with secondary resistance to dasatinib.10 Other
mutations associated with clinical resistance to dasatinib
include V299L, T315A, and F317I,4,10 which are largely
imatinib-sensitive. Nilotinib is a structural derivative of
imatinib that is similarly clinically vulnerable to a small
number of BCR-ABL kinase domain mutations, including
T315I,14 Y253H, E255V/K, and F359C/V.15,16 Collectively,
the three approved BCR-ABL kinase inhibitors appear to
retain activity against all BCR-ABL kinase domain mutants
associated with single nucleotide substitution, with the
exception of BCR-ABL/T315I mutants. Both dasatinib and
nilotinib have been approved for the treatment of CML in
the second line, and more recently in the first line, as a
result of improved cytogenetic and molecular response rates
associated with these agents relative to imatinib.17,18
Commonly, the choice of a second-line BCR-ABL inhibitor
in patients is guided by the specific mutation responsible for
imatinib resistance, as well as comorbid conditions that may
influence the choice of agent. Complex substitutions, including polyclonal (more than one mutation in a single patient)
and compound mutations (more than one mutation on a
single BCR-ABL allele), have also been reported to develop
in patients whose leukemias relapse on sequential treatment with BCR-ABL inhibitors,10,19 which makes the choice
of subsequent inhibitors less clear. Until recently, the T315I
mutation has remained a vexing problem. Encouragingly,
the investigational third-generation BCR-ABL inhibitor
KEY POINTS
686
Clinically relevant mechanisms of resistance to targeted therapies can provide insight into oncogenic
signaling and disease pathogenesis.
Resistance can be divided into mechanisms that reactivate kinase activity (on-target), indicating
maintained critical reliance on the target kinase, and
those that bypass kinase activity (off-target).
Secondary resistance to BCR-ABL and FMS-like tyrosine kinase 3 (FLT3) inhibitors is largely mediated
by on-target mechanisms, specifically kinase domain
mutations.
Increased levels of primary resistance in advanced
phases of chronic myeloid leukemia and in FLT3
internal tandem duplication mutant acute myeloid
leukemia may be indicative of cooperating lesions in
these diseases that allow bypass of oncogenic signaling mediated by the target kinase.
As more potent second- and third-generation BCRABL and FLT3 inhibitors with broader activity
against drug resistance causing kinase domain mutations become available, it is anticipated that resistance will be increasingly associated with off-target
mechanisms of resistance.
ponatinib has demonstrated substantial clinical activity in

patients with the T315I mutation in an interim analysis of a
pivotal phase II study,20 with 57% of patients with CP-CML
with the T315I mutation achieving a MCyR, defined as 35%
or less Philadelphia chromosomepositive bone marrow
metaphases. Ponatinib also demonstrated clinical activity
against other drug-resistant BCR-ABL mutant isoforms in
this study.20 Encouragingly, this agent appears to be highly
invulnerable to kinase domain mutation as a mechanism of
resistance in vitro,21 and translational studies that define
mechanisms of resistance to this agent will be of particular
interest. It is anticipated that the ability to clinically inhibit
all BCR-ABL kinase domain mutant isoforms will likely
render off-target mechanisms of resistance increasingly relevant.
Decreased expression of the drug transporter human
organic cation transporter 1 (hOCT1), responsible for imatinib influx, has also been correlated with suboptimal response to imatinib, though this could be overcome with
higher doses of imatinib.22 Pretreatment expression of
hOCT1 has also been correlated with clinical outcomes to
imatinib including response rate, progression-free survival
(PFS), and overall survival (OS).23 It is likely, however, with
evidence of improved molecular response rates with more
potent second-generation BCR-ABL inhibitors used in the
first line17,18 that this mechanism of clinical resistance will
become less relevant.
Off-target Resistance: BCR-ABLIndependent
Mechanisms of Resistance
To date, the molecular mechanisms of clinically relevant

BCR-ABLindependent resistance remain largely uncharacterized. It is clear that certain additional clonal cytogenetic
abnormalities are associated with progression from CP-CML
to AP-CML, decreased TKI responsiveness, and poorer prognosis. In a large study of 1,151 patients treated with imatinib, major-route cytogenetic abnormalities defined as a
second Ph chromosome, trisomy 8, isochromosome 17q, or
trisomy 19 were clearly associated with longer times to
achievement of complete cytogenetic response and major
molecular response and overall decreased PFS and OS.24 A
recent report suggested an association between translocations involving ecotropic viral integration site 1 (EVI1) and
TKI resistance in myeloid blast crisis disease.25 Additionally, the substantially lower initial hematologic response
rates to imatinib in Ph-positive ALL and AP-CML suggest
that advanced phases of disease may have decreased reliance on BCR-ABL signaling.2,3 The association between
acquired clonal cytogenetic abnormalities and disease progression despite TKI therapy and the lower overall TKI
response rates observed in advanced phases of disease
suggests that select cooperating genetic alterations may
functionally bypass dependence on BCR-ABL tyrosine kinase activity. Multiple in vitro studies have suggested that
CML cells may bypass dependence on BCR-ABL signaling
through aberrant activation of parallel or downstream signaling pathways by other means, including overexpression
of SRC-family kinases,26 overexpression of STAT5,27 or
activation of the PI3K/AKT pathway.28 However, the specific molecular mechanisms mediating this process in patients remain poorly defined.
RESISTANCE TO TARGETED THERAPIES IN AML AND CML
Fig. 1. Schematic depicting primary and secondary

tyrosine kinase inhibitor resistance, and on-target/offtarget resistance mechanisms.
AML: Emerging Targets and Emerging Resistance
The success of imatinib in CML has driven the desire to

translate the paradigm of targeted therapy to other cancers,
including other hematologic malignancies such as AML.
Unlike CML however, the genetic drivers underlying AML
pathogenesis are heterogeneous, and though multiple recurrent molecular abnormalities have emerged in recent years,
to date, no therapies targeting these lesions have been
approved. It is likely that, in contrast to the experience with
BCR-ABL, many of these molecular lesions represent cooperating rather than initiating genetic events in disease
pathogenesis, which common wisdom would deem to be
unlikely therapeutic targets.
Among the most common genetic lesions in AML are
constitutively activating mutations of FLT3. ITD mutations
in FLT3 are found in approximately 20% of patients with
AML and are associated with poor prognosis.29,30 Another
5% to 10% of patients with AML express constitutively
activating point mutations in the tyrosine kinase domain.31
Despite early enthusiasm, initial attempts to target FLT3
with small molecule inhibitors were largely unsuccessful,
resulting in only transient reductions in peripheral blasts
with few bone marrow responses.32,33 Emerging data have
further suggested that FLT3 mutations are likely secondary
rather than initiating genetic lesions in AML, hinting
that FLT3 mutations may not be critical for disease
pathogenesis.34-36
The recent success of the investigational FLT3 inhibitor
AC220 in achieving a composite complete remission rate of
45% in an interim analysis of a multinational phase II study
in patients with relapsed/refractory FLT3-ITDpositive
AML has rekindled interest in FLT3 as a therapeutic target.
The evolution of tyrosine kinase domain mutations in eight
of eight patients whose leukemias relapsed after a bone
marrow response on AC220 confirmed that clinical response
in these patients was achieved by inhibition of FLT3, and
that relapses were mediated by the reactivation of FLT3
kinase activity.37 Similar to BCR-ABL/T315I, F691, the
gatekeeper residue in FLT3, was mutated in a subset of
patients in whom AC220-resistant disease developed (to
leucine; 3 patients), but mutations at the activation loop
residue D835 developed in the remaining six patients (including one patient in whom both a F691L and D835V
mutation developed), suggesting that substitutions at the
D835 residue may be a common cause of AC220 resistance.
These findings validate FLT3-ITD as a therapeutic target
and further suggest that the activity of cooperating genetic
events that are believed to occur relatively late during
evolution of the malignant clone, such as FLT3-ITD, can
nonetheless be critically important to the survival of cells in
which they arise. Though this limited study suggests that
on-target resistance mechanisms are largely responsible for
secondary resistance to AC220, the lower overall response
rates to AC220 imply that off-target mechanisms are likely
to be important in both primary and secondary resistant
cases, as has been similarly postulated in cases of AP-CML.
It is hypothesized that cooperative genetic events in APCML and FLT3-ITDpositive AML can be placed into two
general categories, those that preserve reliance on the activated oncogene and those that bypass it. The latter category
may explain the lower initial response rates in blast phase
CML and FLT3-ITDpositive AML when compared with
chronic phase CML.
It has been suggested that another potential clinically
relevant FLT3-dependent mechanism of primary resistance
may involve increases in FLT3 ligand levels following chemotherapy, which is predicted to increase the concentration
of inhibitor required to effect cytotoxicity.38 This mechanism
may be more relevant in patients treated concurrently with
FLT3 inhibitors and chemotherapy, and further studies are
necessary to more accurately define its clinical importance.
Other targeted therapies currently undergoing clinical
development in AML include inhibitors of mammalian target of rapamycin,39 PI3K, and MEK,40 but the lack of clear
clinical efficacy to date precludes the establishment of clinically relevant resistance mechanisms. As was the case with
inhibitors of BCR-ABL and FLT3, elucidation of clinically
relevant resistance mechanisms necessitates inhibitors that
effect deep clinical responses when administered as monotherapy and detailed analysis of samples isolated from
patients at the time of secondary evolution.
687
SMITH AND SHAH

Conclusion
The overall success of small molecule inhibitors targeted

against BCR-ABL in CML has spurred the development of
inhibitors targeting other pathologically activated kinases.
The on-target BCR-ABL TKI resistance mechanism paradigm has been implicated as operative in a number of other
malignancies associated with clinically active kinase
inhibitors,41-43 now including FLT3-ITDpositive AML. In
particular, the concept that many relapses occur because of
the reactivation of oncogenic kinase activity mediated by

on-target mutation, amplification, decreased drug influx, or
increased ligand activation seems to be particularly clinically relevant and has led to rekindled interest in developing
potent selective FLT3 inhibitors. As inhibitors with increased potency/selectivity and decreased vulnerability to
resistance-conferring mutations become available, it is anticipated that off-target mechanisms of resistance, such as
activation of parallel or downstream signaling pathways,
will be more prominent in both CML and AML.
Author
Catherine C. Smith
Neil P. Shah
Employment or
Leadership
Positions
Consultant or
Advisory Role
Astellas
ARIAD; BristolMyers Squibb;
Novartis
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
Ambit; ARIAD;
Bristol-Myers
Squibb;
Plexxikon
REFERENCES
1. OBrien SG, Guilhot F, Larson RA, et al. Imatinib compared with
interferon and low-dose cytarabine for newly diagnosed chronic-phase chronic
myeloid leukemia. N Engl J Med. 2003;348:994-1004.
2. Sawyers CL, Hochhaus A, Feldman E, et al. Imatinib induces hematologic and cytogenetic responses in patients with chronic myelogenous leukemia in myeloid blast crisis: results of a phase II study. Blood. 2002;99:35303539.
3. Ottmann OG, Druker BJ, Sawyers CL, et al. A phase 2 study of imatinib
in patients with relapsed or refractory Philadelphia chromosome-positive
acute lymphoid leukemias. Blood. 2002;100:1965-1971.
4. Shah NP, Tran C, Lee FY, et al. Overriding imatinib resistance with a
novel ABL kinase inhibitor. Science. 2004;305:399-401.
5. Bradeen HA, Eide CA, OHare T, et al. Comparison of imatinib mesylate,
dasatinib (BMS-354825), and nilotinib (AMN107) in an N-ethyl-Nnitrosourea (ENU)-based mutagenesis screen: high efficacy of drug combinations. Blood. 2006;108:2332-2338.
Science. 2001;293:876-880.
7. Burgess MR, Skaggs BJ, Shah NP, et al. Comparative analysis of two
clinically active BCR-ABL kinase inhibitors reveals the role of conformationspecific binding in resistance. Proc Natl Acad Sci U S A. 2005;102:3395-3400.
8. Hochhaus A, Kreil S, Corbin AS, et al. Molecular and chromosomal
mechanisms of resistance to imatinib (STI571) therapy. Leukemia. 2002;16:
2190-2196.
Cancer Cell. 2002;2:117-125.
10. Shah NP, Skaggs BJ, Branford S, et al. Sequential ABL kinase
inhibitor therapy selects for compound drug-resistant BCR-ABL mutations
with altered oncogenic potency. J Clin Invest. 2007;117:2562-2569.
11. Soverini S, Colarossi S, Gnani A, et al. Contribution of ABL kinase
domain mutations to imatinib resistance in different subsets of Philadelphiapositive patients: by the GIMEMA Working Party on Chronic Myeloid
Leukemia. Clin Cancer Res. 2006;12:7374-7379.
12. Soverini S, Hochhaus A, Nicolini FE, et al. BCR-ABL kinase domain
mutation analysis in chronic myeloid leukemia patients treated with tyrosine
kinase inhibitors: recommendations from an expert panel on behalf of
European LeukemiaNet. Blood. 2011;118:1208-1215.
13. Talpaz M, Shah NP, Kantarjian H, et al. Dasatinib in imatinibresistant Philadelphia chromosome-positive leukemias. N Engl J Med. 2006;
354:2531-2541.
14. OHare T, Walters DK, Stoffregen EP, et al. In vitro activity of Bcr-Abl
inhibitors AMN107 and BMS-354825 against clinically relevant imatinibresistant Abl kinase domain mutants. Cancer Res. 2005;65:4500-4505.
15. Kantarjian H, Giles F, Wunderle L, et al. Nilotinib in imatinibresistant CML and Philadelphia chromosome-positive ALL. N Engl J Med.
2006;354:2542-2551.
688
16. Kantarjian HM, Giles FJ, Bhalla KN, et al. Update on imatinibresistant chronic myeloid leukemia patients in chronic phase (CML-CP) on
nilotinib therapy at 24 months: clinical response, safety, and long-term
outcomes. ASH Annual Meeting Abstracts. 2009;114 (abstr 1129).
17. Kantarjian H, Shah NP, Hochhaus A, et al. Dasatinib versus imatinib
in newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med.
2010;362:2260-2270.
18. Saglio G, Kim DW, Issaragrisil S, et al. Nilotinib versus imatinib for
newly diagnosed chronic myeloid leukemia. N Engl J Med. 2010;362:22512259.
19. Smith CC, Brown M, Chin J, et al. Single Molecule Real Time
(SMRT) sequencing sensitively detects polyclonal and compound BCR-ABL
in patients who relapse on kinase inhibitor therapy. ASH Annual Meeting
Abstracts. 2011;118 (abstr 3752).
20. Cortes JE, Kim D-W, Pinilla-Ibarz J, et al. Initial findings from the
PACE trial: a pivotal phase 2 study of ponatinib in patients with CML and
Ph ALL resistant or intolerant to dasatinib or nilotinib, or with the T315I
mutation. ASH Annual Meeting Abstracts. 2011;118 (abstr 109).
21. OHare T, Shakespeare WC, Zhu X, et al. AP24534, a pan-BCR-ABL
inhibitor for chronic myeloid leukemia, potently inhibits the T315I mutant
and overcomes mutation-based resistance. Cancer Cell. 2009;16:401-412.
22. White DL, Saunders VA, Dang P, et al. Most CML patients who have a
suboptimal response to imatinib have low OCT-1 activity: higher doses of
imatinib may overcome the negative impact of low OCT-1 activity. Blood.
2007;110:4064-4072.
23. Wang L, Giannoudis A, Lane S, et al. Expression of the uptake drug
transporter hOCT1 is an important clinical determinant of the response to
imatinib in chronic myeloid leukemia. Clin Pharmacol Ther. 2008;83:258-264.
24. Fabarius A, Leitner A, Hochhaus A, et al. Impact of additional cytogenetic aberrations at diagnosis on prognosis of CML: long-term observation of
1151 patients from the randomized CML Study IV. Blood. 2011;118:67606768.
25. Paquette RL, Nicoll J, Chalukya M, et al. Frequent EVI1 translocations
in myeloid blast crisis CML that evolves through tyrosine kinase inhibitors.
Cancer Genet. 2011;204:392-397.
26. Donato NJ, Wu JY, Stapley J, et al. BCR-ABL independence and LYN
kinase overexpression in chronic myelogenous leukemia cells selected for
resistance to STI571. Blood. 2003;101:690-698.
27. Warsch W, Kollmann K, Eckelhart E, et al. High STAT5 levels mediate
imatinib resistance and indicate disease progression in chronic myeloid
leukemia. Blood. 2011;117:3409-3420.
28. Quentmeier H, Eberth S, Romani J, et al. BCR-ABL1-independent
PI3Kinase activation causing imatinib-resistance. J Hematol Oncol. 2011;4:6.
29. Abu-Duhier FM, Goodeve AC, Wilson GA, et al. FLT3 internal tandem
duplication mutations in adult acute myeloid leukaemia define a high-risk
group. Br J Haematol. 2000;111:190-195.
30. Kottaridis PD, Gale RE, Frew ME, et al. The presence of a FLT3
internal tandem duplication in patients with acute myeloid leukemia (AML)
adds important prognostic information to cytogenetic risk group and response
RESISTANCE TO TARGETED THERAPIES IN AML AND CML

to the first cycle of chemotherapy: analysis of 854 patients from the United
Kingdom Medical Research Council AML 10 and 12 trials. Blood. 2001;98:
1752-1759.
31. Abu-Duhier FM, Goodeve AC, Wilson GA, et al. Identification of novel
FLT-3 Asp835 mutations in adult acute myeloid leukaemia. Br J Haematol.
2001;113:983-988.
32. Stone RM, DeAngelo DJ, Klimek V, et al. Patients with acute myeloid
leukemia and an activating mutation in FLT3 respond to a small-molecule
FLT3 tyrosine kinase inhibitor, PKC412. Blood. 2005;105:54-60.
33. Knapper S, Burnett AK, Littlewood T, et al. A phase 2 trial of the FLT3
inhibitor lestaurtinib (CEP701) as first-line treatment for older patients with
acute myeloid leukemia not considered fit for intensive chemotherapy. Blood.
2006;108:3262-3270.
34. Kelly LM, Liu Q, Kutok JL, et al. FLT3 internal tandem duplication
mutations associated with human acute myeloid leukemias induce myeloproliferative disease in a murine bone marrow transplant model. Blood. 2002;
99:310-318.
35. Shih LY, Huang CF, Wu JH, et al. Internal tandem duplication of FLT3
in relapsed acute myeloid leukemia: a comparative analysis of bone marrow
samples from 108 adult patients at diagnosis and relapse. Blood. 2002;100:
2387-2392.
36. McCormick SR, McCormick MJ, Grutkoski PS, et al. FLT3 mutations
at diagnosis and relapse in acute myeloid leukemia: cytogenetic and patho-
logic correlations, including cuplike blast morphology. Arch Pathol Lab Med.
2010;134:1143-1151.
37. Smith CC, Chin J, Wang Q, et al. Validation of FLT3-ITD as a
therapeutic target in human acute myeloid leukemia. ASH Annual Meeting
Abstracts. 2011;118 (abstr 937).
38. Sato T, Yang X, Knapper S, et al. FLT3 ligand impedes the efficacy of
FLT3 inhibitors in vitro and in vivo. Blood. 2011;117:3286-3293.
39. Rizzieri DA, Feldman E, Moore JO, et al. A phase 2 clinical trial of
AP23573, an mTOR inhibitor, in patients with relapsed or refractory hematologic malignancies. ASH Annual Meeting Abstracts. 2005;106 (abstr 2980).
40. Odenike O, Curran E, Iyengar N, et al. Phase II study of the oral MEK
inhibitor AZD6244 in advanced acute myeloid leukemia (AML). ASH Annual
Meeting Abstracts. 2009;114 (abstr 2081).
41. Pao W, Miller VA, Politi KA, et al. Acquired resistance of lung
adenocarcinomas to gefitinib or erlotinib is associated with a second mutation
in the EGFR kinase domain. PLoS Med. 2005;2:e73.
42. Cools J, DeAngelo DJ, Gotlib J, et al. A tyrosine kinase created by
fusion of the PDGFRA and FIP1L1 genes as a therapeutic target of imatinib
in idiopathic hypereosinophilic syndrome. N Engl J Med. 2003;348:12011214.
43. Choi YL, Soda M, Yamashita Y, et al. EML4-ALK mutations in lung
cancer that confer resistance to ALK inhibitors. N Engl J Med. 2010;363:
1734-1739.
689
TOWARD SUCCESSFUL TARGETING OF THE

PI3 KINASE PATHWAY IN CANCER
CHAIR
Paul Workman, PhD
The Institute of Cancer Research
Sutton, United Kingdom
SPEAKERS
John C. Byrd, MD
The Ohio State University Comprehensive Cancer Center
Columbus, OH
Carlos L. Arteaga, MD
Nashville, TN
Translating PI3K-Delta Inhibitors to the

Clinic in Chronic Lymphocytic Leukemia:
The Story of CAL-101 (GS1101)
By John C. Byrd, MD, Jennifer A. Woyach, MD, and Amy J. Johnson, PhD
Overview: Targeted therapy with imatinib has transformed

the treatment of chronic myeloid leukemia (CML). Unlike CML,
chronic lymphocytic leukemia (CLL) lacks a common genetic
aberration but does demonstrate constitutive activation of
PI3-kinase (PI3K) as compared to normal B cells. This constitutively active PI3K in CLL likely relates to tonic B-cell
receptor signaling that is present across a wide variety of
B-cell malignancies. Although PI3K is quite proximal and
represents an ideal target to pharmacologically modulate, the
complexity of this pathway on which many normal functions
are dependent had for many years been problematic. The p110
HE SUCCESS of imatinib in CML has prompted great

interest in developing targeted therapies for other
types of blood cancers. Unlike CML, most other blood cancers lack a genetically altered common target that is amendable for therapeutic targeting. Instead, pathways that are
active in normal cells at select times are often aberrantly
continuously activated. An example of this is B-cell receptor
(BCR) signaling in normal B cells, which facilitates proliferation, migration, and survival in the setting of activation by
an antigen but which is promptly downregulated as exposure to this diminishes. In contrast, BCR signaling is often
constitutively active in select B-cell malignancies, including
CLL and NHL. The etiology of this pathway activation is
multifactorial, as a result of enhanced microenvironment
activation, mutation of activating genes, or silencing of
specific phosphatases involved in proximal B-cell receptor
signaling. Although this pathway is quite complex and offers
several therapeutic targets for pharmacologic intervention,
one of the more proximal kinases activated is PI3K.
The PI3K Pathway
The PI3K pathway is a key component of cell survival in

many cancers, including CLL and low-grade NHL. PI3K is
activated by receptors or the small GTPase Ras and includes
multiple isoforms.1 There are three classes of PI3K isoforms;
however, only the class I isoforms phosphorylate inositol
lipids to form second messenger phosphoinositides. Specifically, class I PI3K enzymes convert PtdIns(3,4)P2 into
PtdIns(3,4,5)P3 that recruit downstream signaling proteins
such as BTK, PDK, AKT, ILK, and Rac GEF1,2 that can
generate survival signals. Class I isoforms are made up of
two subsets (IA and IB). Class IA encompasses p110 alpha,
p110 beta, and p110 delta (catalytic domains) that are bound
by p85, p50, or p55 (regulatory domains). Class IB is made
up solely of the p110 gamma (catalytic domain) bound by
the regulatory domain p101. The p110 alpha and p110 beta
isoforms are ubiquitously expressed, and knock-out mice for
both are embryonic lethal.3 It is thought that this widespread functionality of PI3K signaling is at least partially
responsible for the significant cellular toxicity associated
with pan-PI3K inhibitors such as LY294002.4 However, in
recent years it has been shown that these catalytic isoforms
have different expression profiles in different cell types.5-8
delta isoform of PI3K is relatively specific to hematopoietic

cells, and elegant mouse studies where p110 delta was genetically inactivated demonstrated only a selective B-cell defect.
Subsequent development of a potent, selective p110 delta
inhibitor prompted translation into the clinic for the treatment
of CLL and low-grade non-Hodgkin lymphoma (NHL). From the
first patient treated where a dramatic early nodal response
was noted, considerable excitement has developed for this
class of drugs in CLL and NHL. We will summarize the
development process of CAL-101 (now GS1101) in the treatment of chronic lymphoid malignancies such as CLL.
The expression of PI3K-delta is generally restricted to hematopoietic cells.9 Mice with deleted or kinase dead PI3Kdelta exhibit a B-cell defect, with a lack of B1 lymphocytes,
decreased mature B-cell numbers, and impaired antibody
production.3,5,10 Biochemically, B cells derived from PI3Kdelta knock-out mice also demonstrate less AKT phosphorylation when activated, decreased PIP3 levels, and
decreased ability to phosphorylate a YXXM phospho-specific
peptide.3 These mouse studies suggest that specific targeting of the PI3K-delta isoform may be cytotoxic to B cells with
minimal toxicity to other hematopoietic cell types. Forced
expression of PI3K-delta was shown to be transforming in
cell lines.11 These collective findings provided a rationale
to pursue isoform-specific PI3K inhibitors preclinically and
clinically in hematologic malignancies, including CLL. The
acquisition of a platform of novel PI3K-delta inhibitors by
Calistoga, a small biotechnology company with a very collaborative research team, prompted rapid exploitation of
this potential target in a variety of hematologic malignancies including CLL, which we will describe in this review.
Preclinical Targeting of the PI3K-delta
Pathway in CLL
With the goal of determining the feasibility of targeting

the isoform PI3K-delta in CLL, our group collaboratively
worked with Calistoga to initially examine 20 patients with
CLL for expression of the different PI3K isoforms. In these
patient samples, abundant PI3K-delta and PI3K-gamma
protein expression was found, but very modest PI3K-alpha
and essentially no PI3K-beta.12 Beyond this work, we confirmed a previous observation that CLL cells have increased
resting activity of PI3K activity as compared to normal B
cells. This finding was concurrently corroborated by Lan-
From the Department of Internal Medicine, Division of Hematology, the Comprehensive

Cancer Center at The Ohio State University, Columbus, OH; Division of Medicinal
Chemistry, College of Pharmacy, and the Comprehensive Cancer Center at The Ohio State
University, Columbus, OH.
Address reprint requests to John C. Byrd, MD, OSUCCC Building, Room 445B, 410 West
12th Ave., Columbus, OH 43210; email: john.byrd@osumc.edu.
1092-9118/10/1-10
691
BYRD, WOYACH, AND JOHNSON
nutti and colleagues, who demonstrated downstream constitutive phosphorylation of Threonine-308 on AKT, suggesting
activated PDK1 and PI3K in CLL cells as compared to
normal cells.13 Collectively, these findings demonstrated
that CLL cells may depend on PI3K-delta for constitutive
AKT activation and survival. To further confirm this finding, small interfering RNA targeting p110-delta was performed in CLL cells demonstrating knock-down of target
protein, diminished AKTThr-308 phosphorylation, and apoptosis. These findings provided rationale for further targeting of CLL with a potential isoform-specific inhibitor of
PI3K.
CAL-101 was an example of a PI3K-delta-specific inhibitor that is 100-fold more selective for PI3K-delta compared
to other PI3K isoforms.13 Whole blood studies in basophils
demonstrated appropriate inhibition of PI3K-delta at 50 M
concentrations, whereas PI3K-gamma-related targets were
not inhibited until M concentrations were attained. Furthermore, a kinase screen demonstrated significant selectivity for PI3K. The selectivity for PI3K-delta and other
favorable drug properties prompted the choice of CAL-101 as
a lead candidate for clinical development in hematologic
malignancies and initiation of preclinical studies with this
agent. Here we found that CAL-101 exhibits dose-dependent
induction of apoptosis in CLL tumor cells that is independent of patient genomic features, including immunoglobulin
gene variable heavy gene (IVGH) mutational status and
del(17p13.1).12 In contrast, CAL-101 did not promote cytotoxicity toward T cells or natural killer cells. We next
confirmed the effect of microenvironment stimuli on CLL
cells, including that soluble factors such as CD40L, BAFF,
and TNF- and contact factors (fibronectin and stromal cell
coculture) all can activate PI3K and downstream AKT,
thereby promoting CLL survival.12 However, treatment
with CAL-101 in the context of any of these microenvironment protective features was shown to abrogate the AKT
phosphorylation and protection from cell death. CAL-101
modulation of microenvironment protection was confirmed
later by the Burger laboratory that also demonstrated disrupted CLL cell migration, adhesion, and diminished CLL
production of chemokines that recruit immune cells to tumor
sites.14 Similar findings in mantle cell lymphoma and other
types of B-cell lymphoma,13 as well as multiple myeloma,15
were also shown. Collectively, these in vitro studies provided
a strong rationale for moving forward to the clinic with
CAL-101 with particular focus on targeting CLL and other
types of B-cell lymphomas as the lead study diseases.
KEY POINTS
692
PI3-kinase-delta in genetic knock-out models has

predominately a B-cell phenotype.
GS1101 is a PI3-kinase-delta-specific orally administered agent in clinical trials for chronic lymphocytic
leukemia and nonHodgkin lymphoma.
GS1101 demonstrates a unique pattern of response
with early lymphocytosis concomitant with reduction
in lymph-node and spleen disease.
GS1101 has a favorable side-effect profile with a
dose-limiting side effect of transient transaminitis.
Additionally, these preclinical studies provided for biomarkers of target inhibition such as AKTThr308 phosphorylation
loss by flow cytometry, cytokine production by tumor and
normal immune cells, and also chemokine production by
tumor cells.
CAL-101: Early Clinical Development
The selectivity of CAL-101 and also a very favorable

toxicity profile observed in preclinical toxicology at doses
where PI3K-delta was inhibited allowed an initial study in
healthy volunteers. This phase I study demonstrated favorable pharmacokinetics during a week of CAL-101 oral administration with absent lymphocyte count changes or
notable toxicity. A phase I study of CAL-101 in select
lymphoid malignancies was initiated in July 2008 beginning
at a 50-mg once daily dosage. Unlike many phase I studies
that struggle through enrollment with only modest efficacy,
dramatic reduction in tumor size was noted in the first
patient at the 1-month response evaluation. Accrual to this
phase I trial was quite rapid, with efficient data communication and extensive investigator involvement typical of
early studies performed with small biopharmaceutical companies. Dose escalation proceeded through several dose
levels with a loss of linear pharmacokinetics at higher doses
examined. The most predominate toxicity observed with
CAL-101 was reversible transaminitis early during therapy,
which usually resolved with temporary discontinuation of
the drug. Reinitiation of CAL-101 at lower doses generally
was possible without recurrence of transaminitis. Perplexing, however, was the finding that this toxicity occurred
much more frequently in patients with lymphoma rather
than with CLL. This liver toxicity and the nonlinear pharmacokinetics observed with CAL-101 required exploration of
multiple dose levels that ultimately delayed development of
this compound proceeding to phase II formal testing. Results for the CLL and low-grade lymphoma cohort were
most promising, prompting focused expansion in each of
these groups. In contrast, single-agent activity in aggressive
large cell lymphoma, refractory acute myeloid leukemia, and
multiple myeloma was not observed.
Disease-Specific Activity and Toxicity
of CAL-101 in CLL
With respect to CLL and its exploration in the initial

phase I study, a dose of 150 mg every 12 hours was identified
to be ideal for the phase II dose in CLL based on tolerability
and minimal increase in plasma Cmax at increasing doses. Of
54 patients with CLL enrolled in this trial, 84% achieved a
decrease in lymph node and spleen size of 50% or greater. An
increase in peripheral lymphocyte count of more than 50%
was seen in 58% of patients, consistent with other BCR
antagonists. Lymphocytosis peaked at 2 months and resolved over time in a subset of patients. As a consequence of
the lymphocyte count not falling to 50% of baseline, response
across all patients enrolled was 24%. This response was
independent of high-risk genetic features, bulky adenopathy, prior therapy, or presence of cytopenias. Median
progression-free survival was 15 months, with 46% of patients continuing on therapy at the time of presentation.16
Side effects with this agent have been mild and have
included rare cytopenias and pneumonia. Approximately
THE STORY OF CAL-101
6% of patients developed grade 3 or 4 transient liver function abnormalities during the early phase, which was reversible with holding therapy and generally did not recur
with resumption at a lower dose level. Several of the pharmacodynamic studies optimized with the preclinical CAL101 work confirmed target inhibition of PI3K-delta in vivo in
patients receiving this agent, including serial loss of AKTThr308
phosphorylation, diminished chemokine, and also
cytokine levels in plasma. These data collectively provide
support that CAL-101 has significant clinical activity in CLL
and can be administered as continuous therapy for an
extended period of time with very modest toxicity.
Based on the favorable toxicity observed with CAL-101
monotherapy, target inhibition of PI3K-delta, and early
lymphocytosis observed that was believed to be representative of the ability of CAL-101 to mobilize CLL cells from
protected stromal cell niches, combination studies with
other therapies were clearly justified. CAL-101 studies with
either monoclonal antibodies (rituximab or ofatumumab) or
chemotherapy agents (bendamustine, bendamustine and
rituximab, or fludarabine) in the phase I setting have been
pursued. In these studies, CAL-101 was dosed at 100 mg
twice daily or 150 mg twice daily with other drugs administered as standard. Therapy was well-tolerated, with no
toxicities in addition to those expected with the single
agents. Response data have been presented for the CAL-101
studies combined with rituximab, bendamustine, and bendamustine and rituximab. For patients receiving CAL-101
with rituximab or bendamustine, 90% or more of patients
achieved a reduction in lymph node size of 50% or more. For
three patients treated with bendamustine, rituximab, and
CAL-101, all patients achieved a lymph node response.17
Using traditional IWCLL 2008 CLL response criteria, more
than 80% of patients receiving each of these three regimens
met criteria for response. Collectively, these data provide
evidence that CAL-101 can be safely combined with other
therapies used in CLL and also can contribute to durable
remissions with these agents.
Future Directions for CAL-101 and Other PI3K

Inhibitors in CLL
The documented success of CAL-101 has prompted transition of the compound from Calistoga, a very small biotechnology company, to Gilead, a much larger pharmaceutical
company. This transition resulted in both a renaming of
CAL-101 to GS1101 and also the initiation of more definitive
registration studies for eventual marketing approval in CLL
and low-grade NHL. As an expected consequence of the
success of CAL-101, other PI3K-delta specific inhibitors are
now being explored in CLL and related B-cell malignancies.
Additionally, trials with more broad PI3K inhibitors have
also been initiated with documented clinical activity. In
particular, PI3K inhibitors targeting the PI3K-alpha isoform may prove worthwhile in CLL. Burger and colleagues
have previously demonstrated that PI3K-alpha antagonists
abrogate stromal microenvironment protection against
chemotherapy-mediated death.18 Additionally, the success
of CAL-101 has brought forth inhibitors of other kinases
downstream in the BCR signaling pathway including BTK,
which have also demonstrated dramatic clinical responses.
At this point, the field of CLL has been mesmerized by the
durable activity of these BCR antagonizing agents and also
the true potential of these therapeutics to favorably affect
the natural history of CLL.
Acknowledgements
The authors wish to acknowledge the employees of Calistoga
Pharmaceuticals intricately involved in CAL-101 development
(Neill Giess, PhD, Brian Lannutti, PhD, Carol Gallagher, Albert
Yu, MD, David Johnson, and Langdon Miller, MD); the early
phase I investigators on this trial (Ian Flinn, MD, Brad Kahl,
MD, Richard Furman, MD, Jennifer Brown, MD, and Nancy
Bartlett, MD); and most importantly, the patients who enrolled
in these trials. We are grateful for research support that has
made possible our BCR kinase work from The Leukemia and
Lymphoma Society, the National Cancer Institute (P50
CA140158, PO1 CA95426, PO1 CA81534, 1K12 CA133250), Mr.
and Mrs. Michael Thomas, The Harry Mangurian Foundation,
and The D. Warren Brown Foundation.
Author
Employment or
Leadership
Positions
Consultant or
Advisory Role
Stock
Ownership
Honoraria
Research
Funding
Expert
Testimony
Other
Remuneration
John C. Byrd*
Jennifer A. Woyach*
Amy J. Johnson*
REFERENCES
1. Cantrell DA. Phosphoinositide 3-kinase signalling pathways. J Cell Sci.
2001;114:1439-1445.
2. Wymann MP, Zvelebil M, Laffargue M. Phosphoinositide 3-kinase signallingwhich way to target? Trends Pharmacol Sci. 2003;24:366-376.
3. Vanhaesebroeck B, Ali K, Bilancio A, et al. Signalling by PI3K isoforms:
insights from gene-targeted mice. Trends Biochem Sci. 2005;30:194-204.
4. Xu GB, Lu XG, Luo LH, et al. Detection of soluble Apo-1/Fas in plasma,
pleural and ascites fluid of malignant tumor patients and its clinical significance (in Chinese). Zhejiang Da Xue Xue Bao Yi Xue Ban. 2003;32:335-338.
5. Jou ST, Carpino N, Takahashi Y, et al. Essential, nonredundant role for
the phosphoinositide 3-kinase p110delta in signaling by the B-cell receptor
complex. Mol Cell Biol. 2002;22:8580-8591.
6. Fruman DA. Towards an understanding of isoform specificity in phosphoinositide 3-kinase signalling in lymphocytes. Biochem Soc Trans. 2004;
32:315-319.
7. Papakonstanti EA, et al. Distinct roles of class IA PI3K isoforms in

primary and immortalised macrophages. J Cell Sci. 2008;121:4124-4133.
8. Min SH, Abrams CS. Why do phosphatidylinositol kinases have so many
isoforms? Biochem J. 2009;423:e5-e8.
9. Vanhaesebroeck B, Welham MJ, Kotani K, et al. P110delta, a novel
phosphoinositide 3-kinase in leukocytes. Proc Natl Acad Sci U S A. 1997;94:
4330-4335.
10. Okkenhaug K, Bilancio A, Farjot G, et al. Impaired B and T cell antigen
receptor signaling in p110delta PI 3-kinase mutant mice. Science. 2002;297:
1031-1034.
11. Kang S, Denley A, Vanhaesebroeck B, et al. Oncogenic transformation
induced by the p110beta, -gamma, and -delta isoforms of class I phosphoinositide 3-kinase. Proc Natl Acad Sci U S A. 2006;103:1289-1294.
12. Herman SE, Gordon AL, Wagner AJ, et al. Phosphatidylinositol
3-kinase-delta inhibitor CAL-101 shows promising preclinical activity in
693
BYRD, WOYACH, AND JOHNSON

chronic lymphocytic leukemia by antagonizing intrinsic and extrinsic cellular
survival signals. Blood. 2010;116:2078-2088.
13. Lannutti BJ, Meadows SA, Herman SE, et al. CAL-101, a p110delta
selective phosphatidylinositol-3-kinase inhibitor for the treatment of B-cell
malignancies, inhibits PI3K signaling and cellular viability. Blood. 2011;117:
591-594.
14. Hoellenriegel J, Meadows SA, Sivina M, et al. The phosphoinositide
3-kinase delta inhibitor, CAL-101, inhibits B-cell receptor signaling and chemokine networks in chronic lymphocytic leukemia. Blood. 2011;118:3603-3612.
15. Ikeda H, Hideshima T, Fulciniti M, et al. PI3K/p110{delta} is a novel
therapeutic target in multiple myeloma. Blood. 2010;116:1460-1468.
16. Coutre SE, et al. Phase I study of CAL-101, an isoform-selective
694
inhibitor of phosphatidylinositol 3-kinase P110d, in patients with previously

treated chronic lymphocytic leukemia. J Clin Oncol. 2011;29,451s (suppl;
abstr 6631).
17. Sharman J, de Vos S, Leonard JP, et al. A phase 1 study of the selective
phosphatidylinositol 3-kinase-delta (PI3Kdelta) inhibitor, CAL-101 (GS1101), in combination with rituximab and/or bendamustine in patients with
relapsed or refractory chronic lymphocytic leukemia (CLL). Blood (ASH
Annual Meeting Abstracts) 2011;118,?? (abstr 1787).
18. Niedermeier M, Hennessy BT, Knight ZA, et al. Isoform-selective
phosphoinositide 3-kinase inhibitors inhibit CXCR4 signaling and overcome
stromal cell-mediated drug resistance in chronic lymphocytic leukemia: a
novel therapeutic approach. Blood. 2009;113:5549-5557.
PI3 Kinase in Cancer: From Biology to Clinic

By Paul Workman, PhD, and Paul Clarke, PhD
Overview: The discovery and clinical development of smallmolecule inhibitors of the phosphatidylinositide 3-kinase (PI3
kinase) family of lipid kinases have marked a remarkable
20-year journey that follows the progressive developments in
cancer biology over the last few decades: from hypothesisdriven, basic cancer research that began with viral oncogenesis and developed in the 1960s and 70s, through the
discovery of individual mutated oncogenes and tumor suppressor genes in 1970 and 80s and the linkage of these cancer
genes to signal transduction pathways in the 1990s, to all
large-scale genome-wide sequencing, functional screening,
and network biology efforts today. Thus, PI3 kinase research
began with the discovery in 1985 of a new type of enzyme
activity associated with viral oncogenesis. It benefited greatly
from the discovery of wortmannin and LY294002 as PI3 kinase
HE PI3 kinase family of lipid and protein kinases

regulates an intracellular signaling network that controls many features of cell behavior, including growth,
survival, motility, metabolism, and additional specialized
functions (Fig. 1). There are four distinct PI3 kinase subfamilies that are categorized by their substrate specificities,
primary structures, modes of regulation, and domain content. Of these, the class I isoforms (p110, , , ) and class
IV PI3 kinase-related protein kinase mTOR have been the
most intensively explored as targets for small-molecule
therapeutics.1-3
The rationale for the development of the class I PI3 kinase
group of lipid kinases is the observation of frequent genetic
and epigenetic alterations that result in activation of the PI3
kinase pathway.2,3 The class I PI3 kinases catalyze the
addition of a phosphate group to the 3-hydroxyl position of
the inositide ring present in membrane phosphatidylinositides. This produces products, the most notable of which is
phosphatidylinositol-3,4,5-trisphosphate (PIP3), which acts
as a second messenger that recruits PKB/AKT to the cell
membrane (Fig. 1). The phosphatase PTEN, a negative
regulator of PI3 kinase signaling that dephosphorylates
PIP3, is one of the most commonly mutated tumorsuppressor proteins in human malignancy.4 In contrast, the
gene encoding PIK3CA, the p110 catalytic subunit, is
amplified, overexpressed, and frequently mutated in many
cancers.5
A greater understanding of the specific and geneticsdependent roles of the class I isoforms in tumorigenesis
has been established. It has been shown that p110 is
critical for the growth of tumors driven by PIK3CA mutations and is activated by KRAS and also by receptor tyrosine
kinases that are in turn activated by ligand or by oncogenic
mutation, amplification, or translocation.2,3 In contrast,
p110 has been identified as the principal isoform mediating
tumorigenesis in PTEN-deficient backgrounds.6-8 The deltaisoform, p110, has also emerged as a potential therapeutic
target for hematological malignancies, notably acute myeloid leukemia, and perhaps in neuroblastoma, melanoma,
and breast cancers also.1,2,9,10 Finally, there is potential for
all the class I PI3 kinases to be activated in cancer cells
through mutation of the p85 regulatory subunits that recruit the catalytic subunit to growth factor receptors.
inhibitors and chemical tools in late 1980s to mid-90s. Alongside these tools, genetic validation of PI3 kinase as a target
initially involved activation by upstream oncogenic receptor
tyrosine kinases and RAS mutation, together with overexpression and amplification of the p110 catalytic isoform of PI3
kinase and frequent loss of the tumor suppressor and negative
regulator of PI3 kinase activity, PTEN. As PI3 kinase drug
development began, further stimulus came from the discovery
through genome sequencing of mutations in PIK3CA, which
encodes p110 and is the most frequently mutated kinase in
the human genome. From these beginnings, there are now
many PI3 kinase inhibitors in clinical trials and more in
preclinical development. We review progress, current challenges, and future opportunities in this article.
Based on this broad platform of research validation,

inhibition of the class Iand especially the class IA subclass
(p110, , ) has emerged as an important strategy for the
development of novel molecular cancer therapeutics. Looking further ahead, PI3 kinase inhibitors are anticipated to
have a significant impact on the discovery and development
of new personalized medicines in the oncology setting as well
as finding utility in other disease areas.10
Discovery of PI3 Kinase Inhibitors
Since the elucidation of the mechanism of action of the

natural product wortmannin and the discovery of the synthetic flavone LY294002, both of which inhibit the class I
PI3 kinase isoforms and have served as valuable chemical
tools, considerable progress has been made in the design of a
plethora of small-molecule inhibitors.10-12 High-throughput
screening of compound collections supported by rational
structure-based design combined with medicinal chemistry
optimization has led to the discovery and development of
many chemically diverse inhibitors that possess a range of
PI3 kinase subtypeselectivity profiles.
A number of these inhibitors, such as GDC-0941, GDC0980, NVP-BEZ235, NVP-BMK120, GSK-2126458, GSK1059615, PF-04691502, XL147, XL765, CAL-101, and
derivatives of wortmannin and LY294002 (as reviewed by
Shuttleworth and colleagues10), have entered early-phase
human trials over recent years and are in some cases
progressing to phase Ib expansion cohort and phase II
single-agent efficacy studies and early combination trials.
A paradigm for the rational discovery and development of
a potent and selective PI3 kinase inhibitor is GDC-0941 that
has a well-defined and informative case-history.13 In 2007
Hayakawa and colleagues reported the identification of a
pyridofuropyrimidine lead and improved chemical tool com-
From the Cancer Research UK Cancer Therapeutics Unit, Division of Cancer Therapeutics, The Institute of Cancer Research, Sutton, Surrey UK.
Address reprint requests to Paul Workman, PhD, Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, 15 Cotswold Road,
Sutton, Surrey, SM2 5NG UK; email: paul.workman@icr.ac.uk.
1092-9118/10/110
e93
WORKMAN AND CLARKE

14
pound. This chemical probe, PI-103, exhibited advantages

over wortmannin and LY294002, with excellent potency and
selectivity for class I PI3 kinases compared with a panel of
other protein and lipid kinases. PI-103 exhibited therapeutic
activity against human tumor xenograft models representing various cancer types, including several with PIK3CA
mutations or loss of PTEN, which are predicted to be
addicted to the PI3 kinase pathway.15 The antitumor
activity of PI-103 was also associated with effects on pharmacodynamic pathway biomarkers, including reduced phosphorylation of downstream substrates such as AKT,
providing evidence of target engagement. The biomarker
results were fully consistent with the therapeutic mechanism being PI3 kinase inhibition, and, importantly, clearly
demonstrating proof-of-concept for targeting this pathway.
At the same time, it was also recognized that PI-103 had
some liabilities, specifically limited aqueous solubility and
extensive metabolism, which were considered serious drawbacks in terms of its drug-like properties and thus its
suitability for clinical development
Subsequently, a large number of chemical analogs from
two chemical starting pointsthe pyridofuropyrimidines
exemplified by PI-103 and the thienopyrimidine compound
12were designed, synthesized, and tested in an iterative
fashion.14,16 These integrated medicinal chemistry, and
biogical evaluation efforts identified the more advanced
thienopyrimidine lead compounds, PI-540 and PI-620, which
retained the potency and selectivity of PI-103 but with much
improved aqueous solubility and reduced in vitro metabolism.17 Although the improved absorbtion, distribution, metabolism, and excretion properties of these compounds
resulted in correspondingly greater antitumor activity as
compared with PI-103, in vivo metabolism caused by conjugation of a phenolic hydroxyl group remained an obstacle
to the clinical development of this series.17 Indazoles were
investigated as replacements for the problematic phenolic
hydroxyl, with the intention that they might markedly
reduce metabolism while maintaining the necessary target
interaction with PI3 kinase, as revealed by protein-ligand
X-ray cocrystal structures (Fig. 2). This strategy paid off and
GDC-0941 (PI-728) was identified as a potent and selective
KEY POINTS
e94
The PI3K pathway is frequently activated by oncogenic mutations in many different cancers.
Inhibiting the PI3K pathway is an attractive approach to exploit the resulting oncogenic addictions
and vulnerabilities.
A number of potent and selective PI3K inhibitors
have been discovered and are now in clinical studies.
Although fit-for-purpose target engagement biomarkers are now available, identifying predictive biomarkers for patient stratification remains a major
challenge; current predictive biomarkers are useful
for enrichment of early clinical trials with patients
more likely to respond.
Other key challenges being addressed in the clinic are
to define optimal PI3 kinase selectivity profiles and to
explore rational drug combinations.
inhibitor pan-class I PI3 kinase inhibitor with excellent

drug-like properties, including good oral bioavailability, a
major advantage over the previous analogs.18 Interestingly,
GDC-0941 retained full therapeutic activity, at least in the
cancer models tested, despite lacking mTOR inhibition,
which was in contrast to previous analogs such as PI-103.17
A subsequent structurally related clinical candidate GDC0980 does inhibit mTOR, providing an interesting comparator.19
The improved pharmacokinetic profile of GDC-0941 resulted in profound and prolonged target engagement, demonstrated by inhibition of PI3 kinase pathway biomarkers in
human cancer xenografts, consistent with the sustained
tumor drug exposure (Fig. 3). In these experiments, tumor
drug levels were well above antiproliferative GI50 concentrations for at least 6 hours, and decreased phosphorylation
of AKT together with other pathway markers was maintained for at least 8 hours (Fig. 3).17
As a result of these enhanced pharmacokinetic and pharmacodynamic properties, GDC-0941 exhibited excellent
dose-dependent therapeutic activity in the PTEN null, PI3
kinase pathway-addicted U87MG human glioblastoma
xenograft model (Fig. 3). Activity was also demonstrated in
another PI3 kinase-addicted tumor, the IGROV-1 ovarian
cancer carrying both an activating PIK3CA mutation and
lacking PTEN expression and also in additional human
tumor xenografts.17,20
The linkage of pharmacokinetic exposure, biomarker
changes in the PI3 kinase pathway, and therapeutic tumor
response provided a convincing pharmacologic audit trail
(PhaT21) for GDC-0941, which formed a solid platform for
subsequent clinical studies. On the basis of its very attractive molecular, pharmacologic, and therapeutic profile, together with its excellent selectivity profile for class I PI3
kinase compared with a wide range of other kinases, GDC0941 was selected for clinical development. As mentioned
earlier, there are now a considerable number of PI3 kinase
inhibitors undergoing clinical trials and still more in preclinical discovery and development.10
Current Challenges and Future Opportunities
It is important to assess the critical challenges that we

face in the development of PI3 kinase inhibitors for cancer
treatment.22 A major question is certainly the desired specificity toward the various individual PI3 kinase family
members. There has been some real progress, particularly in
designing and evaluating inhibitors with different isoform
selectivities, often facilitated by structure-based design (Fig.
2). However, the definition of the optimal PI3 kinase isoform
selectivity profile or profileswhich may well depend on
tumor geneticsis still a major ongoing research activity.
For example, we remain unsure about whether the addition
of mTOR inhibitory activity to a class I PI3 kinase inhibitor
activity is advantageous, e.g., for overcoming feedback loops.
In addition, it is also not yet clearly established whether
individual compared with combinatorial inhibition of class I
isoforms is advantageous especially given that most clinical candidates inhibit all 4 class I isoforms with low nanomolar potency and appear to be generally well tolerated,
at least in the short to medium term. The preferred profiles
will emerge with time from clinical experience coupled with
mechanistic preclinical studies.
The issue of biomarkers is a critically important one. The
PI3 KINASE IN CANCER
Fig. 1. The PI3 kinase network. Class I PI3 kinases are activated to varying extents by receptor tyrosine kinases (RTKs) and G-protein coupled
receptors (GPCRs). PI3 kinases phosphorylate the 3-hydroxy position of inositol ring of phosphatidylinositides, yielding lipid products that
include phosphatidylinositol-3,4,5-trisphosphate (PIP3), generated from phosphatidylinositol-4,5-bisphosphate (PIP2). PIP3 is a second messenger molecule that recruits certain protein kinases with PH domains, such as PDK1 and AKT, to the cell membrane (dashed line), resulting in
their activation and stimulation of the PI3 kinase signaling network. PI3 kinase may promote cancer through both AKT-dependent and
AKT-independent mechanisms, the latter via PDK1 and the serine/threonine-protein kinase 3 (SGK3).31 Activation of the PI3 kinase pathway is
very common in a wide range of cancers. PIK3CA, which encodes the p110 catalytic subunit of PI3 kinase, appears to be the most commonly
mutated kinase in the entire human cancer genome (12% of all cancers) and is also amplified in some tumors. PTEN, which encodes the opposing
phosphatase to PI3 kinase, is the second most commonly affected tumor-suppressor gene after p53.32 Activation of PI3 kinase-signaling in
cancer also occurs as a result of mutation or increased expression of RTKs, AKT, and RAS. Activation of PI3 kinase signaling has been shown to
reduce cellular dependence on growth factors, attenuate apoptosis, and facilitate tumor growth and invasiveness. Class I PI3 kinases are shown
here as molecular targets of PI3 kinase inhibitors. Figure adapted with permission from Clarke and Workman. 2012 American Society of
Clinical Oncology. All rights reserved.22
use of proof-of-mechanism pharmacodynamic biomarkers to

demonstrate target and pathway modulation in both the
preclinical discovery phase and the early clinical development of PI3 kinase inhibitors is crucial for the implementation of the PhaT.21 This enables rational optimization of
dose and schedule of administration as well as go/no-go
decision making.23 Available proof-of-mechanism biomarkers of signaling output, such as phosphorylation of ribosomal
protein S6, PRAS40, 4EBP1, and AKT have been used as
indicators of PI3 kinase-pathway inhibition in both tumor
and surrogate tissues and certainly appear to be fit for
purpose.10
In contrast, we do not yet have biomarkers that are
robustly predictive for sensitivity and resistance in the
clinic. Several nonclinical studies have defined intrinsic
factors such as HER2 amplification, PIK3CA mutations, or
combined defects in these genes as factors that may influence the sensitivity of breast tumors to PI3 kinase inhibitors.10 On the other hand, the presence of a KRAS mutation
may be a dominant determinant of resistance in other types

of cancer.10,13 The currently available therapeutic response
biomarkers are probably best viewed as enrichment biomarkers, which can be used in early clinical trials for
selecting patients with tumors having molecular characteristics that make them somewhat more likely to respond.
It is clear that much more work needs to be done to
identify, validate, and clinically qualify biomarkers that
may be truly predictive. Additional biomarkers may well be
needed for drugs with different PI3 kinase-selectivity profiles. For example, tumor cells harboring activating mutations in the PIK3CA gene are potentially dependent on or
addicted to this isoform, whereas cancer cells deficient in
PTEN expression can exhibit a dependency on p110.1
As a further complication, it should also be considered
that in addition to their direct therapeutic action on cancer
cells, PI3 kinase inhibitors can exert effects on tumor angiogenesis, immune cells, and other tumor microenvironmental
interactions. Hence, it is quite conceivable that there may
e95
WORKMAN AND CLARKE
Fig. 2. Structural biology of PI3 kinase and inhibitors. The solution of the X-ray crystal structures of PI3 kinases in complex with ATP or
inhibitors acting at the ATP site has facilitated the understanding of their potency and selectivity and educated the prospective design of new
ones. a) Ribbon diagram of human p110 (protein data bank [PDB] code 3dbs) illustrating the five-domain structure of the enzyme. The
Ras-binding domain (RBD) domain is shown in red, the C2 domain is depicted in yellow, the helical domain is illustrated in green, and the bilobal
catalytic domain is displayed in purple. The N-terminal adaptor binding domain (ABD) preceding the RBD-domain is shown in orange. The
ATP-binding site is located in between the N- and C-terminal lobes of the catalytic domain, as indicated by the arrow. b) Binding of ATP to porcine
p110 (PDB code 1e8), showing the hydrogen bond interactions of ATP with the hinge region. The protein is shown in pink and ATP is displayed
in light blue. The hydrogen bonding interactions are shown as dotted lines. Shown in gray are the two metal ions coordinating the phosphate
groups of ATP. c) Binding of LY294002 to porcine p110 (PDB code 1e7v) illustrates the crucial hydrogen bonding interaction (dotted line)
between the oxygen of morpholino group of this weakly potent and relatively unselective inhibitor and the PI3 kinase hinge region. The protein
is shown in blue with key PI3 kinase residues displayed and labeled. LY294002 is shown in orange. Despite the fact that LY294002 extends into
the affinity pocket (right-hand side), it does not fill the pocket so efficiently as does the indazole portion of the GDC-0941 molecule (see d).
d) Binding of GDC-0941 to human p110 showing the hydrogen bond interactions (dotted lines) that anchor into the inhibitor deeply in the
ATP-site. p110 is displayed in purple, and GDC-0941 is depicted in light green. e) Ribbon diagram of the p110/p85 structure, showing p110
in yellow and the p85 niSH2 domain in blue (PDB code 2rd0). The ATP from the ATP-bound p110 structure superimposed on the p110/p85
complex is shown. ATP is depicted in cylinder representation with its semi-transparent molecular surface superimposed in blue. The oncogenic
mutation hotspots Glu 545 and His1047 are highlighted in red. f) Superposition of the GDC-0941-bound p110 structure and the human
p110/p85 structure. p110 is displayed with its solvent-accessible surface colored yellow and GDC-0941 molecule is shown with its
semi-transparent surface in light green. It can be seen that based on this superposition, the GDC-0941 fits extremely well into the p110 ATP site,
showing only a minor clash of the indazole ring with the wall of the ATP site (right-hand side). Figure reprinted with permission from Workman
and colleagues 2010.13
not be a single biomarker of sensitivity but instead a

predictive molecular signature, perhaps involving a geneexpression profile. Studies in preclinical systems, including
large molecularly characterized cancer-cell panels24-26 and
human tumor xenografts, together with genetically engineered mouse models, may contribute to the definition of
candidate biomarkers. However, it is likely that clinically
effective predictive biomarkers will emerge from molecularpathology profiling of clinical tumor materialincluding
global cancer genome sequencing, gene expression analysis,
e96
and proteomicsand the correlation of these with therapeutic response and outcome. In addition, continued longitudinal profiling will also be essential to understand the basis for
mechanisms of drug resistance that, based on experience
with protein kinase inhibitors, is likely to develop with
prolonged exposure to inhibitors.
Another major challenge that we now face is the identification and development of rational mechanism-based combinatorial treatments for optimal therapeutic efficacy with
PI3 kinase inhibitors. In general, PI3 kinase inhibitors are
PI3 KINASE IN CANCER
Fig. 3. In vivo oral antitumor activity

and associated pharmacokinetic and pharmacodynamic properties of GDC-0941 in
the PTEN null U87MG human glioblastoma
xenograft model. a) Chemical structure of
GDC-0941. b & c) Relative mean tumor volume (percentage of initial volume before
therapy) and mean final tumor weights following doses of 25, 50, 100, and 150
mg/kg daily oral doses of GDC-0941 administered for 19 days. Tumors were
grown subcutaneously bilaterally in female NCr athymic mice, and controls received equivalent volumes of drug vehicle.
d) Tumors from the same efficacy study
were taken 4 and 8 hours following the
final dose. AKTSER473 phosphorylation and
total AKT were measured by a quantitative
electrochemiluminescent
immunoassay. For therapy data, results are mean
standard error (SE) of 16 mice. For pharmacodynamic target engagement biomarker
data, results are mean SE of 6 mice.
Adapted with permission from Raynaud
and colleagues 2009.17
not strongly proapoptotic but rather induce a strong and

durable effect on cell-cycle progression.15,24 More profound
therapeutic activity is likely to require combinatorial treatments to facilitate the death of cancer cells. Many combination studies are underway, but because of the sheer number
of potential combinations, it will certainly take some considerable time to identify optimal combinations; both preclinical and clinical studies will contribute to this. Certain
rationally based combinations, for example, using PI3 kinase together with MEK inhibitors that block the important
RAS-RAF-MEK MAP kinase pathway, have obvious mechanistic rationale and these are being given a high priority in
clinical studies.25 Both horizontal pathway targeting and
also vertical targeting within the PI3 kinase pathway are
being investigated. Another interesting opportunity is the
combination of PI3 kinase inhibitors with agents targeted to
the androgen receptor in prostate cancer.29 In addition to
combinations with other molecularly targeted agents, use of
PI3 kinase inhibitors together with cytotoxic agents is also
being pursued.
The picture that we currently see emerging from the clinic
is that PI3 kinase inhibitors are generally well-tolerated
agents and show promising signs of biologic and clinical
activity. Concerns about potential effects on glucose metab-
olism appear to have been alleviated, with only mild and

manageable changes being seen, at least with the doses and
schedules used to date.27 The main side effects commonly
include skin rash and urticaria. Encouragingly, there is
early evidence of single agent therapeutic activity in the
current clinical trials in patients with cancer,30 although
stable disease is often seen and it may require combination
studies to reveal the full therapeutic potential.
Concluding Remarks
The genetic validation provided by the frequency of PI3

kinase pathway deregulation in human cancers and evidence from mouse models encourages continued efforts to
pursue inhibitors of PI3 kinases. The dramatic increase in
the number of PI3 kinase inhibitors that display fascinatingly distinct selectivity profiles now provides us with many
powerful tools with which to further investigate disease
mechanisms and assess therapeutic potential in different
pathogenetic contexts. Continued efforts to define clinically
useful predictive biomarkers for patient stratification, together with rational drug combination studies, will allow us
to uncover the full clinical potential of PI3 kinase inhibitors
in cancer treatment.
Author
Paul Workman
Paul Clarke
Employment or
Leadership
Positions
Consultant or
Advisory Role
Chroma
Therapeutics;
Nextech Invest;
Piramed (Roche);
Wilex
Stock
Ownership
Piramed (Roche)
Honoraria
Piramed Pharma
Research
Funding
Piramed (Roche);
Yamanouchi
Expert
Testimony
Other
Remuneration
Institute of
Cancer
Research
e97
WORKMAN AND CLARKE
REFERENCES
1. Vanhaesebroeck B, Guillermet-Guibert J, Graupera M, et al. The emerging mechanisms of isoform-specific PI3K signalling. Nat Rev Mol Cell Biol.
2010;11:329-341.
2. Yuan TL, Cantley LC. PI3K pathway alterations in cancer: variations on
a theme. Oncogene. 2008;27:5497-5510.
3. Liu P, Cheng H, Roberts TM, et al. Targeting the phosphoinositide
3-kinase pathway in cancer. Nat Rev Drug Discov. 2009;8:627-644.
4. Chalhoub N, Baker SJ. PTEN and the PI3-Kinase Pathway in Cancer.
Annu Rev Pathol Mech Disease. 2009;4:127-150.
5. Samuels Y, Diaz LA Jr., Schmidt-Kittler O, et al. Mutant PIK3CA
promotes cell growth and invasion of human cancer cells. Cancer Cell.
2005;7:561-573.
6. Vogt P, Gymnopolous M, Hart JR. PI 3-kinase and cancer: changing
accents. Curr Opin Genetics and Dev. 2009;19:1-6.
7. Wee S, Wiederschain D, Maira SM, et al. PTEN-deficient cancers depend
on PIK3CB. Proc Natl Acad Sci U S A. 2008;105:13057-13062.
8. Jia S, Roberts TM, Zhao JJ. Should individual PI3 kinase isoforms be
targeted in cancer? Curr Opin Cell Biol. 2009;21:199-208.
9. Billottet C, Grandage VL, Gale RE, et al. A selective inhibitor of the
p110 isoform of PI3-kinase inhibits AML cell proliferation and survival and
increases the cytotoxic effects of VP16. Oncogene. 2006;25:6648-6659.
10. Shuttleworth SJ, Silva FA, Cecil AR, et al. Progress in the preclinical
discovery and clinical development of class I and dual class I/IV phosphoinositide 3-kinase (PI3K) inhibitors. Curr Med Chem. 2011;18:2686-2714.
11. Arcaro A, Wymann MP: Wortmannin is a potent phosphatidylinositol
3-kinase inhibitor. The role of phosphatidylinositol 3,4,5-trisphosphate in
neutrophil response. Biochem J. 1993;296:297-301.
12. Vlahos CJ, Matter WF, Hui KY, et al. A specific inhibitor of phosphatidylinositol 3-kinase, 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one
(LY294002). J Biol Chem. 1994;269:5241-5248.
13. Workman P, Clarke PA, Raynaud FI, et al. Drugging the PI3 kinome:
from chemical tools to drugs in the clinic. Cancer Res. 2010;70:2146-2157.
14. Hayakawa M, Kaizawa H, Moritomo H, et al. Synthesis and biological
evaluation of pyrido[3,2:4,5]furo[3,2-d]pyrimidine derivatives as novel PI3
kinase p110 inhibitors. Bioorg Med Chem Lett. 2007;17:2438-2442.
15. Raynaud FI, Eccles S, Clarke PA, et al. Pharmacologic characterization
of a potent inhibitor of class I phosphatidylinositide 3-kinases. Cancer Res.
2007;67:5840-5850.
16. Hayakawa M, Kaizawa H, Moritomo H, et al. Synthesis and biological
evaluation of 4-morpholino-2-phenylquinazolines and related derivatives as
novel PI3 kinase p110 inhibitors. Bioorg Med Chem. 2006;14:6847-6858.
17. Raynaud FI, Eccles SA, Patel S, et al. Biological properties of potent
inhibitors of class I phosphatidylinositide 3-kinases: From PI-103 through
PI-540, PI-620 to the oral agent GDC-0941. Mol Cancer Ther. 2009;8:17251738.
e98
18. Folkes AJ, Ahmadi K, Alderton WK, et al. The identification of 2-(1Hindazol-4-yl)-6-(4-methanesulfonyl-piperazin-1-ylmethyl)-4-morpholin-4-ylthieno[3,2d]pyrimidine (GDC-0941) as a potent, selective, orally bioavailable
inhibitor of class I P13 kinase for the treatment of cancer. J Med Chem.
2008;51:5522-5532.
19. Sutherlin DP, Bao L, Berry M, et al. Discovery of a potent, selective,
and orally available class I phosphatidylinositol 3-kinase (PI3K)/mammalian
target of rapamycin (mTOR) kinase inhibitor (GDC-0980) for the treatment of
cancer. J Med Chem. 2011;54:7579-7587.
20. Friedman L. GDC-0941, a potent selective orally bioavailable inhibitor
of class I PI3K. Proc American Assoc Cancer Res 2008;LB-110.
21. Yap TA, Workman P. Exploiting the cancer genome: Strategies for the
discovery and clinical development of targeted molecular therapeutics. Annu
Rev Pharmacol Toxicol. 2012;52:549-573.
22. Clarke PA, Workman P. Phosphatidylinositide-3-kinase inhibitors:
Addressing questions of isoform selectivity and pharmacodynamic/predictive
biomarkers in early clinical trials. J Clin Oncol. 2012;30:331-333.
23. Yap TA, Sandhu SK, Workman P, et al. Envisioning the future of early
anticancer drug development. Nat Rev Cancer. 2010;10:514-523.
24. Barretina J, Caponigro G, Stransky N, et al. The Cancer Cell Line
Encyclopedia enables predictive modelling of anticancer drug sensitivity.
Nature. 2012;483:603-607.
25. Garnett MJ, Edelman EJ, Heidorn SJ, et al. Systematic identification
of genomic markers of drug sensitivity in cancer cells. Nature. 2012;483:570575.
26. Workman P, Clarke PA, Al-Lazikani B. Personalized medicine: Patientpredictive panel power. Cancer Cell. 2012; in press.
27. Guillard S, Clarke PA, Te Poele R, et al. Molecular pharmacology of
phosphatidylinositol 3-kinase inhibition in human glioma. Cell Cycle. 2009;
8:443-453.
28. Hoeflich KP, Merchant M, Orr C, et al. Intermittent administration of
MEK inhibitor GDC-0973 plus PI3K inhibitor GDC-0941 triggers robust
apoptosis and tumor growth inhibition. Cancer Res. 2012;72:210-219.
29. Carver BS, Chapinski C, Wongvipat J, et al. Reciprocal feedback
regulation of PI3K and androgen receptor signaling in PTEN-deficient prostate cancer. Cancer Cell. 2011;19:575-586.
30. Bendell JC, Rodon J, Burris HA, et al. Phase I, dose-escalation study of
BKM120, an oral pan-class I PI3K inhibitor, in patients with advanced solid
tumors. J Clin Oncol. 2012;30:282-290.
31. Vasudevan KM, Arbie DA, Davies MA, et al. AKT-independent signaling downstream of oncogenic PIK3CA mutations in human cancer. Cancer
Cell. 2009;16:21-32.
32. Forbes SA, Tang G, Bindal N, et al. COSMIC (the Catalogue of Somatic
Mutations in Cancer): A resource to investigate acquired mutations in human
cancer. Nucleic Acids Res. 2010;38:D652-D657.
This publication is supported by an educational

donation provided by:
Support for this program is funded through

2012 Asco

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2012 Asco

Uploaded by

Copyright:

Available Formats

AMERICAN SOCIETY OF CLINICAL ONCOLOGY

2012 EDUCATIONAL BOOK

American Society of Clinical Oncology

American Society of Clinical

Editor: Ramaswamy Govindan, MD

2012 American Society of Clinical Oncology

ASCO gratefully acknowledges Amgen

American Society of Clinical

How Does Biology Affect Local Therapy Decisions?

New Looks and Challenges for Cooperative Groups

Gastrointestinal (Colorectal) Cancer

Head and Neck Cancer

Research and Standard of Care: Lung Cancer in Romania

Practice Management and Information Technology

2012 ASCO Annual Meeting Disclosure

Employment or Leadership Positions

Continuing Medical Education Information

20112012 Cancer Education Committee

Gastrointestinal (Colorectal) Cancer

Lymphoma and Plasma Cell

2012 Annual Meeting Faculty List

Abenaa M. Brewster, MD, MHS

Howard Colman, MD, PhD

Pamela S. Hinds, PhD, RN

Hedy Lee Kindler, MD

Henry Orlando Otero, MD

Neil P. Shah, MD, PhD

2012 Educational Book Expert Panel

Mary Lou Smith, JD, MBA

2012 Annual Meeting Supporters*

CONQUER CANCER FOUNDATION MISSION ENDOWMENT SUSTAINING DONORS

Lilly USA, LLC

Millennium: The Takeda Oncology Company

ASCO and Conquer Cancer Foundation Boards of

Doris Duke Charitable Foundation

ASCO Cancer Research Committee

ASCO Clinical Practice Committee

Estate of Sandra Syms

National Cancer Institute

Letter from the Editor

he theme of the 2012 American Society of Clinical

Girard N. Thymoma: From Chemotherapy to Targeted

CONTROVERSIES IN THE ADJUVANT TREATMENT

Adjuvant Therapy for Older Women with

Overview: Breast cancer is a disease of aging. However,

Evaluation of the biology of breast cancer by patient age

cancer death in the general population decreased 2.5% per

OWUSU, HURRIA, AND MUSS

Adjuvant systemic therapy refers to the administration of

The undertreatment of older women with breast

The majority of older patients present with hormone

TREATMENT FOR OLDER WOMEN WITH BREAST CANCER

should reinforce the importance of adherence to maximize

strategy is a reasonable approach. For women with low

Cytotoxic chemotherapy can be considered for older

OWUSU, HURRIA, AND MUSS

with early-stage breast cancer were randomly assigned to

individualization of cancer treatment, it is indeed a useful

TREATMENT FOR OLDER WOMEN WITH BREAST CANCER

Chemotherapy with trastuzumab

Chemotherapy with trastuzumab

patients younger than 65. Regardless, in the absence of

Until recently, the guideline recommendation, regardless